# EPIC Online Technology Meeting on Optical Metrology Solutions for the Industry

https://www.youtube.com/watch?v=TSe3ZIiUZmc

[00:43] So good afternoon everybody and welcome to this new online technology meeting.
[00:48] My name is Antonio Castello, I'm technology manager and epic.
[00:52] I'm I'm very happy to be today here to talk about Optical Metrology solutions for the industry.
[00:59] We have a very nice set of speakers to talk about different Technologies and also different applications.
[01:07] We are going to cover uh Metrology for different kind of materials, different kind of products and I think it's going to be a quite interesting interesting um uh meeting.
[01:12] today just a few words at the beginning.
[01:15] about epic in case uh there is somebody in the room that doesn't know epic.
[01:17] we are the European photonics industry Consortium uh and as you can see in the screen we are Association of more than 800 uh companies and research centers that are somehow related with the photonics business and our goal is to connect people to ensure the sustainability of the photonics business in in Europe uh.
[01:36] we offer different benefits to our members so we help with their technological requests uh.
[01:41] we also help us in this meeting to uh uh have good networking activities and connections with companies from the whole supply chain uh.
[01:50] we have also with um market reports and access to different markets uh and also collaborate with uh website infot tonics uh for human resource support and of course we put in contact investors with companies spinoffs and startups that could need the money for for growing and for the next activities uh.
[02:08] just a short summary about
[02:12] the future events um as you know we started uh with a lot of physical events in 2023.
[02:17] um there are still some of them uh in this uh last part of the year.
[02:24] uh we have one in toptica Munich about the quantum photonics technology.
[02:26] another one about laser application for the battery battery manufacturing processes in stutgart.
[02:33] um also microelectronics and photonics and two two size of the GU coin in Munich and the photonics assist cancer pathology and surgery in ANP in Belgium.
[02:43] and combine it also with members delegation trip uh that we still continue doing in different countries.
[02:50] and also an investors meeting in in ovin in Netherlands for all the eping members that want to be connected with investors.
[02:57] and also some offline Meetings online meetings that that we will continue doing.
[03:00] um some meetings in in uh related with technology as you can see Li on chips Optical design simulators and new meeting for developments uh and we continue doing also marketing meetings.
[03:12] uh also technology reports like this one.
[03:15] in freeform Optical system that I guess.
[03:17] could be interesting for uh some of you.
[03:19] uh and also in sustainable photonics so.
[03:21] I encourage you to go to our website and check our next event and register for the ones that are interesting for for you or uh come back to me if you want to discuss the details of any of them uh.
[03:32] we are also involved in different European projects and you can see in this slide and in particular for this meeting we have the collaboration of two of them uh.
[03:40] methab that is a pilot line that is dedicated to help companies in the uh to advance the developments of medical devices based on photonics Technologies uh.
[03:49] and also pulsate that is a network that help companies to um try or to implement laser based advanc and additive Manufacturing in the processes uh.
[04:00] so if any of these cases are interesting for you uh please also let us know so we can put you in contact with both uh networks and let's go with the meeting today as I mentioned Optical Metrology is a very interesting topic it covers I
[04:13] think most of the industries nowadays from semicon Optics micro Optics Automotive metrologies everywhere uh with inline measurements ofline laboratory R&D uh and first we want to thank our sponsors because all these uh meetings that we organize um they need also the support of eping members so today we want to thank uh the sponsors actor Advan Coatings also ENT Optics and also modulite um so now I'm going to introduce you first taras from Ming Optics that he wants to say a few words about his company so Tas whenever you're ready you muted Tas yes thank you very much it's pleasure to be here uh hello to everyone it's both a pleasure and privilege to support the Epic events like this one um I will grab a few moments of your
[05:14] time to briefly explain what we do and how we do this.
[05:17] we are 14 years on the market uh designing and building spectal 4 Mets specifically to measure codings.
[05:23] inspecting codings for Trans and reflection.
[05:30] uh we do this both on flats and lenses and to shorten my presentation I will focus today only on Flats.
[05:40] so on this slide on the left hand you see the instrument which is called Photon RT Spectrum photo meter.
[05:48] this is a most hot selling system on our our portfolio.
[05:51] we were able to place six different configurations in one body uh relative to to the wh range.
[06:00] so the shortest range is from 380 to 1700 NM and the widest range is the most knowledge intensive.
[06:08] covering uh the we range from 185 to 5.2 Micron and we do this uh for
[06:16] transmission for reflection for Flats cubes prisms from 0 to 75 degrees angle of incidence measuring s and average all vales are absolute.
[06:26] um unique feature is that we are able to make measure really small samples like four nanometers a standard.
[06:33] but this year we were able to measure the sample with just 700 microns physical Dimension and on the right hand you see the system which was introduced this year in munic uh designed specifically for longwave infrared measurements and this is not FTI based technology this is a truly dispersive type system with gradings and polarizers and side same case fully automatic capable to measure from 0 to 60° angle of inance also s and average polarization.
[07:12] um for this system the quite unique feature is that we can place here a sample like your palm that
[07:19] size um all systems have very big advantage of very small beam Divergence.
[07:26] so the beam is really narrow uh fully automatic operation very high accuracy.
[07:30] measurement results are being delivered by these systems and if to Antonio you can switch over to the next slide.
[07:38] I think that values they um speak more than any words so these are the two examples of real measurements.
[07:48] on the upper part of your slide you see Broadband laser mirror reflection measurement and the yellow area highlighted is shown on the right hand.
[08:00] you see the scale is just from 95 to 100% very low noise very high accurate measurements no values exceed 100%.
[08:09] and on the bottom part of your slide you see again laser mirror but the customer requested to measure transmission and the challenge.
[08:19] was to measure at 42 44 and 46 degrees
[08:23] angle of incidence S&P
[08:25] polarization and you see on the Right
[08:29] image
[08:30] the area is scaled from 0 to just
[08:33] 0.8%
[08:34] transmission you can see the very good
[08:38] resolution of each Spectra and very low
[08:40] noise which is standard in our system
[08:44] these are the Technologies which we are
[08:45] delivering to our customers worldwide
[08:48] already for 14
[08:51] years thank you very much for the time
[08:54] given to speak a little bit about our
[08:56] company thank you Tas and any anybody
[08:59] interested you can uh reach him here on
[09:02] the chat and also contact us later on so
[09:04] you can be connected to to Tas and I was
[09:08] our next U sponsor today is Mite and I
[09:11] think Tommy is uh here in the room so
[09:14] you thank you very much Antonio so my
[09:18] name is Tom Mulin I work as a product
[09:20] manager here at modulite so modulite is a laser manufacturer based in Tampere Finland with sales process in the US us.
[09:30] the company was founded in year 2000 so we have over 20 years of experience in various laser Technologies such as laser diets fiber lasers DPSS and vexels.
[09:43] in 20121 we did a successful IPO and got listed in NASDAQ first North Helsinki.
[09:50] we are probably best known for our work in life sciences and more specifically in our key strategic areas oncology.
[09:59] oftalmology genetics and Diagnostics and we are the only vertically Integrated Medical laser manufacturer in the world from Chip level to system.
[10:10] but how this relates to uh today's agenda um our product line ml 6600 is a flexible subsystem platform with wavelength range spanning from UV.
[10:23] to2 Micron so that enables wide range of applications in the context of optical meterology.
[10:30] it is used as a light engine so PLU and play Laser Source integrated with all the necessary electronics and optionally also beam shaping Optics relevant.
[10:44] applications are semiconductor material analysis thin fi Metrology and waer DET defect detection to name a few.
[10:50] a unique differentiator of this product line is cloud connectivity which enables for instance predictive maintenance but also new business models like paper use and should you have any questions.
[11:03] I'll share my contact details in the chat so you can reach me afterwards.
[11:08] thank you very much thank you Tommy and our was last sponsor today uh we have ACM that they cannot be here due to some schedule conflict but they asked me to
[11:24] to present this slide for them uh ACM is the German sary of actor and as probably most of you know actor is an Israeli company they specializ in the development and applications of ultra back and light absorbing Coatings and also coated foils uh with three main applications for the products suppress straight light absorb laser power and also create high emissivity surfaces so if you're are interested in this kind of light and sorbing and Ultra coins uh please let us know to put you in contact with um actor group and ACM and let's go with the uh group of companies that we are present today as you can see we have a qu competed supply chain with uh companies from the different um perspectives and point of view uh from the whole supply chain and it was quite interesting because I saw at the beginning when uh preparing this slide that um of course we have companies present that are Metrology manufacturers that they are happy to
[12:24] reach and users also to reach uh perhaps coding and filters companies so they can supply the products for the characterization but also a lot of interaction between them and the Optics and mitro Optics manufacturers as they can use the product for the characterization but also develop Optical components for some of them uh the same with the R&D groups uh because some of R&D groups can be users but also can be developers of some new technology so I see a lot of interactions and of course with the final equipment as we have some companies that are um working in manufacturing uh devices for different IND industry and Metrology is also very important for them so as you see a lot of synergies today so I hope the the presentations of our speakers are going to be as as interesting as I I I discuss with them and let's go with the agenda uh so we will start with a presentation from T Shar s micro Optics then we will have T Steel Frank vanderbeck Imaging optic Rafael porar thetis Kevin contas then AC Optics Andre
[13:27] dentes Mur Dei from opto Fidelity and we
[13:31] will close with autocrat and Christian
[13:33] Brooks so uh to I think you can share
[13:36] your screen you are quite far in Asia in
[13:39] Hong Kong so thank you for being here so
[13:41] late for you uh I will stop sharing my
[13:43] screen and you can the floor is
[13:48] yours thank you Antonio for this uh
[13:53] production and welcome everybody
[13:57] everybody oops just one second here Al
[14:00] though I'm not sure what you see now
[14:03] anymore maybe like this now it's good
[14:06] thank you for this introduction uh I'm
[14:08] happy that I can present a little bit
[14:09] insight into the Metrology aspects and
[14:12] problems we face when we do mitro micro
[14:15] optic characterization I think you know
[14:18] probably already a little bit there
[14:20] microoptics as a player and microoptics
[14:23] Fabrication on different markets so we
[14:25] are we are producing classic micro optic
[14:29] components like micr lenses beam Shapers
[14:32] in F silica in etching technology we produce also printed Optics on glass uh
[14:34] polymer on Glass Technology we put we have a portfolio on defra of optical elements and we also go a little bit now into two and a half Dimensions with more functionalities like you see on the left lower corner with prism lenses every recall it
[14:50] so this is a mixture of lenses on one side and prisms on the other side to change directions of the different um light path
[14:59] we produce in uh Switzerland uh with um with more than 130 people and uh in our own FB uh and we um try to grow as fast as we can uh to serve the market with different objects
[15:19] so if you go a little more technical this what Antonio has asked for uh our main um dimensions are shown a little bit on
[15:29] this slide here we have usually lenses.
[15:31] and we have sometimes as you have seen prisms or steps which we need to characteriz and to measure where we of course want to have very high uh uh quality of the measurement in surface roughness measurements and shape regularities and all these kind of things.
[15:44] where sometimes you break this thing down to Optical functions uh sometimes you break it down to surfaces very often the surfaces which we have to characterize um while you see here the different parameters.
[15:57] usually we don't go larger than 5 mm with the lenses usually it's 2 mm and Below uh and we start at about 20 and we have height limitations for different Technologies which does not exceed 500 Micron and eding we usually go not higher than something like 5080 micrometer for instance.
[16:16] we realiz sperical aspherical uh surfaces as a stard product and some free form surfaces as a little bit of special product for our customer which causes of course then.
[16:29] trouble in the characterization you will see this in a second.
[16:33] how this should look like um don't want to get into give you a few examples of these structures later so I don't go over the whole details of this slide here.
[16:43] so when we do a measurement of our surface usually the client is specifying the surface and not the function so we need to measure the surface.
[16:51] so we have instruments which can do this uh which creates usually lot of data because we have one waer with we work on 200 mm FB we usually have something like three 30,000 lenses more or sometimes less but this is a rough estimation uh and we usually have structures which are getting more and more challenging in slope and lateral resolution.
[17:13] so the small structures of course are challenging lateral resolution the other ones in slope and because you're doing micro Optics many of these instruments and of the things we are using go with a magnification and uh like in the microscope uh and so this causes of
[17:31] course a lot of trouble once you have the measurement of the surface you still need to have data preparation.
[17:39] you still need to have a kind of communication Channel with your customer which is usually an ideal surface which he has defined and then you try to fit this and see the deviations from the surface uh which means that you discuss with your customer the surface parameters and regularities uh and this is actually the agreement which you use usually have on delivering a good or bad sample and of course all this and this is the most important part sometimes should be done on a level where we have calibration standards where we have a well-known standards even already for the measurement principle itself because otherwise we cannot certify our measurement technology due to ISO or even erf standards which is necessary in our case because we are erf certified um think this gives you a little bit the outline how this works usually.
[18:32] um so if we look what would be the ideal measurement to of today because we need to measure surface shapes it should be a tool which has no contamination risk so actually best is non non touching system.
[18:47] it should be of course fast because we have many objects to measure which get more and more challenging because we have increasing numbers of objects customer wants to have a design of everything.
[18:59] so something like 1 second per 1 mm Square fet would be kind of an acceptable value of course large feet of few is necessary because the objects get bigger and bigger and we don't want to stitch actually too much because this is very slow one very important fact is that we need to attack a more and more High slope angles so it gets more on 45 degrees and even bigger slope angles gets a very uh uh challenge situation because with standard microscope observation technique you will not be able to see this very easily I give you an example of how this looks like in a
[19:32] second and a recent development on our side is that we all push now for the 300 mm cases.
[19:37] so we want to go on 300 mm surface substates uh which means that you also have to have high Precision tables to manipulate your object to move it and to set it precisely.
[19:50] all this and of course very very low noise of the measurement system itself because if you want to measure 5 nanometers depths of course your noise has to be lower than this.
[19:58] has to be extremely reproducible you don't want to have limitations on the surface means that material should not play a certain role which you can imagine as a problem for instance for touching instruments because the tip touches different materials in different way and can lead to problems if you have a scanning profile for instance.
