43rd edition of the 5th year of SmartDrivingCars
S. Chirls, Dec. 20, “Three people were killed and more than 70 were injured as an Amtrak Cascades train derailed early Dec. 18 while traversing a curve leading into an overpass at Interstate 5 southwest of Tacoma, Wash., sending a locomotive and passenger cars crashing onto the highway below. The National Transportation Safety Board has identified the cause of the wreck as an overspeed condition, citing a lapse in situational awareness as a potential contributing factor…
Preliminary information indicated that the emergency brakes deployed automatically … pretty worthless automated emergency brakes, Siemens. They must be of the same design that the SAE designed for cars… applied for crash mitigation purposes, not crash avoidance. And how good was the mitigation this time?! So bad!!… and were not manually activated by the engineer, NTSB member Bella Dinh-Zarr said, citing data from the locomotive’s event recorder. Positive Train Control had been installed on the right-of-way, but wasn’t operational, said Geoff Patrick, spokesman for Sound Transit, which owns the right-of-way. The target date for having PTC up and running for the segment of the track where the derailment occurred is the second quarter of 2018. Locomotives and cab-control cars also need to be equipped with PTC… Read more Hmmmm… So once again the NTSB will throw the engineer “under the bus” rather than senior management that didn’t implement PTC in a timely manner. The same thing is happening in the SafeDriving car world where the mentality remains at SAE, NHTSA and the OEMs that the driver is in control, the automated systems are warnings and the mentality is crash mitigation instead of crash avoidance as I’ve commented previously. Maybe this is another wake-up call although there have been many and everyone just hits the snooze button. And then there is the NYTimes with their typical “email server” and “flu symptoms” reporting….
K. Johnson, Dec 20,” … she said the badly damaged cameras in the engineer’s cab — one facing forward, and the other inward, toward the person driving the train — had been sent to the safety board’s laboratory in Washington D.C….Read more Hmmmm… and did the NTSB obtain the cameras from the boardrooms that recorded the stalling, feet dragging that permitted this train to operate without the mandated safety systems? And why haven’t the railway unions been more aggressive in making sure that their ranks are working in safe conditions. And, oh by the way, where is OSHA on this one??? Snooze button!!
So NTSB, please point the finger in the right direction this time and NOT at the engineer. Alain
Episode 16 of the Smart Driving Cars Podcast. The Amtrak crash: who is to blame? Uber’s European problem. Yann LeCun at the Institute for Advanced Study. All this along with the latest on Apple, Volvo and Tesla in Episode 16 of the Smart Driving Cars podcast with Princeton University’s Alain Kornhauser and co-host Fred Fishkin. Listen
L. Alderman, Dec. 20, “…The decision by the European Court of Justice found that Uber operates more like a transportation service than an online platform that matches passengers with drivers. It is likely to restrict the company from expanding services that allowed nonprofessional drivers to offer rides to clients.
While the ruling focused on these so-called peer-to-peer operations, it is likely to be scrutinized by regulators looking more broadly at the gig economy, a growing part of the work force, in which people operate as freelancers or on short-term contracts as opposed to holding permanent jobs…” Read more Hmmmm… A couple of major issues here..
1. What does it take to convert a “gig worker” into a professional? This implies a whole new educational/training activity, probably largely on-line, that will efficiently and effectively allow greater accessibility to many into these professions without the unnecessary hurdles that many professional regulations and their societies place on the professions to create scarcity, and eliminate competition, and
2. If you thought that Driverless Cars were irrelevant to the valuation of Uber/Lyft/Didi … you should think again. Driver costs and driver shortages just went up. How much it will cost, depends how well 1) above can be done. The real cruncher is that it further caps scale, which has always been the Achilles heel of the valuations with and without Driverless Cars.
