With my own design work the first over 300, 400 and 500 mph in dynamic soaring flight and experience building wind tunnel artifacts for advanced visualization and CFD exploration for Boeing and the UofW, I am called by Microsoft Research who; working with EPFL in Switzerland, take the first steps to bring OpenAI into multiple aspects of general aviation and advanced aerial robotics applied to multiple applications spanning design, analysis, intelligent autopilots and much more, all under Microsoft’s Project AirSim, where ChatGPT driven programming of drones is already well established.

One of the first project activities advancing from quadcopter drones is to get a real handle on CFD virtualization.  A lot is out there already, but all leading academics and industrialists will admit the road is long still.  One breakthrough is an idea where physical -and- virtual models are compared in physical -and- virtual wind tunnels with a goal to come to an aggregate ‘truth’ such that intelligent design and analysis tools can be formed up and applied to nearly any aero/hydro application.

Since both teams started with dynamic soaring as their test case, I was contracted to build both physical and virtual wind tunnel models based on the best of the Kinetic 100v2 and the Kinetic 2 meter and FPS, modified to carry instrumentation and eventually a dynamic soaring autopilot.  Being the first person I know ever to be tasked this way by extremely high level data scientists with little aero engineering experience, the lifecycle was based on software development, not aero engineering.  Pressure was extremely high, since every single line item had to be completed error free the first time.  Not only that, both models had to be of the highest possible quality and fidelity, as it was/is intended to build intelligence on both artifacts.

Working with Benoit Valentin Gherardi in Switzerland and observed, mentored and critiqued by senior distinguished technical fellow Dr. Philippe Spalart, I was referred to as ‘The Bedrock’ by the teams, and ‘Dr. Steven’ on occasion by Philippe.  I did 100% of everything virtual and physical alone and on my own, and both teams ended up with ultra precision models and so good data.  Dr. Spalart plotted the pressure tap locations, which were read for 8 seconds per experiment, and authored the final result.  Both models are very exotic, specifically in the design of the wing.  A full discussion is presented on my Projects page.  A couple of videos are posted to my links page.