[20:19] course we want to have a calibration which is if you are lucky automatized so we press a button and self calibration all over and uh what we do today that we have our own evaluation software which gives us freedom in operation but of course it would be much
[20:33] more convenient to have something which
[20:34] you can just buy and which gives you
[20:37] kind of certified output yeah and of
[20:39] course everything should be very easy to
[20:41] use because we don't want to think much
[20:43] if you have to think then usually we are
[20:45] lost somewhere in Translation how we say
[20:48] it in some of the
[20:51] movies today we use all kind of
[20:55] Technologies which you see here in
[20:57] firetry qu focal scanning stylus
[21:00] profilometry and we have coordinate
[21:02] measurement systems so this in this
[21:03] combination with all these we can we can
[21:05] survive and can
[21:07] do uh develop a good measurement
[21:11] methodology uh some of these things are
[21:13] very precise inter defer usually getes
[21:15] very precise results we have a in Comal
[21:18] scanning you have a kind of relatively
[21:19] cheap system it's not as good as it
[21:22] could be probably for the very high
[21:24] slopes we go with stylus profilometer
[21:26] and then for the coordinate measurement
[21:28] system is just for positioning actually
[21:30] this is what we usually use uh is a
[21:33] normal microscope Imaging system this
[21:36] image processing think you are aware of
[21:38] this kind of
[21:40] techniques I give you two examples of of
[21:44] cases where where we have tried to
[21:46] compare a little bit the different uh
[21:48] instruments you see here the different
[21:50] instruments which we have used for this
[21:53] um and we I dra an example of a
[21:55] reference
[21:56] sphere um you see here on the we start
[22:00] on the right side with this
[22:02] [Music]
[22:03] Graphics maybe I can even go for this
[22:07] laser printer here we start with this
[22:08] Graphics here you see the different
[22:10] fields of measurement which we have here
[22:12] the the the dark one is the um Al
[22:19] measurement actually this is a stylus
[22:20] profile meod and we have a 20x objective
[22:24] for the conal scanning and then we have
[22:26] here something is inter which is a
[22:28] stitched measurement on a certain field
[22:30] here and uh and out of this data you can
[22:34] go back to your original data of the
[22:37] sphere which is here sphere with a rate
[22:39] of cature of 1 mm and you see the
[22:41] different quality of the results here so
[22:44] the next few usually provides very very
[22:46] good arms values because it's very well
[22:48] calibrated and maintained the tal surve
[22:51] has a little bit less um good arm is but
[22:55] it's a very very slow measurement system
[22:58] and uh the confocal scanning provides a
[23:00] very fast system but also has a kind of
[23:03] relatively bad measurement quality and
[23:05] is not able to reach the high U the high
[23:08] angles and you see here also of course
[23:10] the the uh the best measurement system
[23:13] which is only measured this graet misses
[23:15] a lot of information which is hidden
[23:16] here on the the the other side you see
[23:19] here a little bit the residuals which we
[23:22] can achieve in a plot you see here there
[23:25] for sure some kind of regession problem
[23:28] in the uh in the conf scanic system this
[23:32] one less and here you see this is
[23:34] equality with a very Co sampling
[23:37] actually of course of the tal system um
[23:41] so uh of course the best would be to
[23:44] have a system which can cover everything
[23:46] in a very fast way but doesn't exist you
[23:48] see this is already a kind of problem
[23:49] here to compare these systems is not
[23:51] actually possible because you don't have
[23:53] the same information
[23:55] available to give you a second example
[23:57] where we we investigated the possibility
[23:59] of measure extreme slope angles so here
[24:02] we have a system which is which works in
[24:05] transmission on digital loc graphy so
[24:07] you Chine light through uh you
[24:09] reconstruct the wavefront and then you
[24:11] can figure out what is going on on your
[24:13] sample this kind of techn technique is
[24:16] not U um traceable back it easily to
[24:20] Conventional uh um standards and it is
[24:23] very difficult to um to make a kind of
[24:27] ification for this kind of measurement
[24:29] doesn't really exist EO norms for this
[24:31] kind of thing at least as far as I know
[24:35] um and it is limited to transparent
[24:38] surfaces and it is a kind of functional
[24:40] test but not really a surface
[24:41] measurement so we have used this in a
[24:44] very special setting to realize a
[24:46] measurement which is really can be
[24:48] considered as an extreme measurement you
[24:50] see here if you you can use these kind
[24:52] of systems in immersion so you put your
[24:55] sample in an immersion field here and
[24:57] what happens is that the deviation of
[24:59] the ray is of course reduced because of
[25:01] the immersion if you have an information
[25:03] about the refractive index of your
[25:05] immersion liquid you can measure now
[25:08] much much larger angles because of this
[25:10] kind of setting and you see here a
[25:11] measurement of a structure which we got
[25:15] from our project partner tum duble in
[25:17] the fabulous project which had a 75
[25:19] degrees slope angle here which we could
[25:22] still reach yeah what you have as a
[25:24] disadvantage on this kind of setting is
[25:26] that it's not really possible to
[25:28] calibrate it very difficult to get
[25:30] absolute measurements you can maybe use
[25:31] it for inspection but not for
[25:33] measurement and the measurement noise
[25:35] also is Amplified with the um with the
[25:37] with the measurement so this is a little
[25:39] bit of a problem but at least you can
[25:41] see what your surface looks like and can
[25:44] access
[25:45] it you go a little bit deeper into the
[25:48] uh uh free form Optical surfaces give
[25:51] you a few arguments but it will be very
[25:54] difficult to to measure free from
[25:55] Optical surfaces on a micro Optical
[25:58] application because very often what we
[26:00] use to describe the surface is a
[26:03] parameterized version of the surface so
[26:04] that we measure four parameters and not
[26:06] have the full surface to be
[26:08] characterized
[26:10] for aspherical surface for instance we
[26:13] go for rock and conic and then tilts
[26:17] maybe but there is a limited set of
[26:19] parameters if you have a free from suace
[26:21] this the set of limited parameters gets
[26:23] very big and you have a in numbers and
[26:27] that's get of very very
[26:28] problematic uh the densely packaging of
[26:32] three form sometimes you don't even find
[26:33] the the structure where you want to
[26:35] measure um what what it means that the
[26:39] that you cannot break it down to
[26:41] identical Optical units which you can
[26:43] then compare so it's a kind of thing
[26:45] which is is has a distributed function
[26:48] over large field and it's very difficult
[26:50] to to to measure this because then would
[26:53] need would mean that you have fields of
[26:56] measurement of about 10 Mill MIM and you
[26:58] need still the nanometer Precision slope
[27:00] angles is always a problem and um and as
[27:04] I said the function is distributed over
[27:06] the whole area and it's very difficult
[27:09] to bring this thing all together to then
[27:11] we say what you measure this is a very
[27:13] challenging situation uh for the moment
[27:17] I think there's no really perfect
[27:18] solution to it only for very very flat
[27:21] um slope angles you can you can probably
[27:23] attack it but then still you are stick
[27:26] with something like a millimeter range
[27:27] of
[27:29] of filter VI while the element usually
[27:31] gets bigger and bigger to 10 mmers you
[27:32] see an example here of some kind of free
[27:35] form machined surfaces which we realized
[27:39] with our partner upmt upmt Vance in so
[27:44] to come to the end of this
[27:47] presentation it's um the variety of the
[27:50] different object makes it very
[27:52] complicated to to to work with a single
[27:54] instrument for the moment doesn't exist
[27:57] f a few slope angle limitations is well
[27:59] known I don't need to tell you something
[28:01] about this but we have realized that
[28:02] operations of the objective is a really
[28:04] really uh challenging problem especially
[28:08] if you want to go to the best quality
[28:09] ever so here you need very very careful
[28:12] calibration and very fair deal every
[28:14] time you measure something in
[28:18] Aras uh usually you can um measure
[28:21] relatively fast uh and and repetitive so
[28:25] that's not really a big problem but if
[28:26] the area has a distributed function and
[28:29] gives gets nonhomogeneous and it gets of
[28:31] course much more complex um for the
[28:34] moment there's no tool which can do
[28:35] everything so we mix the tools and we
[28:37] work on the methodology to have the best
[28:40] uh outcome the um some of the samples I
[28:44] told you already we realized with in
[28:45] fabulous actually and I was just to
[28:47] mention this here so this is the P line
[28:49] for uh anufacturing of free four
[28:52] microoptical structures which is mainly
[28:54] concentrating on Printing and uh uh hold
[28:57] Toole and hold to play
[29:00] technology where we are member and where
[29:02] we used to have a few examples here from
[29:05] it to show you a little bit more fancy
[29:07] things thank you Antonio thank you
[29:09] everybody for listening and if you have
[29:11] question please uh don't be shiny thank
[29:13] you thank you Thal so we have time for
[29:17] questions and um you can I encourage you
[29:19] to raise your hands so you can make your
[29:22] question directly to to
[29:26] toal I can perhaps start with one
[29:29] related the standardization that you
[29:30] mention for example in the case of the
[29:32] extreme angs that you can perform some
[29:35] urement but there is no standardization
[29:37] um how how do a company like Su microp
[29:39] disc Works in this duration do you have
[29:42] contact with committees to go um to a
[29:45] standardization of tools or you are just
[29:48] waiting for news from uh from the
[29:51] industry we are screening regularly the
[29:55] new ISO Norms which are coming coming
[29:57] out we we are part in some of the uh
[30:00] committees but it's kind of moving very
[30:02] slowly and you know usually gets uh it's
[30:06] not very fruitful discussions takes time
[30:08] and then uh it's kind of also a little
[30:10] bit closed community sometimes so um we
[30:14] ask our suppliers so this is actually
[30:16] what we try to do and push them to to go
[30:19] uh with the um this the bodies to to
[30:23] confirm the measurement
[30:25] principle yeah
[30:28] we also have a question from Rafael from
[30:30] iming optic yeah just just a question
[30:33] out of curiosity so thank you toal for
[30:35] the very kind very nice presentation uh
[30:39] you introduced um a clever way to um
[30:44] make samples compatible with the dynamic
[30:47] of instrument sometime which is by
[30:50] immersion and adapting the the
[30:52] refractive index um I was
[30:55] wondering uh for you as a manufacturer
[30:59] how is it uh is it an issue then the
[31:02] cleaning of the of the Optics and how
[31:04] you deal with all this preparation and
[31:07] yeah how this affect let's say your
[31:10] routine or whatever yeah yeah so there
[31:14] um in in some of the fabrication
[31:18] processes you even have an immersion
[31:20] situation can tell it like this if you
[31:22] stamp for instance and replicate you
[31:24] have immersion situations yeah uh in
[31:27] other cases what you usually do is you
[31:29] have a test sample which you throw away
[31:32] afterwards yeah so it's solved in a kind
[31:35] of good good way and you have losses
[31:38] yeah understood thank
[31:42] you we have another question from kri
[31:45] from J hello Jan from ATF kri very nice
[31:50] presentation really appreciate it just a
[31:53] question concerning continue with h with
[31:57] with the this last question is you you
[31:59] have referred about
[32:01] calibration ER how do
[32:04] you how do you proceed it in other words
[32:08] especially for dynamic situation do you
[32:10] do you have a really need as a as a
[32:14] complementary stat perhap for your
[32:16] customers concerning Mir Optical Mires
[32:19] for
[32:21] calibrations uh so what we usually do is
[32:24] we have developed a kind of set of
[32:27] calibration measurements which we do the
[32:29] instruments you which which is linked to
[32:32] step height flatness radius of curvature
[32:35] and and some other maybe objects which
[32:37] we have as calibration objects and of
[32:39] course we need for this calibrated
[32:42] objects which we got which we get from
[32:44] certificate certificated calibrated
[32:46] objects which we get from from bodies
[32:49] which we have to buy uh and some of them
[32:51] are available other are not available
[32:53] yeah okay we are specialist especially
[32:57] in one site inside the ATF group we are
[32:59] very interested in in exploring Synergy
[33:03] synergies with you concerning this type
[33:05] of of small or
[33:07] adaptive me for for your assistance so
[33:12] delighted if possible in a later stage
[33:14] after this meeting to have an informal
[33:17] at least talk with you thank you very
[33:19] much uh so feel free to contact me I'm
[33:22] not sure if Antonio if my contct details
[33:26] I can
[33:28] right good thank perfect yeah thank you
[33:31] final question from Andrea hello
[33:35] from I have a question about the the Lin
[33:39] machine that you presented So when you
[33:41] say there is no standard it means that
[33:43] you get measurement which are let's say
[33:46] relative to each other correct but there
[33:48] might be an offset with respect to
[33:50] reality or can you elaborate a bit more
[33:53] there um it's it's relatively easy you
[33:57] you want to you don't measure the
[33:59] surface shape in this case you measure
[34:01] the optical function somehow and you can
[34:03] use it for inspection to compare one and
[34:06] the other um but you cannot easily trace
[34:10] it really back to the surface shape
[34:12] because there is uncertainties on the
[34:15] liquid which you use uh which as an
[34:18] immersion liquid so if there's a small
[34:20] error there then you have a problem yeah
[34:23] um you also need to know somehow to
[34:26] refective indic of your base material
[34:29] which sometimes is very challenging to
[34:31] get in a very very high Precision so uh
[34:34] and this means that you have a good
[34:36] inspection tool but you will not have a
[34:38] surface measurement
[34:40] tool okay and that's clear thank
[34:44] you thank you Andrea and thank you again
[34:47] toal for your nice presentation and also
[34:49] for showing all the challenges to find
[34:51] in your industry regarding and
[34:53] characterization it was really really
[34:54] nice presentation thank you
[34:57] and uh let's go with our next speaker so
[35:00] we are moving to a different industry
[35:02] and we have a friend Berber it's R&D at
[35:04] alata steel and we are going to learn
[35:07] about the optical measurements also in
[35:08] in his company so Frank the floor is
[35:12] yours yeah thank you very much can you
[35:15] see already my screen everything good
[35:17] everything good yeah okay very good okay
[35:20] so inde need my name is Frank working as
[35:23] scientist at research and development at
[35:26] Tata Steel in a Maran in the Netherlands
[35:30] and yeah in my presentation I will first
[35:32] introduce you a little bit to the the
[35:34] world of
[35:35] Steel uh and just a few minutes and then
[35:38] we'll focus on the optical measurement
[35:40] Technologies especially focused to uh
[35:42] well a bit in the line with the previous
[35:44] speaker uh to the surfice
[35:48] topology um so first uh when we look at
