Wrap-up by Alain L. Kornhauser, Princeton University…

  1. autonomousVehicles have a public perception problem that is rooted in poor use of language
  2. Problem: SundaySupplementsimplicity v. the devil that is in the details
  3. Everybody is confused about what we are talking about.
  4. That’s why I call my newsletter SmartDrivingCars
  5. It is too geeky to talk about the NHTSA or SAE Levels 1,2,3,4 5?
  6. Richard Bishop started to improve things by suggesting that we separate AutomatedCollision Avoidance and Lane Keeping(ACA&LK) from Automated Vehicles (AVs)
  7. although that removes most of the financial & societal benefit from AV and gives it to ACA&LK.
  8. I’d go farther and separate Automated Vehicles into Self-driving and Driverless (Although we need a different name for Self-driving ) There is enormous difference between ACA&LK, Self-Driving (when there is still a driver to pitch in once in a while) and Driverless (where the vehicles can reposition themselves without anyone in them, which is absolutely necessary for autonomousTaxi operation.
  9. Maybe we need to place each of these in their individual modes and let them compete. They each have very different value propositions, technological elements and policy implications.
  10. ACA&LK have fundamental Safety improvements,
  11. Self-driving delivers substantially reduced travel disutility, but substantially increase VMT, and
  12. Driverless promises inexpensive on-demand mobility for all as a share-ride offering that substantially increases PMT while reducing VMT, energy consumption, GHG, other pollution and congestion.
  13. Jim Anderson made a very important comment: We need to raise the playing field of existing transportation. If the sun is behind the traffic light and I can’t see it we need to make the traffic light “seeable” to the conventional driver, then the vision sensors will also be able to see it. (not spend money on fixes that aren’t first-&-foremost useful to conventional drivers. Also our current speed-limit“legislation” is a complete joke. We should have real speed limits and other traffic control “laws” that both humans and algorithms can readily obey.
  14. Promotion by public entities..there is none. This workshop is co-sponsored by I-95 and cruise control has fundamental advantages (energy and congestion), yet I am not aware that there is one sign anywhere along I-95 that promotes or has ever promoted the use of cruise-control by drivers. If agencies haven’t even gotten to the point of promoting the use of cruise control, I wonder if they have the stomach for promoting any of this technology.
  15. Safety… Consumers haven’t bought it. Insurance pays 50% of the tab, but hasn’t promoted it. Regulators just talk about it. We need a disrupter here (probably by insurance, because they are holding the money and one of them may actually see that they can get rich by promoting the technology.
  16. AVO is really 1.0if not lower, today…if AVO is defined as Person-Mobility-miles/Vehicle-miles. The real opportunity with Driverless is to substantially increase AVO to close to 2.0 overall and over 3.0 in peak hour peak directions. That would chop Energy in half, GHG in half, and eliminate most of congestion, while providing safer on-demand mobility for all. Wow!!