2015 marks the third year I’ve been fortunate enough to have been invited to COFES, the Congress on the Future of Engineering Software . Thank you, Cyon Research.
One way to look at COFES is this: invite the glitterati of the CAD and CAE worlds – plus an oddball or two – to a hotel out in the desert (aka Scottsdale) for a couple of days, toss in some catalytic keynote presenters, and then force the folks to wander around and interact. I often feel like the proverbial fly on the wall, which is exactly what Cyon’s Brad Holtz doesn’t want. (Sorry, Brad). But there’s a lot of gray beards and gray matter at COFES and it’s hard sometimes not to just sit back and gawk.
Also at this year’s COFES was a special pre-conference program: several presentations summarizing the first Analysis, Simulation, and Systems Engineering Software Summit (ASSESS). The obligatory joke to make about that event’s acronym is the organizers very carefully and purposefully added the word “Software” to the end of event’s name. I wrote about the post-event news conference back in January.
I’m going to comment on the Report from ASSESS in a separate blog post. So what follows here is a seemingly (and perhaps actually) random collection of comments and impressions from the rest of COFES.
Tech Soft 3D (www.techsoft3d.com) has hosted an annual customer meeting in advance of COFES’ formal start and Ron Fritz is kind enough to allow me (not a customer) to attend. Not only do the Tech Soft 3D guys share their insights into the CAD/CAE world but, for the last couple of years at least, they’ve lubricated the meeting with beer they brew themselves (a bourbon stout and a pale ale if memory serves me). Ah, if I only still imbibed.
The four main influences (driven by customers) on Tech Soft 3D’s products are:
- The Cloud, for scaling to large models or large numbers of models and for collaboration.
- New Device Landscape, aka mobile.
- Deployment and Delivery: People are expecting “web 2.0,” no installs, and frequent updates.
- User Experience (UX): Because everything is moving to a SaaS model, you’d better give your customers some great UX because SaaS makes it easy for them to take their business elsewhere.
Jon Peddie (www.jonpeddie.com) shared some of his firm’s research into the market’s usage of hardware which basically boils down to mobility, collaboration, and ubiquity (i.e. work anywhere, on any device, at any time).
- PC versus tablet is not an either-or proposition, except at the low end of the PC market.
- The PC is not going away. Tablets won’t start outperforming PCs unless software lets us down.
- Virtualization and remote graphics. The market for this was called “staggering” because people want the compute power but lack the IT smarts to do it themselves.
- Augmented and virtual reality.
- Big data has to be handled in the cloud.
Cambashi‘s (www.cambashi.com) Peter Thorne delved into market research on where people are spending money on technical software applications (i.e. CAE excluding EDA and software development tools).
- Year over year growth by market segment
- transportation = 8% (top)
- aero & defense = 6%
- worldwide avg = 4.9%
- AEC < 4%
- Forecast growth rates, 2014 to 2017.
- worldwide market goes from 4.9% to 5.8%
- Americas goes from 4.3% to 6.3%
- By 2017 U.S. is 28% of the world market, more than twice anyone else.
David Brin (davidbrin.com) – physicist, author, futurist – gave the event’s keynote speech. His scifi novel Existence was provided to all COFES attendees. His talk centered around the fact that projecting 5 years into the future is easy, projecting 200 years into the future is easy, but projecting 50 years into the future is really hard.
Onshape (www.onshape.com), and maybe Onshape in the person of CEO Jon Hirschtick, appeared to be the rockstar of COFES. When Jon talks, people listen. When he enters a room, people take notice. And when he and his company host a technology suite, people are crammed in cheek-to-jowl. Obviously smart and successful, he’s also leading what may be the hottest thing in CAD right now – a completely web-based CAD system.
If you have not gone to Onshape’s website and just watched the videos of what can be done in their tool you’re missing something very cool. For old farts like me, just the idea that CAD can be done in a browser over the internet is mind boggling.
Maybe you’re not impressed. But when I witnessed four people simultaneously working on the same model – including someone on an iPhone – and having all their changes simultaneously appear on each others’ screens… Well, let’s just say I was impressed.
Onshape was founded in 2012, now has more than 60 employees, just launched their product in beta, and has “thousands and thousands” of users. You can start using it for free but you’ll be limited to 5 private models. For $100 per month per user you can have unlimited private models. And for enterprise pricing, please give them a call.
