PAVE Panel on ODD – The Question is Where not When!

PAVE Panel on ODD – The Question is Where not When!

Carter Stern of Cruise and I had a great discussion on Operational Design Domains (ODDs) for automated vehicles at the Partners for Automated Vehicle Education (PAVE) panel.

Our overarching theme was “Where, Not When: Why Domains Matter” and it was nice that Carter and I could build off of one another’s perspectives from the tech and auto industries.

Here’s one of my key points:
An Operational Design Domain can be anything. It can be what is sometimes called the ‘Four Ws’ – wherever, whenever, whatever the weather. And that is what is called ‘level 5’ or all ODDs you can imagine. But level 5 driving is similar to infinity in math: it is important to prove your concept, but you never get there.”

Tabitha Colter did a fantastic job moderating the discussion and putting thoughtful questions in front of us. Thanks very much!

Sensors for automated vehicles – contribution to EU panel discussion

Sensors for automated vehicles – contribution to EU panel discussion

Key Vehicle Technologies for Cooperative Connected Automated Mobility (CCAM): Is the level of maturity sufficient for high-level automation?

Robust and accurate environment perception is absolutely essential for highly automated vehicles to enable the safe and unambiguous extraction of reliable information for real-time driving decision-making. Vehicle technologies need to reliably identify, track and discriminate between benign and hazardous objects in the path of the vehicle, under the full range of environmental conditions in which the vehicle is intended to operate. This session addresses the different challenges which must be overcome through advances in vehicle technologies in order to enable CCAM with higher levels of automation to become reality in the near future.

My contribution to the panel hosted by the European Commission discussed the basics of automated vehicle sensors (camera, radar, LiDAR), sensor fusion, and the market outlook.

One topic that I emphasized in particular at the end is recalibration of those sensors once vehicles age and need repair. A recent study showed that just replacing sensors is not enough, they need to get recalibrated for e.g. automated emergency braking (AEB) to work properly. However, those procedures might be costly and not every repair shop might be able to perform those; so that sensors might just get replaced but not recalibrated – and the ADAS won’t work or actually cause additional risk.

Vehicle Electrification Panel Discussion

Vehicle Electrification Panel Discussion

Our Vehicle Electrification panel discussion with Gary SilbergDaron Gifford, and Christopher Thomas hosted by Association For Corporate Growth – ACG Silicon Valley went quite well. We discussed consumer, business, regulatory, technology, infrastructure aspects.

Ok, we could not answer all open questions regarding PHEV, BEV, FCV. But I like that we looked at “both sides of the coin”; actually one needs to peel the onion a bit to figure out what is “the best” solution – if that even exists…

And Gary did a great intro summarizing his recent research Place your billion-dollar bets wisely: Powertrain strategies for the post-ICE automotive industry

Guest on the Driving Green podcast

Guest on the Driving Green podcast

This was a great discussion with Ronald Mikhail AngSiy, Kyle Kennedy, and Kevin Wang on the future of autonomy and transportation. We covered quite a few topics at the following time marks:
05:30​ – History of Autonomy and ADAS
10:15​ – Autonomous Hardware Trends
18:30​ – City vs Suburban Environments
25:00​ – What is Mobility?
35:15​ – What will Autonomy Unlock?
44:00​ – City Design
49:00​ – Ride Sharing
56:25​ – Tesla’s “FSD” Approach
69:53​ – Carbon Footprint

“The Driverless New Normal” – A controversial and passionate debate

“The Driverless New Normal” – A controversial and passionate debate

Great debate between Alain Kornhauser and me whether autonomous vehicles have “the best times” or “the toughest times” ahead of them.

Thanks also to our great panelists Jane LappinBrad TempletonJim Scheinman, and Michael L. Sena who greatly added to the debate. Also thanks to Richard Mudge and Fred Fishkin for great moderation!

We set out with quite some controversial viewpoints but agreed that there are exciting times ahead and it takes much collaboration to make autonomous vehicles happen. We will see them first in very specific applications – maybe trucks or delivery robots – and from there they will grow into the open.

More to come!

Testing & Simulation of Autonomous Vehicles – Stanford Research Park Speaker Series

Testing & Simulation of Autonomous Vehicles – Stanford Research Park Speaker Series

Sven Beiker led a discussion about Testing & Simulation of Autonomous Vehicles – Balancing the real and virtual worlds of future mobility. This event was hosted by Stanford Research Park in collaboration with SAE International and hosted by Ford Greenfield Labs.

“How safe is safe enough?” – that is one of the most essential, most discussed, and still unanswered questions around autonomous vehicles. As the industry prepares for the deployment of self-driving cars, there appears to be only one way to ensure safety: to run over and over again through each and every possible situation those vehicles might encounter during operation. This is done in part through testing on public roads, but to maximize safety and accuracy, closed course testing is also very important. And still, as it is understood that hundreds of millions of miles would need to be driven in order to prove the safety of autonomous vehicles, simulating those situations in the virtual world has also become essential. This presentation and following panel discussion dove into those topics to discuss how the real and virtual worlds need to be balanced for safety, efficacy, and efficiency in testing and how important fidelity of simulation models is in order to draw conclusions from the virtual to the real world. The panelists are members of a core team that gave input to the two most recent SAE International EDGE Research Reports titled “Balancing Virtual, Closed- Course, and Public-Road Testing of Automated Driving Systems” and “Determining Appropriate Modeling Fidelity for Automated Driving Systems Simulation”.

Panelists:
Sven A. Beiker – Managing Director at Silicon Valley Mobility (moderator)
Alexander Kraus – Senior Vice President Automotive at TÜV SÜD
Georg List – Vice President Corporate Strategy at AVL
Robert Seidl – Managing Director at Motus Ventures
Thomas Bock – Director Vehicle Integration & Testing at Samsung Smart Machines

© 2017-2024 - Silicon Valley Mobility, LLC 1102 Clark Way, Palo Alto, CA 94304, USA