[35:51] our sites yeah we have really an
[35:52] integrated steel production so it's from
[35:54] steel making uh iron making and steel
[35:57] making up to Steel coils as you see on
[36:00] the right uh which are then shipped to
[36:03] our customers typically in automotive
[36:05] and in packaging but also for the
[36:07] construction
[36:09] markets uh and to go from really and
[36:13] these primary products so the iron ore
[36:16] and also the coal uh well we have a
[36:18] quite say complex route uh with many
[36:22] factories you can see here uh to really
[36:26] produce these
[36:28] coils uh and in this rout also the yeah
[36:32] the semifinished products these are also
[36:36] often or or sometimes shipped to
[36:38] customers and also these need of course
[36:40] to be inspected either visually
[36:43] typically with cameras as well of course
[36:45] inspected on Geometry aspects Etc uh
[36:48] also surface def effects this is all
[36:50] automatically done um and the material
[36:54] which is shipped uh yeah that is
[36:56] typically
[36:57] galvanized product so coated with think
[36:59] or tin coated products uh and sometimes
[37:03] it's also painted especially for the
[37:05] construction
[37:06] markets uh and we do this uh on this big
[37:11] industrial s it's one of the biggest
[37:14] industrial sites in the Netherlands
[37:15] measuring 3 by 5
[37:18] kilometers uh you see we have an open
[37:20] Harbor connection to the Sea uh where
[37:23] the iron ore and also the coal are trans
[37:26] orted to and then produced further to
[37:29] these steel coils and of course from
[37:32] these products yeah many many different
[37:35] products are made whether these are
[37:37] windmills uh the cars the rails
[37:41] buildings ship building packaging
[37:44] materials Furniture etc
[37:48] etc okay now I will focus a bit more on
[37:51] the topic of today the surface
[37:54] topology um yeah and already of course
[37:56] course in this picture you see quite
[37:58] well how important this is typically
[38:01] when we have cars made of steel uh these
[38:05] car surfaces are inspected in these kind
[38:08] of streets where you have this uh
[38:10] lighting and you see immediately all
[38:12] kind of defects and so the surface
[38:15] topology is really important for the
[38:18] product appearance the look and the
[38:20] feel um but also for the sealability of
[38:24] especially packaging materials uh uh and
[38:27] also the yeah process steps which are
[38:29] done at the at the customer typically in
[38:31] pressing forming painting
[38:35] Etc uh for the manufacturing process
[38:37] itself yeah of course service topology
[38:40] is also very important because uh it has
[38:42] an an impact on the process control and
[38:45] product quality and also uh the surface
[38:49] finish it really is highly dependent on
[38:51] our yeah processing over the entire
[38:55] Road um well if we look then how this
[38:58] surface texture is formed well it's
[39:00] typically in C rolling operations so we
[39:03] have the the work roll uh and these work
[39:07] rolls um these in In Cold Rolling had
[39:10] the well fresh strip enters and then due
[39:13] to the roughness transfer we get a strip
[39:16] with a certain finish coming out and
[39:19] here you see some example of topology in
[39:21] and topology
[39:23] out uh and indeed as it said there this
[39:27] yeah the process itself has a quite
[39:29] large impact on this texture of the
[39:33] strip um and here you see an example of
[39:36] the effect of where of the work R and so
[39:40] this texture it evolves and the yeah
[39:43] roll surface topography on the left you
[39:46] see an image how it looks for a fresh
[39:49] textured work roll and on the right for
[39:52] an worn work roll after about 4 kilm of
[39:56] called rolling uh and you see quite well
[40:00] that especially in this yeah Optical
[40:02] reproduction uh that the say
[40:05] mountains of the surface topology are
[40:09] chipped off after rolling so that's the
[40:11] sort of Aging that you see uh another
[40:15] aspect which is important is the work
[40:16] roll Force so if you have a very light
[40:19] work roll Force then well you see hardly
[40:22] any imprints on the material uh with a
[40:25] normal R Force then you see directly in
[40:28] Yeah a different surface topology with a
[40:31] higher roughness in this
[40:33] case so to um yeah to show what is our
[40:38] ambition what do we ideally want is to
[40:41] inspect the surface topology before and
[40:43] after the rolling
[40:45] process and so this we would like to do
[40:47] inline and real time and use this data
[40:50] as input to a control Loop but yeah the
[40:53] challenges are quite important that we
[40:55] have and it's and we have to measure on
[40:58] moving product so line speed is about 3
[41:02] to 7 m/ second and so normally you would
[41:05] have motion blur for one micron lateral
[41:09] resolution you need to have a sort of
[41:11] snapshot techniques that really acquires
[41:14] all your data in just 200
[41:18] NS uh that's not all also there are
[41:21] quite some vibrations in the production
[41:22] line so it's the object itself the steel
[41:25] strip as we call it it vibrates and also
[41:28] the entire installation basically
[41:30] vibrates so this is making our life
[41:33] difficult but well yeah this could be
[41:35] your challenge as we
[41:37] say uh and then if we can do this then
[41:41] we can if we can measure the surface
[41:43] topology in line then we can also yeah
[41:45] measure the topology of strip in and
[41:48] strip out and then do the roughness
[41:51] transfer analysis in real
[41:54] time uh if we compare it how it's
[41:57] currently done and especially what's the
[41:59] standard reference for the measurement
[42:01] so what is used for product release uh
[42:05] that is still very conventional
[42:07] oldfashioned you can say the mechanical
[42:09] stylist and that is simply the norm and
[42:12] yeah of course we are trying to push
[42:14] also the Norms but as already was
[42:17] discussed in the previous discussion
[42:19] these discussions go very slowly uh so
[42:22] in the optical stylist well you will
[42:24] know there's a contacting needle uh with
[42:26] a very sharp tip of a radius about two
[42:29] Micron and we record the displacement of
[42:32] the needle but also there you can have
[42:35] of course uh issues in this standard
[42:37] they can Beware of the radi aware of the
[42:40] tip uh there can be also yeah the
[42:42] contact force can have an influence the
[42:44] speed that which you measure has an
[42:46] influence and you need also to measure
[42:49] multiple tracks to have a good
[42:51] statistical averaging of these profiles
[42:54] but okay uhuh typically for steel you
[42:56] would get such a type of uh C rolling
[43:00] profile uh in order with a height
[43:03] difference in order of one
[43:05] micron um and uh yeah if we look then
[43:10] typically at what are the say the
[43:12] quantities or parameters which are
[43:14] associated to such an measurement is
[43:17] yeah the the height resolutions to be
[43:19] should be typically be be less than 0.1
[43:23] Micron typically 50 Nano MERS uh lateral
[43:28] resolution in order of one micron and
[43:31] from these profiles height profiles we
[43:33] did use the roughness values typically
[43:36] ra and also Wess uh parameters which are
[43:39] bit the more longer type of uh shape uh
[43:43] variations or topology
[43:45] variations uh and the stylus uh speed is
[43:48] in order of the 1 millimeter per second
[43:51] but okay uh we are also of course
[43:54] looking at other techniques especially
[43:56] in the lab so what we have or have
[43:59] tested in laboratory to measure surface
[44:02] topology uh this either Optical stylus
[44:05] and where you do a scan of course of
[44:07] such a point measurement either working
[44:09] on based on laser triangulation or laser
[44:11] confocal or chromatic confocal
[44:14] techniques uh or preferred you can say
[44:17] Optical imaging technology like confocal
[44:20] Imaging and on the right you see an
[44:23] example of where an yeah an in imager
[44:26] confocal imager is used on a work rle so
[44:29] you see see here such a role uh to
[44:32] measure the the topology of the RO
[44:36] itself uh and some other techniques have
[44:39] been measured before well even
[44:40] holography also mentioned already by the
[44:42] previous
[44:44] speaker um but okay these are lab
[44:46] techniques and we would like to measure
[44:48] really in line so also here for inline
[44:50] measurements different techniques have
[44:52] been investigated and also tried uh one
[44:56] of them is angular resolved scattering
[44:58] and where we project a laser beam very
[45:01] fine beam onto the surface and we look
[45:04] at the angular resolved reflection which
[45:07] tells us something about the
[45:09] distribution of the slopes you can say
[45:11] and this is tested uh well here you see
[45:14] an example installed in line at a
[45:16] distance of about 10 cm from the strip
[45:20] and taking a measurement every 250
[45:24] NS uh uh a second type of technique is
[45:28] to use yeah laser triangulation but
[45:30] really with a nano second pulse laser
[45:32] because of course you have to freeze the
[45:35] motion of the
[45:36] strip um and with such a device indeed
[45:40] you you measure tracks uh and you yeah
[45:43] you can of course only record when in
[45:45] focus and you can deduce roughness and
[45:47] wavess parameters and these techniques
[45:50] these have been
[45:52] evaluated uh in the manufacturing
[45:54] conditions and yeah what we have
[45:56] observed is that okay they are useful to
[46:00] indeed obtain Trends uh in in the
[46:03] roughness during
[46:05] production yeah however uh for really
[46:07] for product release these are not
[46:09] reliable enough
[46:12] unfortunately uh well also more
[46:14] techniques have been evaluated in the
[46:16] lab like speckle Imaging like white
[46:18] light interferometry like wave from
[46:21] sensing um but yeah the general
[46:25] difficulty that always have if you also
[46:27] look from it from an even yeah when you
[46:31] model and the interaction with light
[46:33] with the surfaces is really that the
[46:35] roughness on one hand and the optical
[46:37] wavelength that we are using they are
[46:40] yeah of the same order of magnitude so
[46:42] there is a mix of diffuse of specular
[46:44] and of multiple scattering and this is
[46:46] really what say troubles our situation
[46:50] and to come up with an solution that
[46:52] works uh so finally yeah I can end with
[46:55] with this slide uh well could this be
[46:58] your challenge to inspect uh in real
[47:01] time uh the surface to to poy with some
[47:04] kind of snapshot technology and with
[47:06] that I would like to finish this
[47:08] presentation thank you thank you very
[47:11] much Frank for showing us all your
[47:13] attempts to bring the optical
[47:14] measurements to your industry and also
[47:17] your challenge for the attendees today
[47:20] any questions for
[47:24] Frank
[47:27] I will perhaps start asking about the
[47:29] how how clean is the surface of the
[47:32] metal uh can it be an issue when
[47:36] measuring with yeah it can absolutely be
[47:38] an issue uh of course it can be D on the
[47:41] surface it can also be sometimes or
[47:43] there will always be some little uh
[47:46] layers of oxides as well on the on the
[47:49] surface um and also these
[47:53] oxides uh yeah they have of course also
[47:56] interaction as well as there are so yeah
[47:58] often also foil film layers because
[48:02] that's part of the during this rolling
[48:04] process emulsions are used and yeah
[48:08] these emulsions of course uh uh in
[48:11] situations where they are still on the
[48:13] on the material uh yeah they hamper also
[48:16] your measurement and therefore I only
[48:19] mention these two aspects of well first
[48:21] of all vibrations and secondly of the
[48:23] Dynamics of the whole situation but yeah
[48:26] there's also the entire industrial
[48:29] situation where it has to work uh with
[48:31] dust with even indeed these kind of
[48:33] emulsions or splashes of of water if you
[48:36] like uh so you have to condition your
[48:39] measurement very
[48:42] well good we have time for another
[48:46] question for
[48:51] Frank well if we we don't have any uh I
[48:55] hope that you will perhaps find in the
[48:57] next speaker
[48:59] presentation perhaps you want to discuss
[49:01] with them more in detail about the
[49:03] technology and perhaps it makes sense to
[49:05] apply in your industry as well so thank
[49:08] you very much Frank it was really
[49:10] interesting okay thank you very much
[49:13] byebye and uh let's go with our next
[49:15] speaker we have Rafael porar his
[49:17] scientific coordinator of at Imaging
[49:19] optic and uh he's going to show us the
[49:22] uh industry Optical testing boost by
[49:24] lift wave from sensing so Rafael the
[49:27] floor is yours you can share your
[49:31] screen can you see my screen yeah good
[49:34] everything okay thank you Antonio for
[49:36] the for the introduction and the added
[49:38] pressure um no worries
[49:42] so my name is Rafa POA and today I'm I'm
[49:46] going to discuss an innovation we we
[49:48] brought to the to the market we call
[49:51] lift we from sensing uh and that um that
[49:55] in ultra high resolution in in web front
[49:59] sensing uh and I will also review a few
[50:03] uh use cases we we had with customer in
[50:06] in
[50:07] Industry okay
[50:11] so so imagin optic is a Metrology
[50:15] company um we are developing wont
[50:19] sensors and the and Optical Metrology
[50:22] instrument so this is a non contct
[50:26] um quality testing and uh Optical
[50:29] quality uh in transmission or Surface
[50:33] shape when we when we measure in uh
[50:36] reflection so we we can say we have
[50:39] proboly deployed more than 2,000 sensors
[50:42] all over the world both in industry and
[50:46] Academia uh both in R&D and uh
[50:51] manufacturing so uh one core value of
[50:56] Imagine optic is constant Innovation and
[50:58] this is what I came today to to present
[51:03] um and I'm going to start with a kind
[51:06] reminder of um of um what is a
[51:10] shakman webfront sensor so it consists
[51:13] basically in an array of micr lenses you
[51:16] put in front of a sensor each of these
[51:18] micr lenses focuses centroids on top of
[51:22] the of the camera and what we do is we
[51:26] analyze the lateral displacement of
[51:29] those centroids in order to reconstruct
[51:32] what are the slopes uh that
[51:36] um of of the of the phase uh and which
[51:41] which are the angle of of propagation of
[51:45] of phrase and um from that we we we
[51:48] integrate reconstruct surfaces and and W
[51:51] Front what's important here is that
[51:54] basically sh man w front sentor
[51:55] reconstruct two information per sampling
[51:58] points per micro lens which is the slope
[52:01] in X and the slope in y and from that
[52:04] what you can reconstruct is kind
[52:07] of
[52:09] averaged bit of of surface which are
[52:13] plain surfaces question is from from the
[52:17] row signal we can we get from these kind
[52:19] of sensors is there something more that
[52:22] we that we can do no more than just the
[52:25] analysis of the lateral position of
[52:28] those centroids and if you have a look
[52:30] here to this signal uh I'm I'm sharing
[52:34] with you it seems that it is the case no
[52:36] it seems that all the centroids focused
[52:39] by the micr lenes are not exactly the
[52:41] same this is this one is more round this
[52:44] one or this one for example it seems to
[52:47] be more elongated so is there something
[52:50] more that we we can do and that's
[52:51] exactly the approach of the of the lift
[52:54] innovation
[52:55] that has been initially proposed by our
[52:58] colleague at H Research Center and it
[53:01] consists in on top of analyzing the
[53:05] position of the centroid in analyzing uh
[53:09] the intensity distribution within the
[53:11] the focal spot and because we know that
[53:15] somehow uh when you have aberration uh