At COFES, Onshape presented their API for the first time publicly in order to start building their ecosystem of partners to add functionality to Onshape’s core design tools. Anyone interested in a web-based CFD mesher?
Dennis Nagy from Beyond CAE (www.beyondcae.com) hosted a technology suite on the goal of getting simulation software into the hands of 7 million users. That’s about a 10x growth over current usership. The discussion revolved around the premises that the obstacles to this growth are:
- general purpose CAE software is too expensive
- general purpose CAE software is too hard to learn
- the HPC hardware required for CAE is too expensive to acquire and maintain (see Jon Peddie’s points above)
Siemens PLM Software‘s (www.plm.automation.siemens.com) George Allen gave a very interesting talk on the future of geometry modeling that began by saying that very little has changed in geometry modeling since 1973 when the B-Rep solid model was introduced. Yet 3D scanners, 3D printing, and the advent of GPU computing points us in possible new and radical directions.
- 3D scanners: Using points and meshes as geometry would be evolutionary but voxels would be revolutionary.
- 3D printing: Tackling the issue of embedded materials and their complications would be evolutionary but the ability to handle lattice structures would be revolutionary.
- GPU: Multi-core computing, while brute-force, is evolutionary but algorithms tailored to something between linear facets and NURBS would be revolutionary.
The conversation got a little derailed when George (I think it was him) said these new ideas are all well and good but Siemens PLM will never pursue them because of the need to continue to support their legacy application for current customers. The way to move forward is for a small, nimble organization to make the revolution and then Siemens PLM or someone else will acquire them.
The collective wisdom of the group in attendance was virtually unanimous on one thing: geometry must have a single representation from which all other geometric forms are derived.
Sandia National Lab‘s (www.sandia.gov) Ted Blacker hosted a tech suite on analysis credibility or the lack thereof and proceeded to share with those in attendance an overview of the predictive capability maturity model (PCMM) they apply to their simulations. When your simulations involve nuclear weapons the need to get things right is “important” and their use of the PCMM demonstrates that. PCMM is built-in to a lot of their simulation tools to give simulation users an idea of how credible the answers may be.
- Code verification
- Geometry and mesh fidelity
- testing that mesh points are on the geometry
- seeing how well mesh normal vectors match the underlying geometry’s
- ensuring the correct geometry was meshed
- and of course, mesh cell metrics
- Material and physical model fidelity
- Solution verification including mesh convergence
- Solution validation including comparison with test data
- Uncertainty quantification and sensitivity analysis
Mark Anderson, founder and CEO of Strategic News Service (www.stratnews.com), gave a keynote talk on protecting intellectual property. Mark is apparently the only person known to have predicted both the global financial collapse of 2007-2009 and the more recent drop in oil prices.
The gist of what he said was this. China’s national business model is to steal intellectual property, re-implement in their state-sponsored industries, and sell it back to the world at a vast discount and in the process put the original inventors out of business. It’s not a matter of whether you’ve been hacked by the Chinese; it’s just whether you know about it or not. And when traveling to China he recommends taking a burner laptop and phone. (Note: He is not anti-Chinese. He’s just against their business model.)
Bo Burlingham (boburlingham.com) also gave a keynote presentation based on his book Finish Big: How Great Entrepreneurs Exit Their Companies on Top. The factors that define a good exit are:
- the owners feel treated fairly and adequately compensated
- the owners are left with a feeling of accomplishment
- the owners are at peace with the fate of their (former) employees
- the owners have a new sense of purpose beyond the business
- the companies continue their success with their new owners
Maybe Bo’s most important bit of advice was this: build a company you’ll own forever but could sell tomorrow.
COFES 2016 (cofes.com) will be held on 7-10 April. Between now and then they left us with many things to consider, one of which was this: In 5 years the number of simulation users will explode by a factor of 20. What are you going to do about it?
This year’s offsite event was a visit to the Commemorative Air Force’s Airbase Arizona (www.azcaf.org), home of dozens of beautiful vintage aircraft including the B-17G in the photo below.
Want to read more about COFES 2015?
- Beyond PLM: Product Lifecycle, Supply Chain, and Data Networks
- Waveform: Revenge of the Thought Leaders
- Onshape: Standing Room Only for Onshape at COFES
P.S. After proof-reading this post a couple of times it did turn out to be a random collection of thoughts, notes, and impressions. It probably makes little sense to you. But sometimes I feel better after getting thoughts out of my head and off my papers and onto the computer. Even if (or maybe because) it’s 6 weeks after the fact.