Autonomous Vehicles, LEAF and others...

[GRA]

Via GCC:

IIHS study finds GM front crash prevention systems cut police-reported crashes

https://www.greencarcongress.com/2018/1 ... -iihs.html

An IIHS study of General Motors vehicles with optional front crash prevention systems found that GM vehicles with autobrake and forward collision warning had 43% fewer police-reported front-to-rear crashes of all severities and 64% fewer front-to-rear crashes with injuries than the same vehicles without any front crash prevention technology.

For vehicles equipped with forward collision warning only, the crash rate reductions were 17% for all front-to-rear crashes and 30% for front-to-rear crashes with injuries.

The results echo an earlier IIHS study involving Acura, Fiat Chrysler, Honda, Mercedes-Benz, Subaru and Volvo vehicles, which found that the combination of forward collision warning and autobrake reduced front-to-rear crash rates by 50% for crashes of all severities and 56% for front-to-rear crashes with injuries. Forward collision warning without autobrake cut the rates 27% and 20%, respectively, for vehicles in that study.

[list]The evidence has been mounting that front crash prevention works, and it works even better when it doesn’t solely rely on a response from the driver.

—Jessica Cicchino, IIHS vice president for research and author of both studies[/list]

The new research involves 2013-15 Buick, Cadillac, Chevrolet and GMC brands. GM provided vehicle identification numbers (VINs) for vehicles with and without front crash and other crash avoidance systems.

Now, if we could just get Tesla to provide similar data to back up their Autopilot claims (as opposed to AEB, which we know works for front-to-rear crashes).

[mwalsh]

Saw an Uber autonomous vehicle in Anaheim on Sunday. First time I'd seen one of any stripe.

[GRA]

Via GCC:

Volvo Trucks provides autonomous transport solution to Brønnøy Kalk AS

https://www.greencarcongress.com/2018/1 ... rucks.html

Volvo Trucks has signed a landmark agreement with Brønnøy Kalk AS in Norway to provide its first commercial autonomous solution transporting limestone from an open pit mine to a nearby port.

The solution for Brønnøy Kalk AS consists of limestone being transported by six autonomous Volvo FH trucks on a five kilometer stretch through tunnels between the mine itself and the crusher. Tests of this solution have been carried out successfully and will continue throughout 2018 to become fully operational by the end of 2019.

The agreement follows recent successful automation projects involving mining, sugar cane harvesting and refuse collection. Yet this commercial solution represents an exciting first for Volvo Trucks. Rather than purchasing autonomous trucks, Brønnøy Kalk is buying a transport solution—specifically the transport of the limestone between the two hubs. . . .

The agreement involves a deal whereby the customer buys a total transport service and pays per tonne delivered.

I could almost put this in the 'Mobility as a Service' topic - 'Cargo Hauling as a Service'?

[cwerdna]

Waymo Set to Debut Autonomous Ride-Hailing Service to Select Arizona Users in December
https://www.macrumors.com/2018/11/13/wa ... -december/ points to

Waymo launches its first commercial self-driving car service
Waymo One's on-demand autonomous rides come with human backup for now.
https://www.engadget.com/2018/12/05/waymo-one-launches/

A ride in Waymo One as the self-driving service goes 'live'
Waymo's been promising it'd launch its fully autonomous service by the end of 2018, and with just a few weeks to go the company has met that goal. Kinda.
https://www.cnet.com/roadshow/news/waym ... r-project/

[GRA]

GCC:

Consumer Reports: Self-driving cars bill in Congress still falls short on safety

https://www.greencarcongress.com/2018/1 ... 05-cr.html

As car and tech companies press Congress to approve a bill to speed up the introduction of self-driving cars, Consumer Reports says the bill still falls short of ensuring that these vehicles will be safe for consumers.

Senators have shared draft revisions to the bill—known as the AV START Act—in a last-minute push to win approval as lawmakers prepare to adjourn for the year.

David Friedman, the vice president of advocacy for Consumer Reports and former interim head of the National Highway Traffic Safety Administration (NHTSA), said, “Companies are trying to rush this bill through Congress, but there is no clear need for speed. These changes still do not provide the safeguards we need to make sure self-driving cars will be safe, despite some improvements that have been made in the bill.”

Friedman said senators have added language to require companies to provide more data about the safety and testing of self-driving vehicles, and they have tightened some loopholes that would have allowed companies to shut off vehicle safety systems. However, the bill remains deeply flawed because it would permit companies to put thousands, and possibly millions, of vehicles on the road before they are demonstrated to be safe, he said.

For self-driving cars to succeed, Congress should take the time it needs to draft legislation that provides meaningful protections—even if that means continuing its work into the new year, he said. . . .

• Surveys show that most consumers are already skeptical. If this bill becomes law as written, it could actually set the technology back for years, because the lack of standards could lead to more injuries and deaths, and that would further erode public trust.
  