[53:18] and Optical abberation they have an
[53:20] impact no on how your light is focusing
[53:23] so it's it's really exploiting this
[53:25] this specificity and it allows us to
[53:29] reconstruct per micr lens a combination
[53:32] of up to nine uh different type of
[53:37] aberration uh so it allows us that you
[53:40] have to compare that with the previous
[53:42] technique shakman technique where you
[53:44] only get two modes per micr lenses now
[53:47] we have nine modes per micr lenses and
[53:49] this is um added uh detail at the scale
[53:53] of the micr lenes and this uh um uh
[53:57] translate to increase resolution in the
[54:02] in the measurement so now the web front
[54:04] sensors on top of being versatile easy
[54:08] and robust to to implement they can
[54:10] reach a new level of uh of resolution
[54:13] that we have been uh analyzing uh in
[54:17] order to check the performance of the of
[54:19] the technology so we we sent several
[54:22] complex pattern of uh of phase and
[54:24] increase resolution on our sensors in
[54:27] order to see what is the Improvement in
[54:30] the transfer function of the instrument
[54:33] and um and if we summarize what we can
[54:38] see is an increase in the lateral
[54:40] resolution of of the reference sensor by
[54:43] a factor of four times in in in one axis
[54:47] four time in the other one so overall
[54:50] it's a 16 time uh
[54:53] Improvement if you check check this uh
[54:56] visually um that's an example on the
[54:59] left you have a a map of a wavefront
[55:03] corresponding in this case to a s
[55:04] measurement so we are at 1.5 Micron
[55:08] wavelength and at that wavelength
[55:11] sensors usually has less pixels and then
[55:15] the sensors also have less microl lenes
[55:18] so here you have a measurement of 28 by
[55:20] 28 sampling points and when you apply
[55:23] the new lift Rec constriction you can
[55:26] see how big uh the the Improvement uh is
[55:31] no by simply comparing visually uh the
[55:35] fness in the in the map so um that said
[55:41] um we have um implemented this
[55:45] technology in all our lineup from from
[55:48] our wfront sensors up to our um
[55:53] instrument so we have different families
[55:55] of instrument versatile one where you
[55:58] can combine different modules in order
[56:00] to do your metodology very in a
[56:02] versatile way in the in the lab we also
[56:05] have a a robust automated instrument
[56:08] more sort for manufacturing and um and
[56:11] what I want to show you is different
[56:13] examples of um of this ultra high
[56:17] resolution the first one is in the field
[56:20] of microoptics production so here this
[56:22] is quality testing of um of produced
[56:25] microl lenes that we do at multiple
[56:28] wavelength so here you have uh two
[56:31] pictures of the test bench uh testing at
[56:34] in the red and in the in the green wav
[56:36] lences and this is the kind of wavefront
[56:40] we measure in transmission so these are
[56:43] micro lenses uh the pitch of them is T
[56:46] of Micron so you have to see that here
[56:49] we have more or less 700 by 500 points
[56:53] which means that here we still have at
[56:55] the scale of
[56:56] the micro structures more than one
[57:01] something Point by by Micron we can also
[57:04] isolate one micr lens and and check for
[57:07] its
[57:08] parameters okay another example still in
[57:11] the microoptics production but now for
[57:14] produced by um injection
[57:17] molding so what we do with our customer
[57:21] is testing the quality of the mold we
[57:24] are not testing
[57:25] the Optics direct directly but now we
[57:28] are testing the mold that is going to be
[57:30] used in order to to produce these
[57:34] this Optics with micro stors and this is
[57:38] a very complex mold in in one part of
[57:40] the mold you can see it here on the top
[57:44] map W Front map we are testing in
[57:47] reflection um the the polishing quality
[57:51] and you can see here those circles uh
[57:54] that are made by the polishing tool and
[57:57] we do iteration and and we help our
[57:59] customer to to determine when he has he
[58:02] can stop the polishing of his mold and
[58:07] at the same time in another part of the
[58:08] mold we are also testing different micr
[58:12] structures so here we are at the
[58:15] submillimeter um size more or less
[58:20] okay um we have also implemented the
[58:23] lift W sens
[58:25] in
[58:26] automotive uh in order to test
[58:29] protective Windows you have in front of
[58:32] different um sensor in automotive like
[58:35] cameras and and Radars so in this first
[58:37] case we are testing the same part both
[58:40] in uh reflection and in Transmission in
[58:43] reflection we use it in order to again
[58:46] assess the quality of the of the of the
[58:49] mold and and and to relate the quality
[58:53] of the mold and the shape of the mold to
[58:55] the actual shape of the part that has
[58:59] been injected because sometimes you have
[59:01] some some variation we also measuring in
[59:04] transmission so in this case uh with um
[59:07] MTF MTF measurement uh telling us a
[59:11] little bit what what are what is the
[59:14] optical quality in transmission of the
[59:16] of the of the protective window here
[59:19] another one still in the automotive uh
[59:22] here our customers has a very complex
[59:24] Optical part
[59:26] combining plastic together with metal
[59:30] and electrodes those electrodes are kind
[59:33] of introducing some deformation in the
[59:35] in the plastic part so we are assessing
[59:38] what is the impact of those electrodes
[59:40] on the on the optical quality by
[59:42] measuring the uh the transmitted wfront
[59:45] so so again we we can uh make some
[59:48] profiles and or reconstruct in 3D um the
[59:52] shape of the of the part
[59:56] and uh last example in Industry so this
[59:59] is glass
[01:00:00] manufacturing uh we are uh helping our
[01:00:04] customer to uh test a large part large
[01:00:09] large for us meaning uh six 6 in um they
[01:00:14] are very thin plain parallel substrate
[01:00:19] so so we are measuring them in
[01:00:22] transmission and again uh when we filter
[01:00:25] all the low uh spal uh orders we can see
[01:00:30] appearing what are the detail of the of
[01:00:33] the polishing tool um on the on the
[01:00:37] optical
[01:00:38] quality okay as um as a bonus uh here I
[01:00:43] can say that and that something we share
[01:00:46] all the all the Technologies from we
[01:00:48] front sensing we can work at almost any
[01:00:50] wavelength so in this case uh this is an
[01:00:53] alternative to interferometry like physo
[01:00:57] interferometry
[01:00:58] um because this part has to be measured
[01:01:02] at 800 nanometers for his spectral um
[01:01:07] coating uh and and could not be done
[01:01:10] with a with an
[01:01:12] interform okay so as a conclusion um we
[01:01:17] we help developed and commercialize this
[01:01:20] this new innovation we called lift that
[01:01:23] brings now
[01:01:25] we sensors with up to 700 by 500 face
[01:01:28] points per per pupile uh whatever the
[01:01:31] pupile is so so you can really
[01:01:36] have great um resolution Tech is
[01:01:40] acromatic so it works on the whole
[01:01:42] visible range but it also work on on the
[01:01:45] S and we can measure from Optics from
[01:01:49] tens of microns up to six Ines uh both
[01:01:52] in transmission and uh and in reflection
[01:01:55] so that's all for me I thank you for
[01:01:58] your attention and uh I'm up for any
[01:02:02] question I
[01:02:03] hope thank you Rafael so we have one
[01:02:06] already from jge from
[01:02:08] K
[01:02:11] okay you're muted right
[01:02:15] now ER thank you Rafael for the
[01:02:18] interesting presentation just a question
[01:02:21] concerning micr lences which is the
[01:02:24] infuence up to now with your technology
[01:02:28] obviously which is the influence of any
[01:02:31] cross talk or parite Reflections on your
[01:02:36] microsystem of
[01:02:38] lenses is it a very relevant effect you
[01:02:41] should optimize
[01:02:43] perhaps well you are talking about the
[01:02:46] cross talk between C trades yeah no is
[01:02:51] the
[01:02:52] Performing all making some bra out of
[01:02:57] the geometrical the geometric
[01:03:00] approximation the parial one out of to
[01:03:02] the other channel we could say like this
[01:03:05] yeah um so honestly there's
[01:03:10] no there's no cross talk affecting uh
[01:03:14] the Reconstruction when you when you
[01:03:16] reach uh this point where where you you
[01:03:19] might have a cross talk in between the
[01:03:22] centrate produced by
[01:03:24] um micro enes you you already reach uh
[01:03:29] the dynamic of the sensor that's that's
[01:03:32] limit of of the
[01:03:34] measurement uh but uh you have to be in
[01:03:38] very extreme conditions uh because
[01:03:41] without this we can still measure
[01:03:44] dynamics of more than one micron in tilt
[01:03:48] and um and five 500 nanometers in in
[01:03:53] sphere so
[01:03:55] uh it's it's not something happening uh
[01:03:59] that that fast it it seems that they can
[01:04:02] they can they can merage very quickly
[01:04:04] but if you if you do the math it's not
[01:04:07] that uh that Sur okay thank you thank
[01:04:11] you very much we have another question
[01:04:13] from Ian Warhead
[01:04:17] selan hi uh thanks for the presentation
[01:04:20] um a quick question on uh um sha
[01:04:23] harman's the offering can can can they
[01:04:26] work down to ultraviolet wavelengths 355
[01:04:30] 266
[01:04:31] 248 yeah so so the limit is a the limit
[01:04:36] is a sensor um we we use we we usually
[01:04:41] use uh C sensors so they are sensitive
[01:04:45] uh down to 400
[01:04:47] nanometers um in reality a little bit
[01:04:52] less than than that but more more or
[01:04:54] less this is this is the limit for the
[01:04:56] visible range and then we have to to
[01:04:59] jump uh we have other um sensors that
[01:05:03] are specifically designed for the
[01:05:06] UV okay very good
[01:05:08] thanks thank you
[01:05:10] Yan and uh I don't see any other
[01:05:13] questions so thank you very much Rafael
[01:05:15] for your nice presentation and um for
[01:05:18] showing us some examples also for
[01:05:20] different industry that you are covering
[01:05:22] nowadays thank you very much
[01:05:24] and we can go with our next speaker is
[01:05:27] Kevin contas he's R&D manager at V
[01:05:30] Optics uh so Kevin you can share your
[01:05:33] screen yeah okay hello everyone my my
[01:05:38] name is Kevin contas I am &d manager at
[01:05:42] the the Optics so so thank you very much
[01:05:46] for your invitation for to speak today
[01:05:49] so first I
[01:05:52] I uh
[01:05:54] a little introduction about a so the
[01:05:58] Optics founded in two
[01:06:03] 2008 with the heritage of the activities
[01:06:07] A&D the French lab so we have actually
[01:06:11] seven patents three news in progress so
[01:06:14] we are a small company with 14 people
[01:06:19] five members in the area and the team so
[01:06:23] we a strong competence Optics and
[01:06:26] imagine processing so we are raed in
[01:06:30] stud near to theb in France so we have a
[01:06:35] strong expertise on phase shifting
[01:06:38] deflectometry um our ambition is to to
[01:06:43] be your partner Ser def effect in your
[01:06:45] Control process
[01:06:47] so so a little introduction about the
[01:06:51] phase shifting deflectometry is not the
[01:06:53] idea is to present you a
[01:06:57] basic a
[01:06:59] basic
[01:07:04] sorry sorry there is some some
[01:07:07] background noise I don't know oh okay
[01:07:09] it's not
[01:07:10] me sorry
[01:07:14] G ah okay so I continue so so the
[01:07:18] deflectometry measurement methods us the
[01:07:21] low of the reflection refraction
[01:07:25] and then knowledge the between the
[01:07:28] camera
[01:07:30] Andor so a liquid crystal display
[01:07:32] monitor to draw conclusions about the
[01:07:35] state of the Surface by means observing
[01:07:38] the deformation of the mirror IM for
[01:07:42] automatic visual
[01:07:43] inspection it's necessary to to to know
[01:07:49] uh all system parameters for example the
[01:07:52] size and the position of the L T monitor
[01:07:54] relative to this camera sensor and
[01:07:57] obviously thex camera
[01:08:00] parameters so uh the our device are
[01:08:06] based on different configuration and
[01:08:08] transmission and reflection you can see
[01:08:10] in this slide for
[01:08:13] example some pictures
[01:08:16] about the Distortion the evasion of the
[01:08:19] the light Ray when is transmission and
[01:08:23] reflection
[01:08:24] so with this Optical technique is
[01:08:29] possible just to to St to to to St the
[01:08:32] amplitude and information maps to obtain
[01:08:36] gradient and
[01:08:38] gradient direction to evaluate the slope
[01:08:43] altitude uh
[01:08:46] we we have a a Rob algorithm for the pH
[01:08:49] reconstruction so the basic of this um
[01:08:54] our device is based on this this
[01:08:56] technique so so for example some
[01:08:59] industrial application for example for
[01:09:03] the for the detection of this defs of
[01:09:06] differ origin of the during the
[01:09:09] manufacturing process for example scrash
[01:09:11] di inclusion bubbles the effect orated
[01:09:16] by Diamond tning for example so it's
[01:09:19] possible with the deflectometry to
[01:09:21] evaluate it and to detect it this
[01:09:24] this effect so it's possible also to
[01:09:27] evaluate it or to measure the flatness
[01:09:30] obviously the accuracy depends of the FI
[01:09:33] of view so the deflectometry is very is
[01:09:36] very well adapted to the industrial
[01:09:39] applications
[01:09:41] uh the the main advantage is to have a
[01:09:46] large field of view but your price to
[01:09:50] pay is to obtain a data dep accuracy so
[01:09:54] for example in this is I show you the
[01:09:56] detection of the spash on a wafer the
[01:09:59] sizes around some microns in a large
[01:10:03] field of view and to right you can see
[01:10:06] the difference measurement about the
[01:10:09] fless for two uh for both sides of the
[01:10:13] transparent
[01:10:15] P so uh in reflection we have two
[01:10:22] instruments two divice quality sensor
[01:10:25] 100 and quality sensor 200 so with
[01:10:28] different feel of view we can to it's
[01:10:30] possible to control of SP surface
[01:10:33] painting texture and
[01:10:35] surface uh properties plastic and
[01:10:39] composite materials injection molds also
[01:10:42] paper carton mirror glasses metals and
[01:10:45] Ceramics so we can detect we can we can
[01:10:48] detect it orange peel for example
[01:10:52] or in the pink
[01:10:55] so we have other other versions about
[01:10:59] this um this configuration we have the
[01:11:02] quality sense 400 so for the inspection
[01:11:07] of the effect of the wind shell and qual
[01:11:12] lens 200 for the watch glasses for
[01:11:16] example and Optical lenses filters for
[01:11:19] example in the picture we to see uh
[01:11:24] the W glasses is very critical in the
[01:11:26] industry because it's very important for
[01:11:28] the luxury
[01:11:30] industry with
[01:11:32] this device or based on the
[01:11:35] deflectometry is possible to insect all
[01:11:38] th of the the material not just in the
[01:11:41] surface you know really in the def in
[01:11:46] the
[01:11:47] material so for theic we have a qu lens
[01:11:52] 100 uh
[01:11:55] very a system adapted for the stalic
[01:11:59] industry so properties we can very
[01:12:02] excellent rep repe and
[01:12:05] reputability we have not alignment the
[01:12:09] lenses it's a fast measurement user
[01:12:13] friendly
[01:12:15] interface about
[01:12:17] the this
[01:12:19] software we have the flexibility for
[01:12:24] new uh
[01:12:26] functionalities we can customize it
[01:12:32] develop so for finally for the for the
[01:12:36] application and the industry we have
[01:12:38] other device on the qual lens 150 for
[01:12:42] the inspection for the automatic Optical
[01:12:45] inspection for the inocular lenses this
[01:12:49] in this case is possible to put
[01:12:52] different um
[01:12:55] larg number of the of the device of this
