Thursday, June 18, 2020
26th edition of the 8th year of SmartDrivingCars
U.S. Transportation Secretary Elaine L. Chao Announces First Participants in New Automated Vehicle Initiative to Improve Safety, Testing, and Public Engagement
Press release, June 15, “The U.S. Department of Transportation today announced nine companies and eight States that have signed on as the first participants in a new Department initiative to improve the safety and testing transparency of automated driving systems, the Automated Vehicle Transparency and Engagement for Safe Testing (AV TEST) Initiative. The participating companies are Beep, Cruise, Fiat Chrysler Automobiles, Local Motors, Navya, Nuro, Toyota, Uber, and Waymo. The States are California, Florida, Maryland, Michigan, Ohio, Pennsylvania, Texas, and Utah.
“Through this initiative, the Department is creating a formal platform for Federal, State, and local government to coordinate and share information in a standard way,” said U.S. Transportation Secretary Elaine L. Chao. …
This initiative aligns with the Department’s leadership on automated driving system vehicles, including AV 4.0: Ensuring American Leadership in Automated Vehicle Technologies.” Read more Hmmm… Excellent. This is really good because it is promotes and organizes the open sharing of safety information assoiated with automated driving. This is extremely important because safety of these systems is a necessary condition for their adoption.
Unfortunately, a few things seem to be missing from the announcement.
- a budget (Washington may not have any money left after COVID-19),
- any mention of mobility for people or for goods. The testing of safety is conducted without doing any useful mobility, but the value of testing is derived from the delivery of that mobility. Safety in that context requires the active engagement the entities that are being transported. It is very important that this initiative include potential customers and neighborhoods whose streets such automation might use, and
- the whole Northeast including New Jersey seems to have not “signed on”, nor is Ford/Argo, Zooks, Aurore, Voyage Amazon, Apple…. 🙁 Alain
SmartDrivingCars Pod-Cast Episode 161 – Alberto Stochino
F. Fishkin, June 17, “Is less data sometimes more when it comes to driverless vehicle technology? Perceptive Founder and CEO Alberto Stochino joins Princeton’s Alain Kornhauser and co-host Fred Fishkin for that plus the DOT’s new plan for sharing autonomous safety information, the latest from Tesla, EVs from China and more.” “Alexa, play the Smart Driving Cars podcast!“. Ditto with Siri, and GooglePlay … Alain
SmartDrivingCars Zoom-Cast Episode 161 – Alberto Stochino
Video version of SmartDrivingCars PodCast 161 .… Alain
The SmartDrivingCars eLetter, Pod-Casts, Zoom-Casts and Zoom-inars are made possible in part by support from the Smart Transportation and Technology ETF, symbol MOTO. For more information: www.motoetf.com. Most funding is supplied by Princeton University’s Department of Operations Research & Financial Engineering and Princeton Autonomous Vehicle Engineering (PAVE) research laboratory as part of its research dissemination initiatives.
SmartDrivingCar Zoom-inar 004 Insurance: For or Against SmartDrivingCars?
Live: Tuesday, June 23, 2:00pm New York Time
Free Pre-registration is required
Michigan, Fiat Chrysler join federal self-driving car initiative
K. Laing, June 15, “… But safety advocates sharply criticized the transportation department for sticking with an voluntary approach to self-driving regulation that critics have described as toothless, citing fatal accidents involving partially autonomous cars that have been operated in recent years by companies such as Tesla and Uber. ” Read more Hmmm… How much more “teeth” could regulations and US DoT contribute that make a dent in the “teeth” that the economy sink into Uber’s bottom for its pathetic system design and implementation that disregarded stationary objects in the lane ahead. That caused $60B in valuation to evaporate.