On a broader societal note this is all part of the continuing eternal social welfare question in this 100th anniversary of the Bolshevik Revolution. The gig economy has improved the welfare of many, but as with any improvement, there are costs. How well those costs are mitigated will depend on whether the . Alain
During this public lecture held at the Institute on December 12, Yann LeCun, Director of Facebook AI Research and Silver Professor of Computer Science at New York University, explored deep learning and the principles and methods for predictive learning.
This lecture was part of the Theoretical Machine Learning Lecture Series, a new series curated by Sanjeev Arora, Visiting Professor in the School of Mathematics, and was made possible by a gift from Eric and Wendy Schmidt. Read more Hmmmm… Non-trivial that this lecture was at The Institute for Advanced Study!! See video of talk. A Follow-up to his talk at last May’s Princeton SDC Summit. Alain
J. Figas, Dec. 12, “…AI research director Ruslan Salakhutdinov made a presentation this week that revealed more of what the company’s autonomous driving team has been up to. Some of the talk was familiar, but there were a few new examples of how far the fledgling project had come.
To start, Apple has crafted a system that uses onboard cameras to identify objects even in tricky situations, such as when raindrops cover the lens. It can estimate the position of a pedestrian even if they’re hidden by a parked car. Other additions included giving cars direction through simultaneous localization and mapping, creating detailed 3D maps using car sensors and decision-making in urgent situations (say, a wayward pedestrian)…” Read more Hmmmm… Be sure to see the paper (which I’ve previously referenced) . Alain
“This optical illusion, recently discovered by experimental psychologist Kohske Takahashi from Japan’s Chukyo University, is called the ‘curvature blindness illusion’, and it’s just the latest example of how we can’t always trust our eyes when we’re looking right at something….”We propose that the underlying mechanisms for the gentle curve perception and those of obtuse corner perception are competing with each other in an imbalanced way,” Takahashi explains in his paper, “and the percepts of corner might be dominant in the visual system.”
But why do corners trump curves instead of the other way around? Hypothetically speaking, it could be because of how humans have had to accommodate the contrived geometries of the modern physical world around us. “I’d say that our eyes and brain may have been evolutionarily adapted to detect corners more efficiently than curves,” Takahashi …”We are surrounded by artificial products, which have much more corners than [the] natural environment does, and hence our visual…”…” Read more Hmmmm… This and Escher and all sorts of other optical illusions. Alain
P. Abbeel, Dec 2017, “Many Unsolved Pieces to the AI Robotics Puzzle” Read more Hmmmm… Good overview. Alain
J. Dean, Dec. 2017, “…Deep learning is transforming how we design computers…” Read more Hmmmm… More excellent perspectives. Alain
J. Walker, Sept. 4, 1987, “This was my most outrageous excursion into Commodore 64 programming. It’s a complete neural network associative memory pattern recogniser implemented in fewer than 250 lines of BASIC. You can train the network on a variety of patterns (letters and numbers) and then recognise similar patterns. I submitted this (serially) to all of the Commodore magazines, each of which rejected it as “too esoteric”. I prefer to think of it as simply before its time. ..” Read more Hmmmm… We have advanced, but are we still not close??? Alain
