Interview by Ralph Grabowski with Jacques Delacour
Optis was founded in 1989 by Jacques Delacour, who has a background in optical simulation, such as lens design. He developed software for determining the path lasers take through optics. The Solstis (later Speos) software initially ran standalone on DOS and then on Windows.
But then he found that designers wanted his analysis software to run inside their CAD programs, and so now it is integrated into Solidworks, Catia v5, Creo, and NX. "We try to make it as easy as possible so that users do not need to be experts; the designer just needs to specify the light sources," Mr Delacour told me over WhatsApp.
Over the last 28 years, the company changed its focus to develop software that analyzes light, instead of optics -- working with both visible and invisible light, such as infrared and ultraviolet. (See figure 1.) "What you see is light; we want to generate the image from that," explained Mr Delacour. "One of our major differences is that we model human vision -- we start with light and then create rendered images."
Figure 1 Left: Virtual model in Speos; right: physical model
Optis likes to anticipate trends for products around the corner, and so is concentrating these days on developing software for VR [virtual reality] and self-driving cars. "We can validate the VR equipment itself, as well as what users see in the VR headset."
When it comes to self-driving cars, cockpit design is becoming especially important, because these cars need to have a UI that is different from today. (See figure 2.) For instance, there are the problems created for heads-up displays from the many light reflections inside cars. The software from Optis determines how light reaches the driver so that he gets the right info. "It is the future of the human interface," he says, "And so we have simulators that let designers anticipate things very early, like the glare on windows."
Figure 2 Driving simulator
One of the concerns of self-driving cars is the lack of communication between driver and pedestrian: how do the pedestrian and the car negotiate who will move first at an intersection? "For cars in the future, headlamps will project light and information onto the road. The headlamp gets data, such as from the GPS, and then displays an arrow on the road to show the driver which way to drive -- or that the car will drive, so that everyone else sees it too. Pedestrians will be able to tell which way the car will be going."
Automotive manufacturers are increasingly interested in mood lighting in cars, something that Optis analyzes for them.
Headquartered near the Mediterranean Sea in southern France, Optis now has 230 employees, most of them engineers, and is growing at a rate of 30% a year. It sees a third of its sales in its top market of Japan due to the country's big automotive and electronics industries. For instance, software from Optis is used to design the lenses of cell phones and projectors.
"Our biggest market is in automotive," explained Mr Delacour. "The manufacturers want to shorten the time to market by designing their cars without [physical] prototypes, and so we must reproduce the real world for them. We tell them that our software doesn't make their products look beautiful; it makes products look real -- because if it is not real, then it is not beautiful in the real world."
Customers fall into two groups: suppliers of equipment (such as radios, GPSes, and dashboards) like Bosch and Continenal; and OEMs who make the cars -- like Bentley, Ford, Jaguar, BMW, Tesla, Honda, and Mazda -- and who want to validate that the equipment from suppliers comes together in a way that is safe for the driver.
The number #2 market for Optis is electronics and lighting. At Solidworks World, the company is unveiling a new version of Speos for Solidworks that is focused on lighting.
I asked who his competitors are. There are several, but none that do exactly what Optis does. The others concentrate on lens design or on simulators, but not both design and simulation.
Future of Optis
"We believe in utter realism for VR, and so we needed to add another dimension to excite another sense: hearing the sounds that are related to what we are seeing. It is more realistic." And so in 2015 Optic acquired Genesis, also from France, to combine sound with vision, such as for driving simulators.
"We had the same kinds of customers, and so it made sense. For instance, self-driving cars have to have speakers to emit sounds that alert pedestrians." The company plans to add sound to VR.
"We also work with luxury goods, where our software reproduces the optical effects of materials, such as leather and brushed metal," he said. "With Genesis, we can do the same with acoustics, like the sounds that watches make -- very important to the luxury watch manufacturer -- or the click made by the cap going onto the top of an expensive bottle of perfume. We design the sound of the slam of a car door, where our software lets designers add materials to get the right quality of sound."
The company recently deployed Speos HPC, the multi-server version to speed up calculations. It can take a long time to determine light pathways, and so this new version is very efficient: 1,000 cores run the software nearly 1000x faster. If a rendering took 10 minutes before, it takes 0.6 seconds now. The aim is to produce very quick results so that designers can make very subtle changes. With HPC, designers can fine-tune designs in a way that could not be done before.
Mr Delacour (see figure 3) sees a big demand for VR and for lighting in new areas, such as lit materials. "Visit your new kitchen in VR! We always want to offer the ability to test something that isn't even available yet. My focus is on innovation, for ourselves and for others."
Figure 3 Opis ceo Jacques Delacour
He concluded the interview by relating to me a story from outer space. When the Huygens space probe plummeted towards Saturn's largest moon Titan, it was carrying instruments that measured all kinds of readings. But it had no video camera with which to record the trip. Optis was asked to take the measurement records, and create a video of what the descent through the orange atmosphere and arrival at the surface would have looked like.