Tesla has somewhat of an excuse to disregard stationary objects in the lane ahead. It explicitly puts that responsibility on the human driver. If bad things happen, it is the driver’s misbehavior that’s the root cause. AutoPilots have always required driver supervision whether they be in Boeings or Teslas. Alain
The Plan to Merge Geely Auto and Volvo Cars, plus more
M. Sena, July, 2020 issue, “The July issue of The Dispatcher continues the theme of the impact of China on the global automotive industry by looking at the proposal of Volvo Car’s parent company, Geely, to merge Volvo Cars and Geely Auto and take the combined company public. In Musings I discuss the issue of vehicle connectivity again, and in Dispatch Central I review current attempts to promote electrification through incentives that are also aimed at helping the automobile industry get through the COVID-19 pandemic….” Read more Hmmm… Another excellent issue. See especially the whole section: Musings of a Dispatcher: Vehicle Connectivity
We’re all speaking different languages. … a follow-up to the most interesting Zoom-tank Zoom-inar 003 on Connectivity in which Michael was the provacteur. Alain
Renault Rattled by Threat From Chinese Electric Cars Imports
T. Patel, June 12, “A flood of Chinese electric cars coming ashore in Norway — one of the biggest markets for battery-powered vehicles in Europe — is a sign of the “ferocious competition” awaiting the region’s automakers.
That warning by Jean-Dominique Senard, chairman of struggling French carmaker Renault SA, comes as European manufacturers roll out more electric cars in the midst of a deep slump brought on by the coronavirus.
Tighter emissions rules across the European Union have prodded automakers to lean into the transition to electric powertrains or face big fines this year. To further spur electric-car sales, the governments of France and Germany have included extra incentives for the purchase of electric vehicles in economic stimulus packages.
As a result, the Chinese models are arriving in a market that’s becoming increasingly crowded with new EVs, including Renault’s updated and best-selling Zoe subcompact and Volkswagen AG’s coming ID.3….” Read more Hmmm… Will this look like the Electric Scooter invassion Re-read Michael Sena’s Dispatchers from March, May and June for more background. Alain
FORD CO-PILOT360™ TECHNOLOGY ADDS HANDS-FREE DRIVING, OVER-THE-AIR UPDATES AND MORE TO HELP FORD CUSTOMERS FEEL MORE RELAXED AND CONFIDENT
Press release, June 18, “Ford Co-Pilot360™ Technology – a comprehensive collection of available driver-assist features – adds new offerings including Active Drive Assist, allowing for hands-free driving on more than 100,000 miles of divided highways in all 50 states and Canada.
“The stress of long highway drives remains a huge issue for drivers around the world,” said Hau Thai-Tang, Ford’s chief product development and purchasing officer. “By introducing driver-assist technologies like Active Drive Assist, Ford’s version of hands-free driving, we’re allowing our customers to feel more confident whenever they’re behind the wheel.”
Active Drive Assist is the next evolution of Intelligent Adaptive Cruise Control with Lane Centering from Ford, adding a first-for-Ford Hands-Free Mode with the potential for more enhancements in the future.1…” Read more Hmmmm…. Excellent… Active Drive Assist sounds very much like GM’s Super Cruise and their TM Co-Pilot360 may well infringe on ALK’s CoPilot wich was trademarked last century. Alain
Cadillac Super Cruise vs. Tesla Autopilot
R. Glon, June 10, “Tesla’s Autopilot technology is one of the most-hyped and best-known suites of electronic driving aids, but it’s not the only (or the best) system of its kind on the market. Cadillac’s Super Cruise isn’t as well known, yet it’s outstandingly safe and, in some ways, smarter than Autopilot. Join us for a look at how these rivaling systems work, the ways they’re similar, and the areas where they differ….” Read more Hmmmm…. Informative. Alain
We’re never going to get meaningful data on self-driving car testing
A. Hawkins, June 15, “The US Department of Transportation launched a new voluntary program to collect and share data from autonomous vehicle operators. Companies testing AVs on public roads are invited to submit information to the government, which will then publicize it online. But given the voluntary nature of the program, safety advocates say the effort is likely to fall short of providing useful data to the public.
