Happy 2018 — and, optimize THIS
I hope you had a wonderful year-end, filled with people you love. Or tolerate, if that’s how you roll. In any event, as peaceful and relaxing as it could be.
I’ve been trying to write a 2017 retrospective and, honestly, the mind boggles. So much happened, on so many fronts, that it’s been impossible to coalesce it into themes. Instead, I’d like to look ahead to 2018 because I think it’s going to be awesome. Here are a couple of reasons:
- We’re going to be optimizing EVERYTHING. Soon. And as a fundamental part of the design process, not as an afterthought. Examples are everywhere. At Autodesk University, CTO Jeff Kowalski spoke about how Autodesk renovated its Toronto office using space optimization (based on geometry: these coffee makers must be this close to these desks) and also on other factors that employees deemed important, that are not based on typical geometric constraints. According to Mr. Kowalski, at any rate, the result was an office layout that hit a lot of quantifiable targets that were much broader than physical constraints. And at Siemens R&D Center in Munich, they’re teaching robots to create customized and optimized assembly instructions so that humans aren’t necessarily in the loop. Show the robot the physical parts, show it the assembled object and let it try all of the possible combinations to arrive at the most efficient process. The implications are staggering: tell your optimization tool(s) what you want, what your constraints and objectives are and walk away. Come back to a huge set of possibilities that you narrow down and then apply your design aesthetic to. Downsides, of course, are that designers and engineers will start to lose some of the physical “sense” they build with experience and that we may become, ultimately, obsolete. That latter won’t happen but the former is a real concern.
- Cool toys are becoming the norm. From augmented and virtual reality (AR and VR) applications to sensor-everything, the way we interact with our objects and technology is going to change. At Siemens R&D we were treated to several examples of headsets guiding technicians and others through operational scenarios. In one case, a complex maintenance situation combined sensor data as input, analysis via Simcenter’s CFD (CD-adapco) engine to outline possible failure scenarios and then tied to Teamcenter to suggest maintenance options. The guy/gal wearing the headset wouldn’t need to know about Simcenter, CFD or Teamcenter; just red is bad and follow the maintenance instructions. It sounds easy but there’s a lot of work to do to bring this to life, not the least of which is creating Edge compute versions of CFD and other codes to make this truly real-time — but progress is being made and this sort of solution should hit the market soon-ish.
- One thing we learned (again) over the break: Technology carries risk. Meltdown and Spectre, the computer chip security flaws that are said to affect most of the world’s processors, are just the latest bad news for anyone doing anything digitally — but the good guys are out there, working hard to keep a step ahead of the bad guys. At Siemens, I had the opportunity to be a black-hat hacker, trying to break into a simulated hospital’s database. I was snooping, with no real aim (which I think is how casual hackers operate; if they get in, they sell that on to someone with a serious goal), and learned how to get around the primitive security mechanisms in this demonstrator. Siemens has, over the decades, put millions of devices into the field, all of which need to be safeguarded from known and expected vulnerabilities. It also has to safeguard its own assets. As a result, the company has a huge effort to teach everyone about these vulnerabilities and has a large and impressive team in place, working with lots of other companies, governments, universities and industry bodies to stay ahead of it all. Their main pieces of advice: create strong passwords and change them often, update to the latest operating system/software version, and be aware of what’s connected to what so that you can isolate something quickly if you need to.
- Data is king, queen and overlord. October, November and December were really busy for Schnitger Corp, as we attended conferences on three continents and spoke with dozens (maybe hundreds?) of companies about what matters to them. Uniformly, there’s regret about data not kept. Storage is cheap, so keep the last/latest version of everything (securely, password protected, offsite, etc.) It costs a lot to recreate the digital version of a design when compared to archiving it until needed, so it seems foolish to hit that delete button. But, of course,
- Data must flow smoothly from where it is to where it needs to be. That’s the sticking point for most legacy data, especially when we’re talking older operating systems or changing CAD (or other) file formats. With the focus on downstream uses of data for AR, VR, operations training and other uses, it’s time for organizations to develop a data plan: what is it, where did it come from, what do we want to do with it next (and what might we want to do in 5 years)? Factor that into decisions on CAD or other tool usage (especially if you’re torqued about license model changes) — it’s not just creating the design in the first place; it’s also what happens in the 5th, 6th and 10th use iterations. That, of course, brings me to hopes for
- A calmer 2018 in the PLMish universe. 2017 saw a lot of turmoil in our little part of the tech industry, as the fallout from perpetual to subscriptions continued and as Autodesk, PTC and others continued their pivots into new areas of business — and that meant that some people we all know and respect were made redundant. But our industry has a history of fostering cool new enterprises from forced spin-offs so expect many of these people to appear again, in a new and exciting venture.
Wishing you and yours an awesome 2018!