Mr. Grabowski nails it in @upFronteZine #924: http://eepurl.com/cyzapf #NativeDWG #TryBCADnow #BetterCAD - BricsCAD USA (@bcadusa) via Twitter
- - -
By your review I have concluded that definitively AutoCAD 360 does not replace the vectorial functionality of previous well-known versions of CAD. On the other hand, if you cannot start from scratch (blank dwg) then that means I still need another application to create them and AutoCAD 360 also? This is totally absurd from Autodesk.
Are we looking at the fall stage of AutoCAD? What other options are available in the market? Too many questions, too little information. - Rafael Espinoza (via WorldcAD Access)
The editor replies: Autodesk withdrew that function from the free version of AutoCAD 360 to get us to purchase the paid version. But we don't need to buy another CAD program to start a new drawing: the workaround is to use a free CAD system like DraftSight to make the starter drawing.
AutoCAD 360 is the first step towards a full version of AutoCAD on the Web, available only by subscription. From statements made by Autodesk execs, it appears this project is at least two years behind schedule -- thankfully!
Mr Espinoza responds: Thanks for your explanation, Ralph. Happy to know that others (as well as me) are aware and worried about the future of CAD. It is a way of life for me as a draftsman.
Re: What the Future Holds for MatureCAD
Just read yourupFront.eZine #923 and I am always interested to read your and others views on products and situations. I definitely don’t want to be considered as one who subscribes to solipsism!
Just a thought you might want to mull over re your comments on cloud-based tech. The value propositions espoused by vendors are many (and possibly varied), but can I also offer the suggestion that cloud-based also allows one to more practically manage (possibly even alleviate) software piracy?
Think on it: If one can turn a (significant) proportion of would-be pirated software users into cloud software users -- because there’s no other route to use one’s apps -- that’s a (financial) win for the software author. This might be obvious, and nothing wrong with that IMHO, but I have not seen much on that cloud value proposition from (any) vendors. "Just my thoughts," as Oleg would say..
The editor replies: While your proposal certainly benefits software authors, it is not on the check list of regular users. Corporations do worry about being charged for piracy, so that's a different matter.
The only cloud service I use regularly is MailChimp and the constant lag is frustrating for me; I expect software to run at my speed. (I have here the fastest business Internet service available.) I use MailChimp only because my ISP (telus.com) wouldn't let me email several thousand newsletters anymore, and I was tired of the years of fighting them.
I think for me the #1 reason to not use cloud software is the unwanted and unneeded changes made to software unexpectedly. MailChimp, for instance, changed its start screen to a dashboard -- completely of no interest to me, but just one more barrier to me getting my job done as fast as possible.
Mr Behrens responds: Just to be clear, I was thinking of the topic regarding vendors propositions, not the users, so not in disagreement with you, just sayin’. One might even surmise that’s why some vendors have the ability to ‘lower’ costs in the first instance, but that’s a longer and possibly adjacent discussion.
I quite understand re your point about software changes, and that’s an issue that most have to face and some/many complain about. On-premise makes that irrelevant but obviously only applicable to large corporate users so no use to you.
On the other hand I’ve spent many hours talking with vendors who’re frustrated that their users complain of missing features/functions when it’s in a new release. Cloud-based makes that issue a non-starter. Damned if you do/damned if you don’t. Also, (some) vendors struggle to justify and maintain their subscription revenues, which is a big revenue stream. Cloud makes that problem go away; again, devil's advocate: good news, bad news.
Re your performance issues. I spoke to Al Dean last year and he was scathing on one cloud-based CAD app on that very topic. No doubt we’ll all look back in 10 years and laugh about the bandwidth/performance we have today -- just like we do when we look back at having to use slow 68000 processor workstations, VAXes, and 8086 PCs for new releases of software (for us who remember those processors).
Same with our connected evolution from modems to ISDN to various incarnations of broadband and associated larger amounts of data that we regularly transmit/receive. Unfortunately, it doesn’t solve your problems, but that’s technology evolution for you. I empathize, but bearing in mind the commercial reality of a company like Autodesk (for example), what’s a good solution? If you can figure that out you’ll make a fortune!
The editor responds: I fired up Onshape last week for the first time in several months (I'm writing about it for a magazine), and it told me about only one up-date -- a down-date, as it turned out -- where free users no longer get the five private drawings; all drawings are public now. (Clearly, they have too many free users.)
What else does it do that's new? Dunno. To learn, I'd have to scroll through past blog entries, which the program does not link to.
(Heh, came up with this line for the article: "If you like your software, you can't keep your software," to misquote a former president. )
Thank You, Readers!
Thank you to readers who donate towards the operation of upFront.eZine.
Should you wish to support upFront.eZine through PayPal, then the suggested amounts are like these:
Should Paypal.me not operate in your country, then please use www.paypal.com and the account of firstname.lastname@example.org.
Or mail a cheque (US$ or CDN$ only, please) to upFront.eZine Publishing, Ltd., 34486 Donlyn Avenue, Abbotsford BC, V2S 4W7, Canada.
"Thinking is not a team sport. Deciding what you think is not a collaborative project to undertake with 1000 of your Very Bestest Fake Internet Friends on Twitter." - Ace http://ace.mu.nu/archives/366811.php