Flutterby™! : Reading the road

Next unread comment / Catchup all unread comments User Account Info | Logout | XML/Pilot/etc versions | Long version (with comments) | Weblog archives | Site Map | | Browse Topics

Reading the road

2011-05-27 18:09:38.279529+00 by Dan Lyke 5 comments

The Netherlands recently hosted something called the Grand Cooperative Driving Challenge, in which a bunch of vehicles and some specially equipped infrastructure engaged in "platooning" behavior, closer following and presumably some smarter cruise control behavior.

These demonstrations are similar to ideas that have been floating around since the 1930s, and as I've mentioned previously they're kinda silly because they ignore two three things:

  1. Equipping a secondary set of infrastructure, traffic signs and signals that broadcast their state, roads that have some embedded track in them, is a ridiculous premise: You're going to have additional cost, fragile equipment, low uptake of machinery that can take advantage of it, and we've got problems with potholes, is anyone going to notice that the batteries in the smart stop sign that's never seen a vehicle that can communicate with it have gone dead?
  2. Deer, pedestrians, and the mattress or tire that just blew off the truck in front of you don't have transmitters.
  3. Adaptive Cruise Control is already being deployed without cooperative communication between vehicles, it's been estimated that just 10-20% of vehicles in heavy traffic using such a system can eliminate the stop-and-go traffic jams.

Obviously at this point you know that cars are starting to understand traffic signs, as in the new Ford Focus, but sometimes it's cool to see what's happening behind the scenes, as in this video of a system reading lanes and road markings.

Bonus: Infiniti's Blind-Spot Intervention System is tied into the electronic stability control to apply brakes on the opposite side of the vehicle if it detects you steering into an obstacle in your blind spot.

[ related topics: Embedded Devices Movies Robotics Interactive Drama Video Invention and Design Boats Handicaps & Disabilities Automobiles Machinery ]

comments in descending chronological order (reverse):

#Comment Re: made: 2011-05-30 23:01:04.908801+00 by: Dan Lyke

Interesting: So what's driving the vehicle-to-vehicle communications stuff? I know about IntelliDrive here in the U.S., and I'm pretty sure that pilot program is doomed, but aside from some high spending research programs that ignore basic human factors and risk assessment stuff, like IntelliDrive, I don't see any real driver to put vehicle-to-vehicle communication in the cars.

At least not now before we've figured out how to do anything useful with it.

So what are the non-safety apps?

#Comment Re: made: 2011-05-30 20:53:50.133037+00 by: TommyTom

Well the wifi chip is going to be there whether we like it or not, it already is in many of the upcoming models we work on for various non safety functions (much like the radar was first used for comfort ACC but now is used for auto brake). Sending out a few safety related bits won't cost anything extra.

#Comment Re: made: 2011-05-28 22:14:07.775149+00 by: Dan Lyke

Eric: Grin, I hadn't noticed.

TommyTom: Sample quantities of a 6 channel radar chip for automotive applications like ACC and crash mitigation these days run less than $15 each. Especially given that cars are going to need all of those sensors anyway (crash mitigation, ACC, etc), the additional cost of a WiFi chip, and complexity of intervehicle communication isn't worth it. The L.C. Davis papers seem to show pretty strongly that there's a good case to be made for widespread ACC, but that only works as a cooperative mechanism if you have a hugely higher density than if you have independent ACC.

The cooperation without the other technologies buys nothing, putting those energies into better autonomous capabilities is far more productive than putting them into communication based systems.

#Comment Re: made: 2011-05-28 21:11:41.156132+00 by: TommyTom

  1. The GCDC vehicles did not need any infrastructure support. The infrastructure was used to judge the vehicles in the race using "independent" equipment. The road was just plain ol' asphalt.
  2. Most GCDC vehicles used additional sensors such as laser/radar/cameras which are standard equipment in many vehicles, for example in the competing Volvo S60's. Non-cooperative objects would be detected easily.
  3. ACC is not good enough to drive at the headway times used in the GCDC, delay in detecting the vehicle just ahead in conjunction with braking system delays would lead to string unstable behavior (i.e. traffic "shockwaves").

In fact, if only a small percentage of vehicles were equipped with Cooperative Adaptive Cruise Control (CACC) you would still see improved traffic flow. CACC is way cheaper than a radar system, just an antenna and a cheap WiFi-chip, so it has potential for wider roll-out than traditional radar-based ACC.

#Comment Re: made: 2011-05-28 01:25:11.912652+00 by: ebradway

Are they simulating this in the Mojave Desert? Check the outdoor temp on the dash in the first Ford Focus video.