[01:12:59] more lenses is possible to to to detect
[01:13:02] the def located and the detection of the
[01:13:06] differ def so we have
[01:13:10] also qu lens 100 for the lenses and I so
[01:13:16] finally
[01:13:17] we 300 for the optical inspection of the
[01:13:22] E glasses where you can
[01:13:25] put between the LCD monitor
[01:13:31] camera so
[01:13:34] finally present of the Optics as for the
[01:13:39] def protection in transparent recting
[01:13:41] materials so we are very optimist about
[01:13:44] the pure of the optical deflectometry
[01:13:46] for preform and micro so thank you for
[01:13:50] your
[01:13:52] attention
[01:13:58] thank you Kevin thank you very much for
[01:13:59] your
[01:14:00] presentation and we have time for
[01:14:02] questions for
[01:14:08] Kevin I have one regarding the the
[01:14:11] famous EP epic question because I saw
[01:14:14] what's the Optics offering uh regarding
[01:14:16] the development of your products do you
[01:14:19] have any any need or any challenge for
[01:14:21] the audience regarding one of the
[01:14:23] components you can use can be
[01:14:24] illumination can be uh Optical component
[01:14:27] that you use um do you have any any need
[01:14:31] or or challenge for the
[01:14:32] audience yes yes yes exactly so we are
[01:14:36] open to to to to work part partnership
[01:14:40] for example for different manufacturing
[01:14:43] Optics so but now now way is is our
[01:14:48] business class is to very oriented to
[01:14:50] theal Industry so
[01:14:54] we are open to to to
[01:14:58] discuss
[01:15:01] good I don't see any other question in
[01:15:04] the audience so Kevin thank you very
[01:15:07] much for presenting your your technology
[01:15:08] and your device and um we can move to
[01:15:12] our next speaker it's a pleasure to
[01:15:14] introduce Andres defuentes I know him
[01:15:17] for a few years already uh CEO at ASC
[01:15:20] Optics and he's going to present his
[01:15:23] high dimension High Precision
[01:15:24] dimensional inspection for industry so
[01:15:27] Andre the floor is
[01:15:29] yours thank you Antonio it's pleasure to
[01:15:33] to talk to you again as always and to be
[01:15:35] here at an epic meeting I appreciate the
[01:15:37] invitation and and being able to um to
[01:15:41] share some uh results from r&t activity
[01:15:44] that we carryed at ASC
[01:15:46] Optics um related to to Metrology that
[01:15:50] could have applications in in industry
[01:15:54] um so
[01:15:58] uh okay so I kind of focus on two areas
[01:16:03] that I I want to discuss that are here
[01:16:04] on the left which is something that
[01:16:06] we've been calling High Precision large
[01:16:08] area
[01:16:09] Metrology and then a second area that we
[01:16:13] that also came out from our R&D projects
[01:16:15] in at ASC is inline micro inspection so
[01:16:19] I I spelled that wrong sorry about that
[01:16:23] um and then just a very quick uh review
[01:16:26] of of AC Optics we're an optical uh
[01:16:29] design and and assembly an integration
[01:16:32] company of optical
[01:16:33] systems uh we have we're small but a
[01:16:36] quite Innovative team in the field of
[01:16:38] Optics and
[01:16:39] photonics um working in in different
[01:16:42] fields one of them being some
[01:16:44] applications in Metrology as you as
[01:16:46] you'll see here we don't uh you know
[01:16:49] kind of a difference from some of the
[01:16:50] other great speakers that have been uh
[01:16:53] before me is we we don't have a a
[01:16:55] off-the-shelf or an instrument per se
[01:16:59] but rather some technologies and
[01:17:01] experience that have applications that
[01:17:03] we can further develop or customize for
[01:17:06] for our our potential customers or
[01:17:09] Partners okay so I'll jump right into it
[01:17:13] so where do this High Precision large
[01:17:15] aial Metrology comes from so at ASC we
[01:17:19] worked for a long time for several years
[01:17:22] in um and developing and and testing a
[01:17:26] basically what could be considered a lar
[01:17:28] system for the inspection inside
[01:17:31] the What's called the TAC the the
[01:17:35] chamber inside of a fusion reactor
[01:17:37] that's being built at the at what the
[01:17:39] experiment is called eer in the south of
[01:17:43] France so it requires an optical
[01:17:47] inspection to help determine the
[01:17:51] erosion that is um due to operating the
[01:17:55] the very high temperatures plasma and
[01:17:57] the and the very high magnetic fields
[01:17:59] inside of this chamber so to do that
[01:18:03] inspection one technique that was that
[01:18:05] came up and that's what we explored was
[01:18:08] was based on optical Metrology so we use
[01:18:12] the scanning system to look through the
[01:18:14] to to to scan areas within a within a
[01:18:18] chamber um to to be able to make a
[01:18:21] reference model of of the 3D reference
[01:18:24] model of the inside of the
[01:18:26] chamber uh difference from what you can
[01:18:28] consider let's say the typical lighter
[01:18:30] for other types of application that's
[01:18:32] why this is definitely more Industrial
[01:18:34] in the use case is it has Micron level
[01:18:36] Precision uh at about 10 meters or or
[01:18:40] less uh which is definitely not not
[01:18:43] typical for a light art system and and
[01:18:47] it and we did have to make it compatible
[01:18:49] with harsh environments which maybe in
[01:18:51] all industri is not required but it does
[01:18:53] give us let's say a unique experience in
[01:18:56] working in very harsh environments
[01:18:59] inside of um in The Ether Tac which is
[01:19:04] high vacuum High radiation High magnetic
[01:19:08] fields um these kind of things that
[01:19:10] complicate a lot in the measurements
[01:19:12] we've done a lot of work in in this area
[01:19:15] U but I kind of focus it here on just
[01:19:17] kind of the results of the Metrology
[01:19:19] itself so we could we ended up being
[01:19:21] able to make a 3D mapping of the
[01:19:23] elements inside The Ether Tac it could
[01:19:26] detect any damage or erosion of the
[01:19:28] components of tiles that that uh the
[01:19:30] chamber uses we did the complete Optical
[01:19:33] system design and integration and the
[01:19:36] whole test campaign to validate the to
[01:19:38] validate and also find the limits of the
[01:19:41] of the uh resulting
[01:19:44] technology and we uh finally we're able
[01:19:46] to package it into something that could
[01:19:49] be put into the what the The Ether
[01:19:51] cartridges what they call their
[01:19:52] cartriges and we did as I mentioned
[01:19:54] before the implementation and
[01:19:58] testing so um you can see my screen so
[01:20:02] basically we ended up finding a combined
[01:20:05] solution that solves both the metrol
[01:20:08] what we called the Metrology and the
[01:20:09] viewing channel so we have uh Metrology
[01:20:13] so measurement but we also have a view
[01:20:15] so a 3D reconstruction and that's what
[01:20:18] I'm showing here is the 3D
[01:20:19] reconstructions which are a lot more fun
[01:20:22] to see but behind this there's there's a
[01:20:24] system of measurements so the Metrology
[01:20:27] channel was based on a a frequency
[01:20:29] modulated lar and this gave us the
[01:20:32] absolute distance
[01:20:33] measurements and then this was combined
[01:20:36] with an amplitude modulated lar which
[01:20:38] gave us the surface viewing so um not
[01:20:43] each of these techniques by themselves
[01:20:45] wasn't enough but we devis a ways to
[01:20:47] combine them to give the functionality
[01:20:50] that in the end was required for the for
[01:20:52] the fin
[01:20:55] application uh so here's just some views
[01:20:57] of the the raster scans for the 3D
[01:21:00] reconstructions we can measure at 10
[01:21:02] meters um you can see here the top
[01:21:05] images
[01:21:07] tiles the this image is just a big stool
[01:21:10] but we wanted to show that you could at
[01:21:12] 10 meters you could actually capture
[01:21:15] this uh this with Precision a larger
[01:21:18] area but also a smaller area and that's
[01:21:20] why we scan this coin here at uh this
[01:21:22] Euro coin also at 10 meters it's a this
[01:21:26] particular solution was very slow but
[01:21:27] that was just due to the to the
[01:21:29] availability of products or components
[01:21:33] that that were compatible with that
[01:21:34] particular harsh environment so that's
[01:21:36] not a not a particular limitation of the
[01:21:38] system itself and then all of this that
[01:21:41] you can see here was captured and
[01:21:43] processed and and displayed by our own
[01:21:47] custom software development
[01:21:50] team and then the the fun part is is
[01:21:54] that
[01:21:54] this we we find that this could have
[01:21:57] applications in industry and that's why
[01:21:58] we're here today and I thought this was
[01:22:00] an excellent meeting to participate in
[01:22:02] uh where have we found some interest
[01:22:04] though we've not um deployed this
[01:22:07] product like I mentioned anything yet
[01:22:10] but we did find interest in say where
[01:22:13] where High Precision is required in
[01:22:14] large areas like for example in in
[01:22:17] construction of of Auto of aircraft or
[01:22:21] maybe highend automobiles but particular
[01:22:23] aircraft where as you could see here
[01:22:25] maybe rivet inspection is an important
[01:22:28] uh key area for the aerodynamics of the
[01:22:31] final assembly of the system maybe some
[01:22:33] of the high Precision Parts involved in
[01:22:35] making some of the turbines in in um in
[01:22:40] Aerospace the previous speaker Frank
[01:22:42] also mentioned some some areas where
[01:22:45] where I will talk about inline
[01:22:47] inspection but where also large area and
[01:22:49] high Precision were required which we
[01:22:50] hadn't even noticed so I'm already glad
[01:22:52] I came to this meting and and finally
[01:22:54] just in general like finished product
[01:22:56] quality inspections when things are
[01:22:58] pretty big so in that sense um you know
[01:23:01] to epic's question what what do we we
[01:23:04] need like a general need I'll have a
[01:23:07] slide for that more specific later but a
[01:23:08] general need is if there's ideas or end
[01:23:11] users where they might find this
[01:23:13] interesting we'd be happy to to discuss
[01:23:17] so that's kind of a wrap up of of what
[01:23:19] we did for the let's say large area lar
[01:23:23] inspection with high Precision that's
[01:23:25] coming out of the of the work we've done
[01:23:27] for EF and that we can apply uh luckily
[01:23:30] to to Industrial Solutions if if needed
[01:23:34] and now we we uh we can jump to kind of
[01:23:37] another area of development that we we
[01:23:40] uh took on some years ago based on or
[01:23:44] coming out of a European uh funded
[01:23:47] project uh it was actually I think it
[01:23:50] was the last of the FP SS at the time
[01:23:53] and this
[01:23:54] is uh basically what we tackled was the
[01:23:58] initial idea
[01:23:59] was making U Micro Devices for use in
[01:24:03] medical applications and one of the uses
[01:24:06] one of the devices was micr fluidic
[01:24:09] chips um our challenge within this
[01:24:12] project was to be able to inspect those
[01:24:14] chips in compatible with with production
[01:24:17] so inline inspection of microfluidic
[01:24:20] chips both in depth so because we wanted
[01:24:23] to be able to to quantify the the
[01:24:28] quality of the of the chambers of the
[01:24:30] microchambers for the fluidics in those
[01:24:33] micr fluidic
[01:24:34] chips and and then also in area so we
[01:24:39] had to be able to measure the details in
[01:24:42] area so we had ended up combining a what
[01:24:45] we call it a 3D High highspeed
[01:24:47] inspection based on on optical coherence
[01:24:49] tomography combined with Machine Vision
[01:24:52] we had a patent behind that as well so
[01:24:54] we started developing our own o system
[01:24:57] with a particular solution to make it
[01:25:00] quite robust and less sensitive to to
[01:25:03] the to the to the parameters or the the
[01:25:08] the noise or the the environment in a in
[01:25:11] a production
[01:25:12] plant um we use the the Machine Vision
[01:25:16] to guide the laser tool but also to to
[01:25:19] do the to do lateral inspection as I'll
[01:25:21] come in a little a little bit later and
[01:25:23] then we finally made some demonstrations
[01:25:26] of the of the system adapted for inline
[01:25:31] inspection so what we could what we
[01:25:33] achieved is death inspection of micr
[01:25:35] channels particular for fluidic chips uh
[01:25:39] combined it with that lateral 2D
[01:25:40] inspection as well so that whole idea is
[01:25:43] that as as the chips were going through
[01:25:45] a production line we could scan um but
[01:25:48] then we found that this could be
[01:25:49] adaptable uh different tool
[01:25:51] configurations we could have multi-layer
[01:25:53] samples as it's depending on how we
[01:25:55] configure the OC we could um uh use it
[01:25:59] to to Define diff to define the
[01:26:02] thickness of different samples as long
[01:26:04] as it was transparent to our wavelength
[01:26:06] and in the end we we were able to
[01:26:08] demonstrate functionality for that
[01:26:09] project and then also for some other
[01:26:12] potential use cases that we that we
[01:26:14] pitched in the past as well which I'll
[01:26:16] show you a little bit here in the next
[01:26:18] uh
[01:26:19] slide
[01:26:21] so just just a quick summary of of of
[01:26:23] how it uh you know how it went into not
[01:26:26] production but into something that we
[01:26:28] could test and demonstrate so we
[01:26:30] actually built the system itself it's an
[01:26:32] OC an OCT combined with Machine Vision
[01:26:35] we have a you know it's it has a
[01:26:37] particular
[01:26:39] illumination uh for the Machine Vision
[01:26:42] uh looking at the sample with a
[01:26:43] particular field of view that's that can
[01:26:45] be adapted depending on on the needs
[01:26:47] because we can we control the Optics the
[01:26:48] needs resolution it's a balance as we'll
[01:26:51] see then then it's divided we could call
[01:26:53] it into two channels the O Channel and
[01:26:56] the Machine Vision Channel then those
[01:26:58] two are combined and again with our our
[01:27:02] uh software team we were able to make
[01:27:04] the the the system to interpret the
[01:27:07] results and present to to a
[01:27:09] user so in the end we're able to to
[01:27:12] obtain a depth measurement lateral um
[01:27:15] depth and measurement and lateral
[01:27:16] measurement of micro features so here's
[01:27:18] an example of a microlitic chip how we
[01:27:20] can measure it in lateral and it's hard
[01:27:22] to show here but you can see the scans
[01:27:24] and this would be
[01:27:26] a just a a depth let's say a a vision of
[01:27:30] a depth scan that we would see actually
[01:27:32] usually would never see this per se but
[01:27:35] it's a information that we get out of
[01:27:37] the depth scan so like I said it's not
[01:27:38] as exciting as seeing the the lateral
[01:27:40] the lateral stuff the more visual things
[01:27:42] but it's that the the power of the
[01:27:44] system is to look through the through
[01:27:46] the material and look in depth again it
[01:27:48] was customized for inline inspection we
[01:27:51] also
[01:27:53] saw that we could adapt the technology
[01:27:54] to make it uh handheld for some
[01:27:57] particular industries that required to
[01:27:59] to do measurements in the field let's
[01:28:02] say um and there's other different
[01:28:04] configurations that that we could make
[01:28:06] for for different production
[01:28:08] lines um we could get depth and lateral
[01:28:11] accuracies down to two
[01:28:12] microns and and as I mentioned before we
[01:28:15] uh we have our software team that was
[01:28:17] able to optimize and and uh prepare the
[01:28:20] data acquisition for good user interface
[01:28:23] and and showing the
[01:28:26] results and finally in terms of and this
[01:28:30] is where again I think it could be