  —David Friedman

[GRA]

ABG:

Uber to fire up its autonomous testing program once again
Ride-hailing company plans cautious restart after tragedy earlier this year

https://www.autoblog.com/2018/12/07/ube ... ram-again/

Uber is coming back to autonomous vehicle testing eight months after one of its test vehicles hit and killed a pedestrian crossing the street in Arizona. The plan this time around is to be overly cautious. Vehicles will run on a one-mile loop between two company offices in Pittsburgh — they will only be allowed to operate in the daylight and in dry weather conditions. Speed will also be limited to 25 mph for every vehicle testing. There won't be any rides given to passengers in these test vehicles, either. Uber doesn't specify how many vehicles it will start with. . . .

Every new autonomous Uber will have two passengers, both sitting in the front seat from now on. This is a huge departure from its past policy of only requiring one operator in the driver's seat, even in high-risk situations like night driving in the city. The loop for these initial test cars is actually the exact same loop Uber started its program with back in 2016.

Gaining the public's trust back after the incident in March of this year won't be an easy task. Still, the company knows this technology may be necessary for its profitability. Eliminating the cost of the drivers is the biggest hurdle Uber faces as a company if it wants to stop having quarters like its most recent in which the company lost $1 billion. . . .

Finally adopting conservative practice, better late than never. RIP Elaine Herzberg - from the Wiki:

The National Transportation Safety Board (NTSB) sent a team of federal investigators to gather data from vehicle instruments, and to examine vehicle condition along with the actions taken by the safety driver.[21] Their preliminary findings were substantiated by many event data recorders and proved the vehicle was traveling 43 miles per hour (69 km/h) when Elaine was first detected 6 seconds (378 feet) before impact; it was unable to determine that emergency braking was needed another 4 more seconds.[9] A vehicle traveling 45 mph (72 km/h) can generally stop within 195 feet (59 m).[22] Because the machine needed to be 1.3 seconds (76 feet) away prior to discerning that emergency braking was required, whereas at least that much distance was required to stop, it was exceeding its assured clear distance ahead,[23] and hence driving too fast for the conditions.[24][25][26][27][9] A total stopping distance of 76 feet itself would imply a safe speed under 25 mph,[22] whereas human intervention was still legally required. Computer perception–reaction time would have been a speed limiting factor had the technology been superior to humans in ambiguous situations;[28] however, the nascent computerized braking technology was disabled the day of the crash, and the machine's apparent 4 second perception–reaction (alarm) time was instead an added delay to the still requisite 1–2 second human perception–reaction time. Video released by the police on March 21 showed the safety driver was not watching the road moments before the vehicle struck Herzberg.[29] Notwithstanding perception and reaction times, 43 mph also commands a braking distance slightly greater than 76 feet.[22].

https://en.wikipedia.org/wiki/Death_of_Elaine_Herzberg

[DanCar]

Waymo flops: https://www.youtube.com/watch?v=L_5xBf-TZ3I

[GRA]

Waymo flops: https://www.youtube.com/watch?v=L_5xBf-TZ3I

No flop, just a very conservative roll-out, which is what Uber should have been doing (and Tesla). Waymo (I think then still the 'Chauffeur' program as part of Google) had their first accident (a 2 mph non-injury fender-bender) after seven years of public driving; by comparison, Tesla experienced their first fatal A/P accident just 7 months after releasing it, because they were pushing the tech way beyond what it was ready for. I don't doubt that Waymo is being even more conservative with this roll-out than they otherwise would have been, given the negative public perceptions from the Uber fatality there, even assuming that AZ isn't requiring a safety driver for now.

If you want to see some of the differences between the various companies attitudes/approaches, I suggest you read "Autonomy: The Quest to Build the Driverless Car—And How It Will Reshape Our World" by Larry Burns, who has been involved in this stuff from back when he was VP of R&D and Planning for GM, then consulted for Google/Waymo, and knows all the major players.

https://www.amazon.com/Autonomy-Quest-D ... 0062661124

The book's so recent it even mentions the Herzberg (and Huang) crashes, although not in as much detail as the NTSB prelim, and it also dissects the Brown crash and why Google/Waymo ultimately decided not to go down the 'driver assist' route on the way to full autonomy.

BTW, that guy at Hyper Change needs to cut back on the caffeine!

[cwerdna]

Uber self-driving cars allowed to operate on Pittsburgh streets again
https://www.wpxi.com/news/top-stories/u ... /892115618

Uber Resumes Self-Driving Car Tests In SF Months After Fatal Accident
https://sanfrancisco.cbslocal.com/2018/ ... -accident/

[cwerdna]

Waymo self-driving vehicles face attacks in Arizona
Slashing tires, throwing rocks, brake-checking, and a game of chicken
https://www.autoblog.com/2019/01/02/way ... le-attacks