There is currently no federal rule requiring AV companies to submit information about their testing activities to the government. Instead, a patchwork of state-by-state regulations govern what is and isn’t disclosed. California has the most stringent rules, requiring companies to obtain a license for different types of testing, disclose vehicle crashes, list the number of miles driven, and the frequency at which human safety drivers were forced to take control of their autonomous vehicles (also known as a “disengagement”). Unsurprisingly, AV companies hate California’s requirements…. “. Read more Hmmmm….Testing with “safety drivers” behind the wheel is really just like driving a Tesla with autoPilot. No need for the Feds or States to require anything.
More importnt is the understanding of the decision process and the evidence that a company will use to offer mobility services without a driver or attendant on-board. Those data and evidence should be captured in a certification process, not a testing process. What process will convince a community, a State, a Federal government, a … to welcome such driverless mobility services in its prescribed Operational Design Domain (ODD). Alain
Tesla admits its approach to self-driving is harder but might be only way to scale
F. Lambert, June 18, “Tesla’s head of AI admitted that the automaker’s approach to self-driving is harder than what most companies in the industry are doing, but he says it’s the only way to scale.
There are dozens of high-profile companies working on solving self-driving and virtually as many different approaches, but there are two main differences: those who rely mainly if not entirely on computer vision and those who rely on HD mapping. Tesla falls in the former category of relying on computer vision.
Andrej Karpathy, Tesla’s head of AI and computer vision, is leading this effort. Earlier this week, he participated in a CVPR’20 workshop on “Scalability in Autonomous Driving” during which he gave an update on the status of Tesla’s program and talked about the scalability challenges:… ” Read more Hmmmm… Must Watch video! At 2:30 in, does AutoPlot actually know the clearance in the lane ahead under the 2nd overpass? I sure hope so.
The problem with “relying” on HD maps is that they contain zero information about the objects moving in the scene ahead, You need something else (vision) to “see” these objects and to tag their location, velocit, acceleration relative to you so you can avoid hitting them or them hitting you. Since vision can also deal with the stationary objects, the HD aspects of digital maps are not necessary and tough to justify even as a redundancy. Alain
New method makes more data available for training self-driving cars
CMU, June 17, “…In the past, state-of-the-art methods for training such a system have required the use of labeled datasets—sensor data that has been annotated to track each 3-D point over time. Manually labeling these datasets is laborious and expensive, so, not surprisingly, little labeled data exists. As a result, scene flow training is instead often performed with simulated data, which is less effective, and then fine-tuned with the small amount of labeled real-world data that exists.
Mittal, Held and robotics Ph.D. student Brian Okorn took a different approach, using unlabeled data to perform scene flow training. Because unlabeled data is relatively easy to generate by mounting a lidar on a car and driving around, there’s no shortage of it.
The key to their approach was to develop a way for the system to detect its own errors in scene flow. At each instant, the system tries to predict where each 3-D point is going and how fast it’s moving. In the next instant, it measures the distance between the point’s predicted location and the actual location of the point nearest that predicted location. This distance forms one type of error to be minimized… ” Read more Hmmmm…. I don’t understand… “the actual location of the point nearest that predicted location” is a label. Something had to classify/label that “actual” point. LiDAR doesn’t classify each point. If it did, you wouldn’t need any “method”, new or not. ???? Alain
Tesla deliveries unsurprisingly crash in US amid pandemic but it’s not that bad
F. Lambert, June 17, “Tesla deliveries have fallen ~37% in California and ~30% in the US based on registration data, but it’s not surprising amid the pandemic and actually not even that bad of a drop. Due to the pandemic, Tesla had to shut down its Fremont factory for just over a month in March and April.
It is where Tesla produces most of its vehicles and therefore, the shutdown is expected to have a major impact on Tesla’s sales and ultimately its financial results during the second quarter.