[01:28:31] interesting for the audience is where
[01:28:34] could we we found some applications
[01:28:36] besides the microlitic chips and inline
[01:28:39] inspection of microtic
[01:28:41] chips um we've we were able to
[01:28:44] demonstrate that we we could determine
[01:28:46] uh faults in microwelds and this
[01:28:49] particular space for needles but in
[01:28:50] general any kind of microwelds where
[01:28:52] they're where they're needed so it's a
[01:28:54] type of surface inspection as well and
[01:28:56] just in general different types of micro
[01:28:58] features that are required for for uh
[01:29:01] let's say this could be a needle patch
[01:29:03] which is something that that we worked
[01:29:04] on but in general different types of
[01:29:07] micro features that have a certain
[01:29:09] amount of area that are in line that can
[01:29:11] be susceptible to to a combination of
[01:29:14] depth scans through LCT and raster scans
[01:29:17] with our with our combined Machine
[01:29:19] Vision and that kind of sums up what I
[01:29:22] wanted to discuss about our our R&D
[01:29:25] developments in in combination of O and
[01:29:29] Machine Vision for inline inspection at
[01:29:32] at let's say relatively high speeds so I
[01:29:36] think the important thing I want to
[01:29:37] conclude with both of those lines that I
[01:29:39] just mentioned is that that we you know
[01:29:41] we have a really in-depth understanding
[01:29:44] of how the things work because we
[01:29:46] developed it and and we're quite open to
[01:29:49] adapting it to particular Solutions and
[01:29:51] work working with a partner partners
[01:29:53] that that want to uh drive us to a
[01:29:56] certain area of applicabilities
[01:29:58] and then I prepared the specific Li
[01:30:01] slide for the the very famous now epic
[01:30:04] question which is what can what can we
[01:30:07] do for you and and what can we do for
[01:30:10] you I summed it up in two areas it's
[01:30:12] customize solutions for optical
[01:30:13] Metrology based on what you've seen
[01:30:15] above I did mention it before but for
[01:30:18] other customers we already make uh you
[01:30:21] know we've helped the develop and make
[01:30:23] custom Solutions custom instrumentation
[01:30:25] for them and we develop them in series
[01:30:27] production which is my next point we we
[01:30:30] don't only stop at the development we
[01:30:32] can also help you uh deliver your
[01:30:34] products for your for your customers or
[01:30:36] or for you if you're if you're the
[01:30:38] customer but it's also very important in
[01:30:40] these uh meetings that that uh that
[01:30:44] there's a lot of things that we still
[01:30:45] would like to to have solve or or or
[01:30:47] things that we might need to improve our
[01:30:49] products or or our um our offerings so
[01:30:53] what can you do for us so some things
[01:30:56] that came to mind that are key were high
[01:30:58] sensitivity fast detectors that that's
[01:31:00] really key for the the types of things
[01:31:02] we're looking at into now so we're
[01:31:04] really open to to hearing some
[01:31:06] developments in that area if if anyone
[01:31:09] has
[01:31:10] any quite open again also related to
[01:31:13] Metrology uh Solutions is fast
[01:31:15] modulation single mode fiber coupled
[01:31:17] lasers that'd be really interesting to
[01:31:19] hear about developments in that area and
[01:31:22] then we're always always looking for
[01:31:24] Reliable partners for Precision custom
[01:31:26] Optics and and and Custom Manufacturing
[01:31:29] of mechanics so reliable meaning you
[01:31:32] know good good quality and good uh good
[01:31:35] good time and Lead times as well that
[01:31:38] would be very very interesting for us as
[01:31:41] well um so with that I conclude my
[01:31:44] presentation and uh thank you again epic
[01:31:46] for uh for this
[01:31:48] invitation thank you very much Andrea
[01:31:50] it's really interesting really nice nice
[01:31:52] projects and also thank you for this
[01:31:54] final slide showing uh the both sides of
[01:31:58] the Epic equation and we have already
[01:32:00] one question from the audience so for
[01:32:03] thank you again Manuel thanks for your
[01:32:07] presentation I have a question for you
[01:32:10] concerning your Optical
[01:32:13] designs for biochips microf frics as you
[01:32:18] say ER can you comment any on your
[01:32:23] Optical experience design including
[01:32:26] black coating to enhance sensitivity and
[01:32:31] contrast sure so uh so for that for this
[01:32:35] P thanks for the question J nice nice to
[01:32:37] speak with you again it's been a while
[01:32:40] um for for this particular
[01:32:43] applications the we the codings that we
[01:32:46] we used for the the Optics themselves
[01:32:49] were were let's say optimized codings
[01:32:52] for the particular wavelengths that we
[01:32:54] were working at which which would be
[01:32:55] near infrared or shortwave infrared so
[01:32:58] the Telecom wavelengths and so the the
[01:33:01] the Coatings are are multi-layer
[01:33:03] dialectric Coatings but I would say
[01:33:06] they're relatively standard codings that
[01:33:08] we would use um because of the
[01:33:11] controlled
[01:33:12] environment uh and the wavs that we're
[01:33:14] working on we weren't too worried about
[01:33:17] stray light from the mechanical
[01:33:19] components but I can say that um that
[01:33:22] that is something that does worry Us in
[01:33:24] other projects and and definitely
[01:33:26] codings is a key thing both so like I
[01:33:30] said maybe not in this particular
[01:33:31] application but in general to reduce
[01:33:34] scattering for sure there's a there's a
[01:33:37] there's value and in you know in
[01:33:39] exploring really good codings for both
[01:33:41] the Optics and and the
[01:33:43] mechanics yes yes yes to to to to
[01:33:46] precise we have a whole a industrial
[01:33:50] channel for other similar application
[01:33:52] concerning
[01:33:54] microoptics biochips assessments where
[01:33:58] black cing is a key
[01:34:00] factor to be included in Optical
[01:34:04] designs and it was surprising for me
[01:34:08] that okay you're not
[01:34:10] surprising that what my question was
[01:34:13] whether this additional not traditional
[01:34:16] I could say traditional codings
[01:34:17] obviously it's a key factor as well but
[01:34:20] black coding for the background of your
[01:34:23] experimental Optical setup whether you
[01:34:26] included it or not and it's okay it's
[01:34:28] okay seems is a challenging strategy for
[01:34:31] you thank you very much thank
[01:34:36] you I saw that Ral uh from mb6 course
[01:34:39] raising the hand but I don't know if uh
[01:34:41] you still have the question
[01:34:45] Ralph if not we can move to Andrea from
[01:34:48] qu print yes so is the super quality of
[01:34:52] the microf chips plays a role in the dep
[01:34:55] measurement and do you have any specific
[01:34:58] aspect ratio that limits the measurement
[01:35:00] say size versus
[01:35:02] depth so the depth so I I did quite
[01:35:05] catch your question about the surface
[01:35:08] quality um there is I um if it is if if
[01:35:13] I think I understood was if the surface
[01:35:15] quality like roughness affects the
[01:35:17] measurement yes okay so it it would
[01:35:20] because if it produces too much
[01:35:22] scattering it might it we would not be
[01:35:24] able to penetrate so there is a level of
[01:35:26] transparency that's required for this
[01:35:28] particular type of
[01:35:30] measurement um so because do you have a
[01:35:33] number like not really but uh the
[01:35:36] everything we expected was we could say
[01:35:38] Optical quality so in terms of RMS is
[01:35:40] probably down you know to pretty pretty
[01:35:43] good polished levels of of
[01:35:46] Optics um if if if what we're looking is
[01:35:49] just surface quality then it's not not
[01:35:51] so important because but since we're
[01:35:54] looking in depth and it was important
[01:35:55] that that it be relatively well polished
[01:35:58] but but the oct could be appli for
[01:35:59] surface and that's okay uh if if it's
[01:36:02] roughness but but for the microtic chips
[01:36:04] and to your second question the depth
[01:36:06] resolution is
[01:36:07] limited uh more by the configuration of
[01:36:11] our Optical uh or let's say our o setup
[01:36:14] which is a balance between the
[01:36:18] wavelength
[01:36:19] band and
[01:36:22] uh and and the and the the sensor that
[01:36:24] we're able to use you know how you know
[01:36:26] the resolution of the sensor um those
[01:36:29] those two things give you a combin and
[01:36:32] then you know the area that we're going
[01:36:33] to scan so in terms of aspect
[01:36:36] ratio we we really I'd say within those
[01:36:40] limits yes if we went uh if we went like
[01:36:43] over a millimeter we probably started
[01:36:45] not getting the resolution down to the
[01:36:48] to the low singled digigit microns that
[01:36:50] that we would um for the inspections
[01:36:52] that were
[01:36:53] required so that's that could be I don't
[01:36:56] know if that answers your question but
[01:36:57] that in terms of that that's where the
[01:36:59] limits were a little bit in in our
[01:37:01] experiment no that's clear thank you
[01:37:06] perfect I I have a final question for
[01:37:08] you andr because you have this
[01:37:10] experience with these large area
[01:37:11] measurements also difficult environments
[01:37:13] what do you think about the challenge
[01:37:15] that Frank from T brought today to the
[01:37:18] to the meeting yeah I took I took notes
[01:37:21] very challenging very challenging
[01:37:23] because it's kind of a from looking at
[01:37:26] it's like if I could combine the you
[01:37:29] know an adapted
[01:37:31] OCT and a large area scan then then we
[01:37:34] can probably help Frank solve that that
[01:37:37] problem I don't dare say anything
[01:37:39] because the resolutions of 0.1 microns
[01:37:41] were were difficult to meet but if we're
[01:37:44] at the Micron level range um because of
[01:37:47] our experience with the the scanning
[01:37:49] mechanisms um I think there there is a
[01:37:51] possibility that there is some
[01:37:53] application there but but again I didn't
[01:37:54] want to be uh I didn't want to overstep
[01:37:57] and say anything before but yeah I don't
[01:38:00] want to compromise just just your
[01:38:01] opinion about
[01:38:03] the yes I do think that there's some
[01:38:06] very interesting challenges that Frank
[01:38:07] put forth that might be uh at least
[01:38:11] tackled in some in some way with what
[01:38:13] we're
[01:38:14] presenting good so thank you very much
[01:38:16] andr for your nice presentation and
[01:38:19] showing us what you do in this in this
[01:38:21] field and um again if you want to reach
[01:38:24] Andre PR let us know or feel free to
[01:38:27] contact him directly for any any uh well
[01:38:30] possibility you have to cover their
[01:38:31] needs or any Metrology um need that you
[01:38:35] could have in the future thank you again
[01:38:37] Andress thank you and we go to uh our
[01:38:40] next speaker is Maura Dei is the
[01:38:44] director in business development at opto
[01:38:45] Fidelity and U the he's going to present
[01:38:48] his optical Metrology in AR is mar
[01:38:51] glasses so
[01:38:53] Murat can you see the presentation you
[01:38:56] are not in the presentation mode but we
[01:38:58] see the screen okay just a minute
[01:39:08] please what about now again is with the
[01:39:11] with the slides and and the notes all
[01:39:14] right sorry about that
[01:39:20] guys
[01:39:23] now it's good
[01:39:26] okay okay so thank you for organizing
[01:39:29] the event once more and introduction so
[01:39:31] my name is Murat de today I will present
[01:39:34] about how Optical Metrology is helping
[01:39:36] us overcome challenges with ar smart
[01:39:39] glasses or in general with immersive XR
[01:39:43] but let me quickly introduce uh who is
[01:39:45] opto Fidelity who we are so uh we began
[01:39:48] in the early 2000s with Optical
[01:39:50] Metrology ology and test automation our
[01:39:53] first VR Optical Metrology tester was
[01:39:55] already developed after the introduction
[01:39:57] of oculus and with the debut of holand
[01:40:00] we also uh introduced our first wave
[01:40:02] guide based Optical AR meterology and
[01:40:05] testing equipment since then we've
[01:40:07] delivered hundreds of different uh
[01:40:09] Optical Metrology systems to this uh
[01:40:11] particular domain although uh our
[01:40:14] company based in Finland we have a very
[01:40:16] Global presence uh our company has
[01:40:18] production facilities Europe USA and
[01:40:21] China we have two different offices also
[01:40:24] in USA where I'm also sitting and
[01:40:26] multiple different locations in the Apec
[01:40:28] and we help our customers uh with their
[01:40:31] all needed Metrology and testing
[01:40:33] equipments when they are developing and
[01:40:35] Manufacturing the next uh next uh
[01:40:38] generation of headsets and smart
[01:40:41] glasses so there are two main topics in
[01:40:44] this presentation one is focused on
[01:40:46] image quality and the other one is
[01:40:47] focusing on the latency but first let's
[01:40:49] even start by talking about uh AR smart
[01:40:52] glasses so I just want to say a few
[01:40:54] words about how they work uh there are
[01:40:57] various designs based on different
[01:40:58] Technologies as you also heard from
[01:41:01] different uh companies there are already
[01:41:03] different commercial products uh however
[01:41:05] to simplify this presentation and use
[01:41:07] the limited time more efficiently uh let
[01:41:10] me just focus on defract waveguide based
[01:41:12] Optical combiners that may or may not
[01:41:15] work with various different uh displays
[01:41:17] and micro displays and projectors such
[01:41:19] as lbs laser scanning alcol or LED based
[01:41:23] Solutions so there are two main
[01:41:26] components in this Essence light engine
[01:41:29] or light display engine so-called and a
[01:41:32] combiner so the light engine uh creates
[01:41:35] the light uh if it's a panel based light
[01:41:37] engine such as an alcos for example uh
[01:41:40] or a
[01:41:41] microed uh it would also use some uh
[01:41:44] illumination and projection Optics to
[01:41:46] cumate the light and direct it into the
[01:41:48] combiner then what's a combiner
[01:41:51] combiner transmits the real world images
[01:41:53] and displays the created images to the
[01:41:55] viewer's eye merging them to create an
[01:41:58] immersive experience so in an optical
[01:42:01] combiner uh an incoupling element
[01:42:04] couples the light into the wave guide at
[01:42:06] Angles above the critical angle so
[01:42:09] through this surface Rel greting for
[01:42:10] example this is based on TI total
[01:42:14] internal reflection phenomenon so the
[01:42:16] light will propagate down to length of
[01:42:18] the wave guide then another element out
[01:42:21] couples to light in a region directly in
[01:42:24] front of the user's eye here so these
[01:42:27] areas are known as input and output
[01:42:29] gradings and between them there is the I
[01:42:31] propile expansion area for example uh
[01:42:35] some of the most uh commercially sold
[01:42:38] products are already using this uh type
[01:42:40] of uh combination so such as lbs plus a
[01:42:44] diffractive optical
[01:42:46] element so now we know how they work so
[01:42:49] let's talk about what is expected from
[01:42:51] them so the requirements for the
[01:42:53] consumer level adoption of AR can be
[01:42:55] simplified into two categories as
[01:42:58] immersion and comfort this
[01:43:00] simplification was originally made by
[01:43:02] Bernard Gres currently the president of
[01:43:04] the SP uh so let's talk about immersion
[01:43:07] which is uh more related to this
[01:43:09] presentation immersion provides the
[01:43:12] feeling of being part of that simulated
[01:43:14] world this can be called also a presence
[01:43:17] presence would require different