Press release, Dec. 12, “Volvo Cars will from today start developing its autonomous drive cars with help from Swedish families who will test its cars on the public roads of Gothenburg and feed back their impressions to Volvo Cars engineers. The first two families, the Hains and the Simonovskis from the Gothenburg area, have now received the Volvo XC90 premium SUVs with which they will support the Drive Me project. Three more families will follow early next year and over the next four years, up to 100 people will be involved in Drive Me…” Read more Hmmmm…. Nice to see Volvo crowd sourcing the DriveMe Gothenberg tests. Alain
T. Simonite, Dec 8, “Rockets, electric cars, solar panels, batteries—whirlwind industrialist Elon Musk has set about reinventing one after another. Thursday, he added another ambitious project to the list: Future Tesla vehicles will run their self-driving AI software on a chip designed by the automaker itself. “We are developing customized AI hardware chips,” Musk told a room of AI experts…” Read more Hmmmm…. I guess Elon wants to be the Henry Ford of Driverless cars by controlling all of the pieces of the supply chain himself. Great theory, but will he really be able to remain competitive with the nVIDIAs of this world?? Alain
J. Laughlin, Nov 24, “…Among the findings:Ride-hailing users skew younger, better educated, and richer, and the service is more widely used in cities than the ‘burbs. Ride hailing causes a 6 percent reduction in bus ridership, and a 3 percent reduction in light rail ridership. On the other hand, the study found ride sharing was associated with a 3 percent increase in commuter rail ridership. Ride-hailing users walked more, too, the study reported….” Read more Hmmmm…. Of course it reduces Transit use because it is BETTER than conventional transit for those with and without autos (technically I consider it as a new form of transit. It is mobility as a service, which is/should be, the fundamental definition of transit.) Of course it doesn’t reduce auto use very much because it isn’t better to auto use if you don’t have an auto. Alain
Some other thoughts that deserve your attention
Half-baked stuff that probably doesn’t deserve your time
C’mon Man! (These folks didn’t get/read the memo)
Calendar of Upcoming Events:
Lectern Session 225
Automated Driving Speed Dating
Monday 8:00 AM- 9:45 AM
46th Annual Princeton University Transportation Program Reunion Banquet
TUESDAY, January 9, 2018
May 16 & 17, 2018
Save the Date
Episode 15 of the Smart Driving Cars Podcast. Hosts Fred Fishkin and Princeton University’s Alain Kornhauser are joined by leading expert Michael Sena from Sweden in a wide open and most entertaining chat ranging from the impact of Ralph Nader to the insurance industry’s role, to the latest from Ford, Lyft, Uber and China’s Didi.
Episode 13 of the Smart Driving Cars Podcast with host Fred Fishkin and Princeton University Professor Alain Kornhauser. This edition In this edition Fred and Alain are joined by Bernard Soriano, the Deputy Director of the California Department of Motor Vehicles. On the agenda: Waymo’s CEO says real driverless testing is coming soon.; Waymo’s autonomous fleet now has traveled four million miles; Lyft gets the green light from California to test self driving on public roads
Episode 11 of the Smart Driving Cars Podcast with host Fred Fishkin and Princeton University Professor Alain Kornhauser. Fred and Alain are joined by leading expert and Internet pioneer Brad Templeton. Waymo makes some history, Thee tech needed to make it work..cameras…lidar or both? Navya bringing new robotic vehicles to Paris. And an accident…as a self driving shuttle is launched in Las Vegas.
F. Fishkin, Oct 25, Episode 9 “Host Fred Fishkin with Princeton University’s Alain Kornhauser and guest Fred Payne, council member from Greenville County, South Carolina. Greenville’s autonomous taxis are rolling. Bank of America analysts see big investment opportunities in vehicle technology. The latest from London, China and New York. And on demand pilotless planes?
