Realities of traffic make automated driving a challenge

Realities of traffic make automated driving a challenge

Nice interview here with SAE International for their monthly magazine “Update“. My piece is titled “Realities of traffic make automated driving a challenge”.

It is a conversation around what is missing to bring autonomous vehicles (automated driving to use the right SAE term) to the real world. I am glad that SAE asked as I have been discussing a lot over the years what is holding us back: technology, regulation, trust…? My view is that we have solutions for all of those, as long as we choose the ODD wisely. However, real-world traffic still poses the biggest challenge.

Along those lines, I created the following chart at some point. On the right you see the graphic of climbing a mountain, which I typically use as a prompt in my talks to say the following.

– If you want to get on top of the world, Mount Everest, you might get on a plane in San Francisco, make a connection in Heathrow, and get to Kathmandu.
– From the airport you travel to the basecamp at 5,300m altitude of Mount Everest
– At that point you have traveled well over 99% or so of the distance, but this is where it begins to get interesting…

Report on Visual Communication Between Automated Vehicles and Other Road Users

Report on Visual Communication Between Automated Vehicles and Other Road Users

I am glad and proud to have another SAE EDGE Research Report published. This one discusses “Visual Communication Between Automated Vehicles and Other Road Users”.

As automated road vehicles begin their deployment into public traffic, they will need to interact with human driven vehicles, pedestrians, bicyclists, etc. This requires some form of communication between those automated vehicles (AVs) and other road users. Some of these communication modes (e.g., auditory, motion) were previously addressed in “Unsettled Issues Regarding Communication of Automated Vehicles with Other Road Users.”

We had the general overview already out earlier in the year, and now this one focuses on visual communication and its balance of reach, clarity, and intuitiveness, and discusses how different visual modes (e.g., simple lights, rich text) can be used between AVs and other road users. A particular emphasis is put on standardization to highlight how uniformity and mass adoption increase communication efficacy.

Unsettled Issues in Determining Appropriate Modeling Fidelity for Automated Driving Systems Simulation

Unsettled Issues in Determining Appropriate Modeling Fidelity for Automated Driving Systems Simulation

Here’s the other SAE International EDGE Research Report that we published already some months ago: “Unsettled Issues in Determining Appropriate Modeling Fidelity for Automated Driving Systems Simulation”

It discusses the challenges of achieving optimal model fidelity for developing, validating, and verifying automated vehicles. The primary questions raised are:

  1. How to make sure that simulation models represent their real-world counterparts
  2. How to define a universal simulation model interface
  3. How to determine the different requirements for sensor, vehicle, environment, and human driver models

Thanks to the team for your great contributions!

https://www.sae.org/publications/technical-papers/content/epr2019007/


Click image for more info

Unsettled Issues in Balancing Virtual, Closed-Course, and Public-Road Testing of Automated Driving Systems

Unsettled Issues in Balancing Virtual, Closed-Course, and Public-Road Testing of Automated Driving Systems

Great to have two new SAE EDGE Research Reports out there! The first report discusses three main issues:

  1. Determining what kind of testing an ADS needs before it is ready to go on the road.
  2. The current, optimal, and realistic balance of simulation testing and real-world testing.
  3. The challenges of sharing data in the industry.

We have been discussing a lot how to develop > test > certify autonomous vehicles and always get to the point of simulation vs. road testing. So we went deeper into this to at least agree on the “right” questions.

Thanks to the team for a great collaboration!

https://saemobilus.sae.org/content/epr2019011


Click image for more info

© 2017-2021 - Silicon Valley Mobility, LLC 1102 Clark Way, Palo Alto, CA 94304, USA