[01:43:19] feelings which our eyes ears skin Etc
[01:43:24] can stimulate so there are multisensory
[01:43:27] ways of activating the presence then
[01:43:30] there is a temporal part of it which
[01:43:32] consists of movements mapping the
[01:43:34] environment and
[01:43:36] orientation uh this is usually studied
[01:43:38] under slam the abbreviation of
[01:43:41] simultaneous uh
[01:43:42] localization uh and mapping another
[01:43:45] important temporal parameter will be
[01:43:47] motion to Photon latency which I will
[01:43:50] specifically talk talk that talk about
[01:43:51] that later then there is the Comfort
[01:43:54] part of the experience which is quite
[01:43:56] subjective but there needs to be some
[01:43:58] sort of uh General agreement on what is
[01:44:00] comfortable and enjoyable however I will
[01:44:03] not go further on that due to limited
[01:44:04] time and it's not really relevant to
[01:44:06] Optical
[01:44:08] Metrology so how can we fulfill these
[01:44:12] expectations earlier we set two primary
[01:44:14] components the combiner and the light
[01:44:16] engine make the AR display architecture
[01:44:19] so to build such an a display we need to
[01:44:21] select the right combination of these uh
[01:44:23] com components we need to select the
[01:44:27] uh uh we need to select the right
[01:44:30] combination by making some tradeoffs by
[01:44:32] making uh trade-offs based on some
[01:44:34] analyzis for example it could be a
[01:44:36] tradeoff between the field of view and
[01:44:38] the angular resolution or uh the
[01:44:41] coupling efficiency of the combiner with
[01:44:43] the brightness for example which is the
[01:44:46] case with the micro LEDs for example
[01:44:49] somehow we must evaluate are design
[01:44:51] selected components individual
[01:44:53] performances and eventually the fully
[01:44:56] assembled AR iwes Performance and
[01:44:59] functionality but uh what do we need to
[01:45:02] have to perform those assessments so we
[01:45:04] need to have some tools that would
[01:45:06] simulate the human visual system as much
[01:45:09] as possible so that meaningful analyzis
[01:45:11] can be made and the results would be
[01:45:14] satisfactory simulating the human visual
[01:45:17] system will require the performance
[01:45:18] knowledge of the human eye such as the
[01:45:21] foe resolution sharpness contrast and
[01:45:25] more mechanical and statistical data
[01:45:26] such as how many what is the ipd uh for
[01:45:30] these uh age criteria and for male and
[01:45:34] females what is the I relief in average
[01:45:36] what is the iul size for males and
[01:45:38] females and what if uh under different
[01:45:42] scenarios such as what is for the
[01:45:43] teenagers what is for the uh Elder
[01:45:45] people so here the challeng is using
[01:45:48] this knowhow and building such
[01:45:49] evaluation tools to mimic the human
[01:45:52] visual system or the human eye in
[01:45:56] particular so now let's uh little bit
[01:45:58] talk about image quality here so on the
[01:46:01] right side you can see that uh what is
[01:46:04] projected into an AR wave guide and the
[01:46:07] resulting uh virtual image that appears
[01:46:09] to the user these are the real images
[01:46:11] that we capture and of course the
[01:46:13] expectation is the what that we are
[01:46:15] giving uh through the input grading or
[01:46:17] through the displays other uh other ways
[01:46:20] and the left side you can see what is
[01:46:22] given again through the input grading
[01:46:23] and what is taken out from the output
[01:46:25] grading so the question comes down jet
[01:46:28] does this product produce a nice image
[01:46:31] if so how nice it is and can I turn this
[01:46:33] into a metric and quantify
[01:46:36] it so before designing an optical
[01:46:39] Metrology system for near to eye
[01:46:41] displays or any type of uh smart glasses
[01:46:44] and their components such as the wave
[01:46:46] guide combiner in this particular case
[01:46:49] one needs to study the human very well
[01:46:51] which our own Optical engineering team
[01:46:53] have been studying studying it for years
[01:46:55] and they have patented several of their
[01:46:57] findings and we turned those inventions
[01:47:01] uh into our core elements when we are
[01:47:03] building this Optical Metrology tool so
[01:47:06] some of those uh core components have
[01:47:08] become for example this human eye
[01:47:10] mimicking lens this opto ey this I
[01:47:12] referred to it in the first uh bullet
[01:47:14] point so this is a type of uh double
[01:47:18] Imaging conoscopic lens that allows I
[01:47:21] point alignments and a single shut of
[01:47:23] over 100° fov 100 uh 100 degree of fov
[01:47:28] which is which is which is a little bit
[01:47:30] smaller than what we are usually having
[01:47:32] as a stereoscopic uh human visual uh fov
[01:47:36] but in real life uh we have this also
[01:47:39] selective uh selective visual uh visual
[01:47:43] visual system so that maybe in maybe all
[01:47:46] the time we are also limiting ourselves
[01:47:48] to less than 100 Dee fov so this is good
[01:47:52] enough to mimic the humanize radiometric
[01:47:55] properties so far and then there is the
[01:47:57] color uh there's the color parameter of
[01:48:00] this so there should be some sort of
[01:48:03] there should be some sort of uh system
[01:48:05] to mimic the colorometric properties of
[01:48:08] the human eye so we invented a filter
[01:48:11] wheel colomer and we coupled that with
[01:48:13] this opto eye lens so is in this picture
[01:48:16] over here so that's how uh we are now
[01:48:19] simulating the human colorometric
[01:48:21] properties then there's the projector so
[01:48:24] the projector is used to project
[01:48:26] something through the input grading if
[01:48:28] you remember the previous slides so you
[01:48:30] have to put something in order to get
[01:48:32] something out of it so with this
[01:48:34] projector uh we are putting all these
[01:48:37] type of different uh retical so that we
[01:48:41] will be able to see different patterns
[01:48:43] uh those are projected from the input
[01:48:45] grading and we are analyzing the results
[01:48:48] through the output grading so in its
[01:48:51] Essence this is type of a powerful and
[01:48:53] customizable uh RGB LED light source so
[01:48:57] we have this reticle changer this is
[01:48:59] motorized so you can have different
[01:49:01] patterns uh comes down to the wave
[01:49:04] guyses input grading and then it changes
[01:49:06] it and then you are seeing it and then
[01:49:07] you are going through the different
[01:49:09] pattern for
[01:49:10] example then uh we have to design we had
[01:49:13] to design a highly accurate robotic
[01:49:16] platform to com to accommodate these
[01:49:18] lenses and also different uh different
[01:49:21] Optical meterology tools so the problems
[01:49:23] were that you have to have uh such a
[01:49:26] tool to compensate a large uh area of
[01:49:30] pantoscopic tilts and face wrap angles
[01:49:32] so pantoscopic tilt uh is the uh is the
[01:49:36] angle of the is the angle of the glasses
[01:49:39] that you put it it's it's on this
[01:49:40] direction and the face wrap is on on
[01:49:43] this direction for example so with is
[01:49:46] multiaxis we turn into a station with 16
[01:49:49] axis with this this allowed us to
[01:49:52] perform this type of measurements so we
[01:49:55] can accommodate both this type of wave
[01:49:57] guys and we can also accommodate uh
[01:49:59] fully assembled glasses for example so
[01:50:02] how we have done this uh the alignment
[01:50:05] uh and also the precise movements were
[01:50:07] very challenging so we utilized lots of
[01:50:09] different autocom and we also uh
[01:50:12] included confocal distance sensors and
[01:50:15] as you can also see it over here we also
[01:50:18] have a highly accurate hexapod over here
[01:50:20] to uh to take uh to take advantage of
[01:50:23] that pile accurate six axis movement and
[01:50:26] uh in addition to that we also have
[01:50:28] different type of machine Vision system
[01:50:31] to look at for example the fiducials of
[01:50:33] the wave guides so that we know where we
[01:50:35] are actually relatively located uh from
[01:50:38] the input grading to the output grading
[01:50:41] once the optical setup and the Robotics
[01:50:43] are integrated so this sort of unique
[01:50:46] air Optical meterology system uh have
[01:50:48] been has been born so with this new
[01:50:50] system now we already have AR customers
[01:50:53] AR companies so they're able to measure
[01:50:56] uh this type of uh image quality metrics
[01:50:59] so the most Commons uh most common uh
[01:51:01] metrics over here are a field of view uh
[01:51:05] Distortion and amtf uh and also I will
[01:51:08] say the contrast for example efficiency
[01:51:10] can be measured quite easily so it's
[01:51:13] basically means that what you are giving
[01:51:15] in terms of uh let's say uh in terms of
[01:51:19] uh light light power and what is what
[01:51:22] are you getting out of out of it from
[01:51:24] the output grading or it could be also
[01:51:25] seethrough efficiency you can just
[01:51:27] imagine that you are putting a
[01:51:28] flashlight through a glass what you are
[01:51:31] what you have been putting it here and
[01:51:32] what is uh what is it that you are
[01:51:34] getting out of it from the other side
[01:51:36] and the ratio would be the efficiency
[01:51:38] for example and here I put a small
[01:51:41] graphs here to to provide you the
[01:51:44] understanding of what kind of spectral
[01:51:46] transmitters curves that we are getting
[01:51:48] these are pretty close to the ideal Tri
[01:51:51] stimulus XY Z
[01:51:54] curves then uh after measuring this
[01:51:57] image quality uh of AR wave guys we then
[01:52:00] started to question is the grading
[01:52:03] quality high enough to even Ensure High
[01:52:05] image quality so while searching for
[01:52:08] this answer we realize that there is not
[01:52:10] even an optical Metrology tool to
[01:52:12] measure the gradings with the needed
[01:52:14] accuracy there are some traditional
[01:52:16] tools such as asms but the accuracy and
[01:52:18] repeatabilities
[01:52:20] were not good enough to measure this
[01:52:21] type of AR wave guides then uh we were
[01:52:24] able to build a lro diffractometer to
[01:52:27] measure grading pitch and
[01:52:28] orientation if you unfamiliar with LR
[01:52:32] diffractometer a lro diffractometer is
[01:52:34] an optical device that uses a laser beam
[01:52:37] to illiminate a sample the grading
[01:52:40] reflects the defraction order back to
[01:52:42] the laser which is then directed toward
[01:52:44] the camera via beam splitter with direct
[01:52:47] reflection from theg grading we can
[01:52:49] calibrate the sample rotation stage to
[01:52:51] measure the fraction angles accurately
[01:52:53] by building this Optical Metrology tool
[01:52:56] we achieved a way to analyze the AR
[01:52:58] waveguide gradings on the left side here
[01:53:01] uh the pixels color uh indicates how
[01:53:04] much the grading period differs from the
[01:53:06] nominal period and it's also even
[01:53:08] detecting the defects which has a direct
[01:53:10] impact on the image quality so this type
[01:53:13] of non-uniformities uh we have done uh
[01:53:16] quick studies uh these are directly
[01:53:19] impact the for example the color
[01:53:21] uniformity of the AR wave GES but the
[01:53:23] future work will be about linking uh
[01:53:26] more creating more links between the
[01:53:28] image quality results and the grading
[01:53:30] problems such as those mentioned
[01:53:32] nonuniformities and how we can perform
[01:53:35] these measurements also
[01:53:40] quicker okay and the last topic is about
[01:53:44] latency so let's now jump into a little
[01:53:46] bit different stuff so we said that AR
[01:53:49] is mostly about immersion and creating a
[01:53:51] physical presence in a non-physical
[01:53:53] world uh if we want to assess the
[01:53:55] temporal performance then we have a
[01:53:56] couple of challenges one of them is uh
[01:53:59] producing end to endend uh non-intrusive
[01:54:01] frame by frame analyzes of the AR
[01:54:03] displays performance then another one is
[01:54:06] producing these analyzis while moving in
[01:54:08] at least three degree of Freedom or
[01:54:10] preferably in six degree of Freedom
[01:54:12] which we would do it in the real life by
[01:54:14] wearing those type of headsets we will
[01:54:17] be moving around we will be walking
[01:54:18] around so it's in three degree of
[01:54:20] Freedom then comparing these results of
[01:54:22] the virtual contents posst to the real
[01:54:24] verse post by doing this we will get the
[01:54:27] knowledge of motion to Photon latency
[01:54:29] and word tracking performance so what is
[01:54:31] motion to Photon latency latency in
[01:54:33] general is the difference between action
[01:54:35] and reaction and in AR we referred
[01:54:37] motion to Photon latency as the amount
[01:54:40] of time between the user's head movement
[01:54:42] that is the action and it's
[01:54:43] corresponding displays output
[01:54:45] Reflections that is the reaction on the
[01:54:47] display users should not experience a
[01:54:49] delay between the physical movement and
[01:54:51] the display output to have the best
[01:54:53] immersion otherwise uh the sense of
[01:54:56] physical presence uh in a virtual world
[01:54:58] will be lost and then anyway any MTP
[01:55:02] latency of more than 20 milliseconds uh
[01:55:05] are introducing uh the dizziness and the
[01:55:07] VR motion sickness a low of motion
[01:55:09] Photon latency will anyway improve the
[01:55:11] Hologram stability so that a good
[01:55:14] immersive experience would require a
[01:55:16] very low motion to Photon latency
[01:55:18] ideally less than 5 milliseconds for
[01:55:21] these reasons uh reducing motion to
[01:55:23] Photon latency was critical to provide a
[01:55:26] a good immersive experience to Consumer
[01:55:28] AR users I also mentioned about word
[01:55:31] tracking performance this is mainly
[01:55:32] referring to find out for example uh
[01:55:35] what happens uh when there's a user
[01:55:37] related changes such as it starts to
[01:55:39] move or what happens the lightning
[01:55:41] conditions are changing in for example
[01:55:43] you in the room or you are going outside
[01:55:46] such bright conditions it affects the
[01:55:47] image quality for example so we we
[01:55:49] needed to have a tool to figure this out
[01:55:51] this Challenge and in that case uh these
[01:55:55] uh these were needed because our
[01:55:57] customers reported those and we took
[01:55:58] this Challenge and patented a way of
[01:56:00] measuring endtoend latency so we made a
[01:56:03] constellation map which is visible here
[01:56:06] that consist of absolute markers with
[01:56:08] binary cost uh shown as blops these
[01:56:11] small uh tiny blobs then we install this
[01:56:14] to the AR device as an application these
[01:56:17] absolute marker patterns move in the
[01:56:19] virtual word uh using the a device's own
[01:56:22] virtual word engine such as the open XR
[01:56:24] or Unity or some other stuff then we
[01:56:26] needed to observe and track these blobs
[01:56:28] as an observer we have a smart camera
[01:56:31] inside the eye slows of a 3D printed
[01:56:33] head just like the human eye would look
[01:56:35] at the glasses we simulated that by
[01:56:37] putting smart cameras into ey Sloss of a
[01:56:39] 3D printed head then we made uh this 3D
[01:56:43] printed head to wear the AR device then
[01:56:45] camera looks at the center of the sphere
[01:56:47] so that the 3D virtual vir orientation
[01:56:50] is detected when the camera when the
[01:56:52] camera is looking at these blobs in the
[01:56:54] virtual world and then we represent the
[01:56:57] results in different uh different angles
[01:57:00] such as in yo pitch and roll and then we
[01:57:02] compare the robot's motion which we take
[01:57:05] it from the encoders then we say that
[01:57:07] okay you are this off in terms of the
[01:57:10] latency when you are viewing this uh