The Wall Street Journal now reports data registration for Tesla in California, Tesla’s biggest market in the US, and it shows a 37% drop in April and Ma:… ” Read more Hmmmm… Impressive. Need to wait for China deliveries in those months. Could be really impressive. Alain
Driver free but virus free? Robo cars hit new speed bump
T. Krisher, June 17, “… An executive with self-driving car company Waymo said Wednesday that the coronavirus pandemic forced it to put its limited ride service in the Phoenix area on hold to make sure human backup drivers and passengers were safe.
The passenger-carrying service hasn’t resumed yet, but testing restarted on May 8. Some of the rides were shifted toward delivery, Patrick Cadariu, Waymo’s head of supply chain operations, said on a webinar….” Read more Hmmmm…. On the positive side, the coders continue to imprve the code Making these driverless cars safe,remains to be the “critical path” to driverless mobility services. This may not have hit a speed bump, but instead has accelerated because the coders may have actually become more productive. Just a thought. Alain
Watch This Stopped Tesla Model 3 Get Violently Rear Ended By A Chevy SUV
S. Loveday, June 11, “Watch This Stopped Tesla Model 3 Get Violently Rear Ended By A Chevy SUV…” Read more Hmmm… See video. So much for the automated emergency braking system on the Chevy SUV.. It either didn’t have one or the one it had didn’t work. Alain
Toyota Looks Pretty Smart Right Now On Autonomous Vehicles
T. Davenport, June 10, “…All of which makes Toyota’s strategy on smart cars the smartest one around. For years it’s been pursuing Guardian—a project at the Toyota Research Institute (TRI) focused on making human driving smarter and safer. …” Read more Hmmm… What I’ve been calling Safe-driving cars. Nice that Toyota continue to work on these. It would be nice if they were available in showrooms. It would be pretty smart if they were in showrooms now. Alain
Draft Program 4th Annual Princeton SmartDrivingCar Summit Postponed until Evening Oct. 20 through Oct. 22, 2020 (Maybe???)
A. Kornhauser, Feb 6, “The focus of the Summit this year will be moving beyond the AI and the Sensors to addressing the challenges of Commercialization and the delivery of tangible value to communities. We’ve made enormous progress with the technology. We’re doing the investment; however, this investment delivers value only if is commercialized: made available and is used by consumers in large numbers. Demos and one-offs are “great”, but to deliver value that is anywhere near commensurate with the magnitude of the investment made to date, initial deployments need to scale. We can’t just have “Morgantown PRT Systems” whose initial deployment has been nothing but enormously successful for 45 years (an essentially perfect safety record, an excellent availability record and customer valued mobility). Unfortunately, the system was never expanded or duplicated anywhere. It didn’t scale. It is a one-off.
Tests, demos and one-offs are nice niche deployments; however, what one really needs are initial deployments that have the opportunity to grow, be replicated and scale. In 1888, Frank Sprague, successfully deployed a small electric street railway system in Richmond, Va. which became the reference for many other cites. “… By 1889 110 electric railways incorporating Sprague’s equipment had been begun or planned on several continents…” Substantial scaled societal benefits emerged virally from this technology. It was eventually supplanted by the conventional automobile but for more than 30 years it delivered substantial improvements to the quality-of-life for many.
In part, the 4th Summit will focus on defining the “Richmond” of Affordable Shared-ride On-demand Mobility-as-a-Service. The initial Operational Design Domain (ODD) that safely accommodates Driverless Mobility Machines that people actually choose to use and becomes the envy of communities throughout the country. ” Read more Hmmmm… Draft Program is in flux. Consider all named individuals as “Invited yet to be confirmed”. Alain
C’mon Man! (These folks didn’t get/read the memo)
Calendar of Upcoming Events:s
Insurance: For or Against SmartDrivingCars?
Live Tuesday, June 23 @ 2pm New York Time
Postponed, until Evening Oct. 20 -> Oct 22.
4th Annual Princeton SmartDrivingCar Summit
On the More Technical Side