The Driverless world is getting closer to where I have been for over three years-But they are still not there
The article — In ‘Race to Autonomy,’ It’s Time to Earn Trust
https://www.eetimes.com/in-race-to-autonomy-its-time-to-earn-trust/#
Incremental epiphany by incremental epiphany, the driverless world is getting to where I have been for over 3 years. After 4 years many, including the press, are “shocked” to learn AV makers use human Guinea pigs. The problem is the industry is still not where it needs to be. Both because of a continued lack of understanding of what the best simulation technology is, and how it can solve the development time, cost and safety issues, and because most people simply go where the wind pushes them versus leading. (One still wonders if the first child or family needs to die as unnecessary test subjects for the final epiphany to occur. While Tesla is the most egregious, for many reasons, virtually every AV maker uses public human Guinea pigs needlessly and will, by the current design approach, have to injure and fatally injure many to get to L4.)
A clear example of someone who goes where the wind blows, hedges and fails to lead is Phil Koopman from Edge Case Research. He still supports “safety driving”. However, he now mentions it can be dangerous and then diverts attention from that to a safety case discussion which in no way addresses it.
On a positive note, Michelle Avary stated — “Speed to the market is great for innovation, but it poses problems for civil society. When AV companies run field trials on public roads, for example, they are “effectively turning your citizens into test subjects,” Avary noted.” The problem here is “speed to market” assume engineering due diligence. Which this industry is nowhere close to doing.
Junko Yoshida, the author of the article, writes — “So far, while testing their vehicles on public roads, AV companies have kept disclosure of their cumulative data as opaque as possible. They have relied on such measures as miles driven and disengagements (when safety drivers had to intervene) to claim the safety of their self-driving vehicles. Painting a positive picture for investors is imperative for covering the enormous AV development costs. Sharing the minimal amount of information might be a justifiable approach during technology research and development phases, but it is indefensible when it comes time to sell AVs to consumers. The endgame is safety, and that means earning the confidence of policy makers, regulators, and consumers. That requires transparency and collaboration among rival AV companies. In a highly competitive development environment, many automakers and tech developers have shown little appetite for this sort of harmony thus far.” From a safety and trust point of view she is right on point. However, the AV makers do not provide all disengagement data. And not providing data in the development phase might be fine if the development environment was not the public domain, using human Guinea pigs. Or involve hyping capabilities and saying one is L4 when they are not.
Beyond this I am in support of pulling together all the disparate and competing “standards” bodies and alliances. But again, the problem is I know of none of them who supports the right solution. Or even comes close to understanding the root cause. Catching on to the use of human Guinea pigs but still supporting it or thinking it is necessary and tenable is not enough. (The UN and UK got closer when they imposed a 10 second handover requirement for Automatic Lane Keeping System (ALKS). The irony here is that requirement not only makes my point but makes it so the mode will rarely be used. See more below.) The problem is the reliance on public shadow and safety driving, supported by gaming-based simulation technology. The solution is to flip the paradigm and use DoD/aerospace level simulation, informed, and validated by the real-world.
Finally, I want to plug the USDOT VOICES group, they get it. They understand the problem and solution. Hopefully, Secretary Buttigieg, and the world actually, will acquire the same understanding and push them to the forefront. (They specific the right simulation needs to be developed. That was written before they became aware of the technology I mentioned.)
More here
The Autonomous Vehicle Industry can be Saved by doing the Opposite of what is being done now
SAE Autonomous Vehicle Engineering Magazine — Simulation’s Next Generation (featuring Dactle)
· https://www.sae.org/news/2020/08/new-gen-av-simulation
UK Government is Ending (and Saving) the Autonomous Vehicle Industry as We Know It
Simulation can create a Complete Digital Twin of the Real World if DoD/Aerospace Technology is used
Simulation Photorealism is almost Irrelevant for Autonomous Vehicle Development and Testing
The Crash of the Autonomous Vehicle Industry
· https://medium.com/predict/the-crash-of-the-autonomous-vehicle-industry-f71fd26c1ed0
Proposal for Successfully Creating an Autonomous Ground or Air Vehicle
Autonomous Vehicles Need to Have Accidents to Develop this Technology
Using the Real World is better than Proper Simulation for AV Development — NONSENSE
The Hype of Geofencing for Autonomous Vehicles
My name is Michael DeKort — I am a former system engineer, engineering and program manager for Lockheed Martin. I worked in aircraft simulation, the software engineering manager for all of NORAD, the Aegis Weapon System, and on C4ISR for DHS.
Key Industry Participation
- Founder SAE On-Road Autonomous Driving Simulation Task Force
- Member SAE ORAD Verification and Validation Task Force
- Stakeholder for UL4600 — Creating AV Safety Guidelines
- Member of the IEEE Artificial Intelligence & Autonomous Systems Policy Committee (AI&ASPC)
- Presented the IEEE Barus Ethics Award for Post 9/11 Efforts
My company is Dactle
We are building an aerospace/DoD/FAA level D, full L4/5 simulation-based testing and AI system with an end-state scenario matrix to address several of the critical issues in the AV/OEM industry I mentioned in my articles below. This includes replacing 99.9% of public shadow and safety driving. As well as dealing with significant real-time, model fidelity and loading/scaling issues caused by using gaming engines and other architectures. (Issues Unity will confirm. We are now working together. We are also working with UAV companies). If not remedied these issues will lead to false confidence and performance differences between what the Plan believes will happen and what actually happens. If someone would like to see a demo or discuss this further please let me know.