[01:57:12] virtual result I'm about to finalize it
[01:57:15] so very quickly uh so AR displays are in
[01:57:19] constant development mode we don't know
[01:57:21] what will be the future maybe there will
[01:57:22] no there will be not a wave guide based
[01:57:25] solution uh in the near future but we
[01:57:27] will work on it we will develop more
[01:57:28] Optical Metrology systems and we are
[01:57:31] constantly listening our customers and
[01:57:32] the market and we are also looking for
[01:57:34] collaborations as mentioned uh we are
[01:57:37] only we are only testing here two
[01:57:39] different uh components essentially the
[01:57:41] light engine and the combiners but there
[01:57:43] are hundreds of different components in
[01:57:45] an AR smart glasses and if you are in
[01:57:47] this domain please come to us and
[01:57:49] explain us more more challenges with the
[01:57:51] different components we are ready to uh
[01:57:54] take those challenges in and turn those
[01:57:55] into a solution and also the mass
[01:57:58] production phase eventually will come
[01:58:00] for AR smart glasses and we are looking
[01:58:02] for ways to improve our testing times so
[01:58:05] if you have uh if you are an automation
[01:58:07] company if you are working on these
[01:58:09] domains also you can contact us uh and
[01:58:12] we can collaborate on ideas of how to
[01:58:14] improve the testing times also so thank
[01:58:16] you very much for listening it was uh
[01:58:18] yeah quite but I hope that I covered
[01:58:21] enough thank you very much Murat for
[01:58:24] this good review about all the
[01:58:27] capabilities all the Metrology tools you
[01:58:28] are building and also the challenges for
[01:58:31] the future because as you mentioned is a
[01:58:33] um we are starting with this kind of
[01:58:34] products and uh it's um there are some
[01:58:37] challenges probably
[01:58:38] undiscovered uh we have time for one
[01:58:42] question or comment regarding potential
[01:58:45] collaboration with Murat and his team
[01:58:54] well if not feel free to contact uh us
[01:58:56] back and um we will put you in contact
[01:58:58] with Murat to discuss more more in
[01:59:01] detail and let's go with the final
[01:59:03] speaker for today is uh I have the
[01:59:05] pleasure to introduce Christian Brock
[01:59:07] say manager at AutoCraft and we are
[01:59:09] moving now to the measurement of the
[01:59:11] optical quality of Windows and lenses
[01:59:13] you using uh the way from Metrology so
[01:59:16] Christian the floor is
[01:59:18] yours
[01:59:20] yeah welcome um and I hope you can see
[01:59:25] my screen now yeah everything
[01:59:27] good uh so I'll talk about the
[01:59:29] measurement of the optical quality of
[01:59:30] Windows and objective lenses using
[01:59:33] wfront Metrology start with a short
[01:59:35] company overview
[01:59:37] so we are
[01:59:41] um we are building the shakman wfront
[01:59:43] sensors is our core um Metrology system
[01:59:48] that our customers use used for
[01:59:49] different applications in R&D in
[01:59:51] production for integration into their uh
[01:59:54] uh manufacturing lines uh then there are
[01:59:56] subsystems and modules which will
[01:59:59] combine the the wfront sensor with a
[02:00:02] light source and will will Define a uh a
[02:00:05] measurement plane and uh then we also do
[02:00:07] the the turnkey Solutions uh for
[02:00:11] dedicated uh tasks so we are
[02:00:15] um uh we are based in U Southern Germany
[02:00:19] 35 people
[02:00:22] currently um we are doing this
[02:00:25] developing the software inhouse and
[02:00:27] doing the mechanical engineering and we
[02:00:28] have our team of optical Engineers to
[02:00:31] take care about these uh applications
[02:00:34] and of course there is uh the other um
[02:00:38] departments and uh since a few years we
[02:00:41] are a part of the the micro Epsilon
[02:00:43] group that uh I think many of you should
[02:00:48] know
[02:00:49] um now for the um for the testing of
[02:00:52] Plano Windows the first application I
[02:00:55] want to talk about uh the basic idea is
[02:00:58] um you have here your Optical system
[02:01:00] under test you have a cated light source
[02:01:03] that you shine onto the uh system and
[02:01:06] then you are measuring the transmitted
[02:01:07] wfront and uh in in case the the system
[02:01:12] is just perfect you will get a wfront
[02:01:14] which is flat which shows no operations
[02:01:17] at all um in case you have an optical
[02:01:20] system which has some
[02:01:22] aberration on one of the surfaces or
[02:01:24] maybe also inside in the bulk material
[02:01:28] um you will get a transmitted wavefront
[02:01:29] that then carries these aberations so in
[02:01:32] my example here we would have a such a
[02:01:35] coma like aberration like you have a
[02:01:38] valley here and a peak here and that
[02:01:39] will imprint onto the transmitted
[02:01:42] wavefront we measure it and we can
[02:01:43] directly qualify the aberration of the
[02:01:46] sample under test in that way
[02:01:49] um and um a real setup would look like
[02:01:55] this so you have a light source maybe a
[02:01:57] fiber coupled like source with a cation
[02:01:59] lens could be the illumination unit on
[02:02:01] one side and then there comes your
[02:02:03] sample under test and then on the other
[02:02:06] side you have a detection unit
[02:02:07] consisting of the shakman wfront sensor
[02:02:09] and the Kepler
[02:02:11] Telescope um
[02:02:14] in and in that way we can also rescale
[02:02:17] basically the the det area of the
[02:02:19] wavefront sensor to the needed uh field
[02:02:22] of view to the needed measurement area
[02:02:24] of the
[02:02:25] sample um this is a typical single pass
[02:02:30] setup um and the advantage of this is
[02:02:32] where we we we can do a very nice
[02:02:35] referencing so we can just take a
[02:02:37] reference measurement of the empty
[02:02:40] system um and then and then we insert
[02:02:42] the sample under test and then we get
[02:02:45] only the operations of the sample under
[02:02:47] test because we can subtract
[02:02:49] the the operations of the empty system
[02:02:51] and in that way we get can get
[02:02:53] accuracies up to 2 nanometers RMS
[02:02:56] measurement
[02:02:59] accuracy um so here's an example where
[02:03:01] we did some lot testing of uh ND filters
[02:03:04] gr filters and you see here the the
[02:03:07] wavefront RMS RMS of the transmitted
[02:03:10] wavefront um and yeah we can clearly
[02:03:14] Identify some of the samples by their uh
[02:03:17] increased RMS value meaning reduced uh
[02:03:20] Imaging
[02:03:21] quality um and just I picked a few
[02:03:24] examples here of this this one on the
[02:03:27] which has a very low RMS value where you
[02:03:29] get more or less a flat wavefront or for
[02:03:32] example this one here where we have a
[02:03:34] strong defocus or another one where
[02:03:36] there's more like an astigmatic
[02:03:38] wavefront you can see that here on this
[02:03:40] sadle shape aberration so on one side
[02:03:43] the slide tells you okay we can when we
[02:03:45] do a lot testing we can do uh good or
[02:03:48] bad bad uh identification of good parts
[02:03:51] of bad parts but in addition what the
[02:03:55] the the transmitted wavefront tells you
[02:03:57] is also the reason at least gives you
[02:03:59] clues or hints on the reason for for the
[02:04:03] for the good or bad quality that you're
[02:04:04] measuring right like in in this case
[02:04:06] here seems like a global defocus on the
[02:04:09] on the plan of window so it's not really
[02:04:11] plain it's it's like warped and and
[02:04:14] defocus or this one here has this settle
[02:04:17] shape operation on it
[02:04:19] okay so um this is a short example of
[02:04:23] lot testing and um yeah quity testing
[02:04:27] now for objective lens testing these are
[02:04:29] not Plano Windows now we have uh Optics
[02:04:32] that create an image of focus um we
[02:04:35] basically we can use this uh um double
[02:04:40] pass setup which is not so different um
[02:04:43] we have here our um our illumination
[02:04:45] unit again consisting of a fiber coupled
[02:04:48] light source col lens now we couple this
[02:04:51] into the experiment via a beam
[02:04:54] splitter and uh as we go through the
[02:04:57] sample under test here now our objective
[02:04:59] lens that we are testing is creating a
[02:05:02] focus um and we need to use a reference
[02:05:04] sphere which is basically a concave
[02:05:06] mirror with a very good quality to uh um
[02:05:10] reflect back the the light that comes
[02:05:12] from the uh from the sample under test
[02:05:15] so each Ray that is leaving here um is
[02:05:17] reflected back into itself and then uh
[02:05:22] is transmitted through the sample a
[02:05:23] second time meaning we are collecting
[02:05:26] the operations of the samples twice um
[02:05:30] and then we have our detection unit
[02:05:32] again here where we have a relay
[02:05:34] telescope that um does sample Imaging um
[02:05:39] right so that way we can test uh for
[02:05:42] example objective lenses used in uh
[02:05:45] mobile devices or in microscopy or also
[02:05:48] in automotive industry U might have an
[02:05:51] example here for microscope objective
[02:05:53] lens which was a 40x 40 magnification
[02:05:56] lens was numerical aperture
[02:05:58] 65 um and we measured two lenses with
[02:06:02] the same same
[02:06:04] design uh so both na 0.5 lenses and one
[02:06:09] uh lens showed uh this strong aberration
[02:06:13] here spherical aberration that you can
[02:06:15] see and the other was a relatively good
[02:06:17] quality that we can also see here we
[02:06:21] measure a St ratio of 75 here we have a
[02:06:23] St ratio of 0 98 and if we look a bit
[02:06:26] into the xiki decomposition uh we can
[02:06:30] see that actually the reason for this uh
[02:06:33] difference in in in imaging quality is
[02:06:36] coma and spherical aberation that we
[02:06:38] have uh here for this uh lens with lower
[02:06:42] Optical quality we have a coma and
[02:06:45] spherical aberration which are nearly
[02:06:48] not present in this lens number two
[02:06:50] which has a very good Imaging
[02:06:55] quality um so as was mentioned before
[02:06:59] the the shakman wavefront sensor has a
[02:07:01] strong point also in uh being able to
[02:07:03] measure at different wavelengths and
[02:07:06] here's an example where we tested such a
[02:07:08] smartphone lens um at three wavelengths
[02:07:10] at the blue the green and the the red
[02:07:12] wavelength range um and as you can see
[02:07:16] the the the transmitted wavefront is
[02:07:18] different
[02:07:19] um and actually the uh the best
[02:07:21] performance of this lens was in the
[02:07:23] green for for 532
[02:07:26] nanom um so it seemed this lens was
[02:07:29] optimized for the for the green uh
[02:07:31] wavelength range and in the blue or in
[02:07:33] the in the red we get a little bit
[02:07:35] increased um wavefront pictor Val value
[02:07:39] meaning a little bit decreased Optical
[02:07:43] quality so in addition this this system
[02:07:46] would do field testing of exis testing
[02:07:48] but here I just wanted to show you the
[02:07:51] um as an example the the the measurement
[02:07:54] capability for different measurement at
[02:07:56] different wavelengths and um what the
[02:08:00] wavefront sensor can do there so um now
[02:08:03] I'm already at the end of my talk um so
[02:08:06] what we can do for you but basically we
[02:08:09] share our um expertise in Optics testing
[02:08:11] and deliver solutions for uh Optics
[02:08:14] testing and um we are always uh looking
[02:08:18] for for for special Optics um and
[02:08:21] special cameras that we use in our um in
[02:08:23] our
[02:08:24] systems um like for example uh special
[02:08:28] Optics would be also off AIS parabolic
[02:08:31] mirrors with with high Optical quality
[02:08:33] such um such
[02:08:36] Parts um current challenges are the
[02:08:40] measurement of large free from Optics
[02:08:41] that's I think this is not only a
[02:08:43] challenge that we are facing that's a
[02:08:45] topic that is of Interest generally and
[02:08:48] uh what we are currently also working on
[02:08:50] is product modularization so that we are
[02:08:53] uh able to measure our to to to to uh
[02:08:56] deliver our uh TurnKey Solutions um in a
[02:09:01] modular way so we have we have the
[02:09:03] possibility to to customize but uh off
[02:09:06] the shelf customization so that's the
[02:09:08] goal okay so with that I would like to
[02:09:11] uh thank you for your attention uh and
[02:09:15] um if you are interested in further
[02:09:18] information you feel free to contact me
[02:09:21] uh us and uh or maybe there's still time
[02:09:24] for one last question
[02:09:26] today we have time for question for you
[02:09:28] Christian if there is somebody in the
[02:09:30] audience I have one at least but I will
[02:09:33] wait if somebody else have a question
[02:09:35] for
[02:09:38] you if not I I would like to ask um if
[02:09:42] if if you try adaptive op Optics
[02:09:46] Solutions in your device devices uh
[02:09:49] because you're talking about this wave
[02:09:51] sensing but I don't know if adaptive
[02:09:52] opst could make difference in some of
[02:09:54] the
[02:09:55] measurement now so um opcraft does not
[02:09:58] provide Adaptive Optics directly with a
[02:10:01] wavefront sensor but basically the
[02:10:02] wavefront sensor is set up to work with
[02:10:05] adaptic Optics like deformable mirrors
[02:10:08] so there uh the software interfaces are
[02:10:10] there and so that's an application that
[02:10:12] we are uh realizing also with some
[02:10:14] customers right M good and we have a
[02:10:18] question from
[02:10:20] Dominque hello hello everyone uh thank
[02:10:22] you very much for your time and your
[02:10:24] presentation I have question about meta
[02:10:27] Optics component like lenses and other
[02:10:30] complex functions are you able to
[02:10:33] discuss with me and my partners about
[02:10:36] this subjects how to measure meta Optics
[02:10:39] components with different Optical
[02:10:43] functions basically yes yeah um so the
[02:10:47] meta
[02:10:50] Optics they in the end what they do they
[02:10:53] also create a wavefront like uh which is
[02:10:55] which is in the at least looked from
[02:10:58] from a distance which is not so
[02:11:01] different in uh in uh in comparison to a
[02:11:04] to a standard optic so in that way if we
[02:11:07] are measuring just the overall quality
[02:11:09] of a of a metal lens that would not be
[02:11:12] so different of course if you're looking
[02:11:15] into the the the surface uh structure
[02:11:18] and characterization of that that would
[02:11:20] be a different topic and that for that
[02:11:21] you would not probably not use the
[02:11:23] shakman wfront sensor yes the I'm
[02:11:26] looking for a simple
[02:11:28] test having first results and not high
[02:11:31] Precision of the details of the
[02:11:34] structure okay may I contact you soon
[02:11:37] yes for sure very good thank you very
[02:11:40] much yeah thank you very good and uh if
[02:11:44] there is no more questions for Christian
[02:11:47] thank thank you very much uh for your
[02:11:49] time and for presenting your Solutions
[02:11:52] uh you see now a poll that uh we will
[02:11:55] really appreciate if you can answer so
[02:11:57] um we will improve uh for our next
[02:12:00] meetings and um just to to finish
[02:12:03] because I know that we are over time uh
[02:12:06] just thank you again to our sponsors uh
[02:12:09] so actar advaned codings Optics and
[02:12:12] moduli thank you very much for
[02:12:13] supporting us in this event and uh thank
[02:12:16] you for all deles for here all the
[02:12:18] discussion all the questions and um very
[02:12:21] happy also to see some points of
[02:12:23] collaboration between the attendees so
[02:12:26] thank you again and I hope to see you
[02:12:28] next
[02:12:29] time thank you byebye thank you by bye
[02:12:34] thank you bye good afternoon everybody
[02:12:37] thanks a lot goodbye goodbye thank you
[02:12:39] byebye