Recent Highlights of:
B. Jones, Dec. 6, “Self-driving cars are expected to make our roads safer. Now, UK insurance company Direct Line is offering a discount to customers who use Tesla Autopilot to facilitate research into its effects…..” Read more Hmmmm… Guess what.. I bet the discount is substantially less than the expected reduction in LOSS! Win-Win!! Why is this not happening in the US? Alain
Personal Sedan Sales in Jeopardy as U.S. Auto Market Transitions to “Islands” of Autonomous Mobility: KPMG Research
Nov 27, ” KPMG predicts that self-driving cars and mobility services will provide options that will reduce consumer desire to own cars, particularly sedans. Pushing a button for mobility services competes with the utility of sedans, and both give consumers the freedom to buy the car they really want to own or utilize mobility by the trip. In fact, KPMG projects that sales of personally-owned sedans in the U.S. will drop precipitously – from 5.4 million units sold today to just 2.1 million units by 2030….” Read more Hmmmm… See video, See full report next, Excellent but they don’t sufficiently differentiate between Self-driving and Driverless (and don’t even bother with Safe-driving which is unfortunate. But excellent anyway because they approach it from the individual trip demand. Alain
S. Masunaga & R. Mitchell, Nov. 20, ” fleet of self-driving Volvo vehicles operated by Uber Technologies Inc. could be ready for the road as early as 2019, marking the ride-hailing firm’s biggest push yet to roll out autonomous cars. Volvo said Monday that it would sell Uber tens of thousands of luxury sport utility vehicles between 2019 and 2021 outfitted with the Swedish automaker’s safety, redundancy and core autonomous driving technologies. Uber will then add its own self-driving technology to the autonomous taxi fleet…” Read more Hmmmm… This is a significant announcement and recognizes that it is going to take another year-plus for the Uber/Volvo existing ‘Self-driving‘ technology stack (which now requires an Uber attendant in the car) to become ‘Driverless’ (can operate safely without an Uber attendant in each car). If these cars don’t become Driverless, their cost per ride will be so prohibitively high that their use will not be sustainable.
24,000 is a reasonable number with which to start These vehicles would become Uber’s work horses. They’ll operate ~20 hours a day and could serve ~5 short trips per hour when concentrated in Uber’s highest demand areas. With some ride sharing they could serve 100 person trips per day, allowing them to serve nearly half of Uber’s current 5.5 million trips a day. Unfortunately, this is the short-trip half. The half remaining is dominated by long trips. Vehicles serve these at only about 1.5 trips per hour. Vehicle (driver) productivity is consequently limited to about 30 trips per day. That means that each day Uber will still need two shifts of 100,000 gig workers each to show up and deliver the mobility services needed to serve their 3 million daily longer trips. To really scale, Uber will need to order many more of these Volvos and and get them to operate Driverlessly in much larger geographic areas so that they can serve some of these long trips.
In the US there are about 1 Billion vehicular trips per day. Many are short, some are long, very few are very long. It is doubtful that a Driverless car could serve more than 2.5 person trips per hour or 50 blended-length trips per day. Thus, to serve 10% of the Billion trips per day would require a fleet of about 2 million Driverless cars. In 2016, 17.5 Million cars & light trucks were sold in the US. By devoting about 10% of the car & light duty truck manufacturing capacity to the production of Driverless vehicles, enough Driverless cars are produced in a year to serve 10% of all US vehicular trips. So the manufacturing capacity exists to enable an Uber or Lyft or Didi or Waymo or … to in a few years serve many/most trips in the US.
This suggests to me that Waymo must have already established a deal/arrangement with a manufacturer to begin very soon to produce thousands of cars that can accept Waymo’s Driverless stack of hardware and software and aggressively begin to serve pockets of those 1Billion daily person trips. Alain
NYTM Editor, Nov 7, “IT IS A truth as visible from space, in the tendrils of our brightly lit megaroadways, as it is under a microscope, in the diesel particles that tar our lungs: America is a car country. This truth has been true for more than a hundred years now. Soon after the turn of the 20th century, the automobile became pervasive in urban centers, establishing a new, briskly perilous pace for city life; by the 1930s, as annual vehicle sales surpassed four million, urban streetcar systems had begun to go defunct, scrapped in favor of gas-burning buses; by the mid-1950s, new cars increasingly rolled off the line into a suburban way of life that the car itself had made possible. Suburbs had already swallowed a quarter of the population when, at the behest of President Dwight D. Eisenhower in 1956, Congress inaugurated the most ambitious infrastructure project in the nation’s history: a 47,000-mile Interstate highway system. Its construction would crisscross the nation and spread even more suburbia as it went, bringing a profusion of new features to the American landscape — not just countless cloverleafs and imposingly stacked overpasses but drive-ins, drive-throughs, cul-de-sacs, acres upon acres of blacktop parking.
