AUVSI Driverless Car Summit Part 2
The AUVSI summit on "driverless" cars last week contained 2 days of nothing but robocars, and I reported on issues regarding Google and policy in part 1.
As noted, NHTSA released their proposal for how they want to regulate such vehicles. In it, they defined levels 0 through 4. Level 2 is what I (and GM) have been calling "super cruise" -- a car which can do limited self driving but requires constant human supervision. Level 3 is a car which can drive without constant attention, but might need to call upon a human driver (non-urgently) to handle certain streets and situations. Level 4 is the fully automatic robocar.
Level 2 issues
Level 2 is coming this year in traffic jams in the Mercedes S and the BMW 5, and soon after from Audi and Volvo. GM had announced super cruise for the 2015 Cadillac line but has pulled back and delayed that to later in the decade. Nonetheless the presentation from GM's Jeremy Salinger brought home many of the issues with this level.
GM has done a number of user studies in their super cruise cars on the test track. And they learned that the test subjects very quickly did all sorts of dangerous things, definitely not paying attention to the road. They were not told what they couldn't do, but subjects immediately began texting, fiddling around in the back and even reading (!) while the experimenters looked on with a bit of fear. No big surprise, as people even text today without automatic steering, but the experimental results were still striking.
Because of that GM is planning what they call "countermeasures" to make sure this doesn't happen. They did not want to say what countermeasures they liked, but in the past, we have seen proposals such as:
- You must touch the wheel every few seconds or it disengages
- A camera looks at your eyes and head and alerts or disengages if you look away from the road for too long
- A task for your hands like touching a button every so often
The problem is these countermeasures can also get annoying, reducing the value of the system. It may be the lack of ability to design a good countermeasure is what has delayed GM's release of the product. There is a policy argument coming up about whether level 2 might be more dangerous than the harder levels 3 and above, because there is more to go wrong with the human driver and the switches between human and machine driving. (Level 4 has no such switches, level 3 has switches with lots of warning.)
On the plus side, studies on existing accidents show that accident-avoidance systems, even just forward collision avoidance, have an easy potential for huge benefits. Already we're seeing a 15% reduction in accidents in some studies just from FCA, but studies show that in 33% of accidents, the brakes were never applied at all, and only in just 1% of accidents were the brakes applied with full force! As such, systems which press the brakes and press them hard when they detect the imminent accident may not avoid the accident entirely, but they will highly reduce the severity of a lot of accidents.
Level 3 Issues
Christian Schumacher of Conti told instead of greater problems with level 3 and above, because in today's cars, the physical abilities of the human driver are the backup systems. For example, if your steering motor fails, your arms physically can turn the wheel. If your brake pumps fail, your foot can apply direct hydraulic pressure. In a full robocar, you can't count on a human as your backup system, and so cars need to be redesigned to have redundancies they did not have before. Contintental, as a leading supplier of such systems, wants people to be aware of this.
Nissan's Maarten Sierhuis indicated Nissan's vision is more of thinking of the self-driving system and the human driver as "team members" in the task of driving the car. They are not ready to go for the higher levels at first where the human goes out of the loop. Think of the car like a home or office or even a space suit, and design that way.
Volvo boldness, and convoy issues
Volvo was, to my surprise, most aggressive of all the car companies I have heard from of late. While almost every car company says things like Nissan did, that they are not trying to eliminate the driver, but rather make travel safer and more convenient with the human still playing a real part, Volvo is willing to say it's going for the more serious levels. (Of course Google and most academic teams also aim there.)
Volvo is doing this as part of the goal they set several years ago, that "nobody dies or is seriously injured in a new Volvo by 2020." Volvo says it is still on track for this, ahead of competitors, but knows that full robocars are needed to make the goal.
Volvo also presented, along with Ricardo, on the Sartre project, which I've covered here before. They showed a new negative result on convoying. This amazing image, from Volvo, shows the radiator of one of the following cars in their Convoy tests. It was so destroyed by thrown stones that it started leaking. The windshield and other elements of the car were also heavily pitted.
It seems that super-close convoying will require vehicles designed to handle it. They also reported difficulty in coming up with a business model, due to the chicken and egg problem. Fuel savings are 15 to 20%, which is worth it, but not the overwhelming value.
Volvo also reported more and more studies that show the public is very ready for this technology. Those who believe that the public simply won't get in these cars appear to be wrong. A Canadian study cited showed 1 in 7 ready to buy one today, and only 1 in 7 on the other side, who think they would never ride one.
The FHWA, which manages infrastructure and not cars, is still getting excited, because they now are realizing that these cars will let them get a lot more out of their infrastructure for the same money.
On the name: Once again, the crowd was not a big fan of "driverless car" as a name, though that name is gaining dominance in the press. Once again, there was no clear other favourite.
The SAE has its own working group on the tech now, which defined 5 levels to NHTSA's 4 -- the extra distinction being handling all roads and only a subset in full-auto mode.
Valeo (who I reported earlier said Google was the enemy) also talked about and demonstrated their various self-parking systems, including a live demo of unmanned cars parking in the parking lot behind the stage.
They are also producing a new low-cost 4-beam LIDAR. This LIDAR was designed by IBEO, and it was reported in the press after CES that Audi had made it. With its low cost, it may change people's minds about the cost of LIDAR, though with only 4 planes it is mainly good for ADAS and parking, and perhaps level 2, at this time.
According to Valeo, if the initial efforts at park assist were level one of effort, their park4U system (where you don't have to help) was level 10, and the remote system which works unmanned is level 100. Doing full remote valet parking would be level 1,000 of effort.
Jan Becker of Bosch gave the most technical of the presentations and this was very welcome. They showed off Bosch's Velodyne based experimental car. They reported that Bosch's version of Level 2 will be ready to deploy by 2016, and speculated on 2020 for their level 3.
There was a presentation by NAVIA, whom I've written about before. They showed some videos I have been meaning to blog. The NAVIA videos can be seen here. What's remarkable here is their fully autonomous vehicles with no steering wheels are doing demos with civilians on crowded pedestrian-only streets, even running vacant at times. They only go 12mph and a staffer was present with a kill switch, but even so it's impressive to see cities willing to do such demonstrations. NAVIA already has a number of campus contracts and plans to expand into cities soon.
There was a lot of talk about trucking. Generally, I have felt trucking was not a great first area to explore, in spite of the obvious commercial market. Nobody building robocars is looking to take away jobs, so why go into that area initially?
However, truck owners are now very keen because they can't hire enough drivers. There are 8.8 million trucks in the USA and only 7 million drivers with commercial licences to drive them.