As the car transformed the American lifestyle, so did it colonize the American imagination. Perhaps the most fascinating artifact of automotive retrofuturism dates to 1958, two years into the Interstate building boom, when Disney produced an hourlong TV program called “Magic Highway, U.S.A.” After a half-hour or so celebrating the car’s ascent, the program pivoted to envisioning its future. At a time when Eisenhower’s Interstate project — inspired, famously, by his awe at traveling Hitler’s autobahn during the Nazi overthrow — was seizing private property through eminent domain around the country, Disney imagined atomic-powered tunnel borers and imperial road-building machines as tall and wide as skyscrapers, cutting through landscapes and leaving fully constructed roadways in their wake. Dreaming further forward, the program extrapolated from the new real-world highways to envision literal high-ways of clear tubes raised far above the urban environment, magnificent air-conditioned arteries that someday would “link together all nations and help create a better understanding among the peoples of the world.”…” Read more Hmmmm… Wow! The whole magazine about SmartDrivingCars. Wow! Must reading. Caution: The only way this utopia doesn’t turn into a disaster is if, when pertinent, we Share Rides! It is not pertinent when the trip density is so low that there isn’t anybody around going in the same direction at about the same time. Then it is appropriate and no problem to go alone.
However, when there are other people going our way at about the same time, then we should go together instead of adding to congestion by going alone. When several people are waiting for an elevator going up, they tend to all get in together and go up. The alternative would need to be for the whole building to be made of nothing but elevator shafts (same as paving our way out of congestion).. Then each could ride alone without waiting. Without more elevator shafts, then each could wait and incur intolerable delay by insisting on riding alone as, happens today on our streets and highways at rush hour. A fleet of centrally managed driverless vehicles create “horizontal automated elevators” that can facilitate sufficient ride-sharing to deliver utopian mobility to all without incurring vehicular congestion. This enables moderately dense living environments to be very attractive. Alain
AP, Nov. 7, 2017 “Waymo, the self-driving car company created by Google, is pulling the human backup driver from behind the steering wheel and will test vehicles on public roads with only an employee in the back seat.
The company’s move — which started Oct. 19 with an automated Chrysler Pacifica minivan in the Phoenix suburb of Chandler, Ariz. — is a major step toward vehicles driving themselves on public roads without human backup drivers. …” Read more Hmmmm… Not to be too critical, but Waymo is still just ‘Self-driving’ . While they moved the ‘engineer’ with the ability to ‘take over and drive the vehicle’ from behind the wheel to the back seat, this is just a step along the broad ‘Self-driving’ continuum which is a vehicle that, under certain circumstance, can drive itself, but does that only if there is a person ready and able to take over if the unexpected appears.
The big-leap/major-step will come when Waymo removes the ‘engineer’ entirely from the vehicle and it is human-less when it arrives to pick up a passenger and drives away human-less after the last passenger(s) disembark. That enormous leap-of-faith in the technology will mark Waymo’s inception of the Driverless Era. (or what Waymo prefers to call ‘Fully Self-driving’ era.)
Just to be clear, when that time comes, I’m sure that Waymo will have telemetry throughout that Driverless vehicle and there will be a room full of engineers in Waymo’s ‘Situation Room‘ ready to take over the driving should the need arise. However, until that time, Waymo is just like all the other wanabes, they are just ‘Self-driving’ without the ‘Fully’.
The reason why ‘remote emergency driving’ is ‘Driverless’ is because it scales. By that I mean that it takes the provision of horizontal mobility on our public streets from needing at least one human per vehicle to needing less than one human per vehicle. Initially the remote driver will monitor one car. Before you know it that person will be monitoring two, four, eight, … vehicles and truly Driverless with zero remote human oversee-ers will be approached asymptotically. But just like the old saw between the engineer and the mathematician: engineer and mathematician were sitting on a bench recalling their youth… Engineer said “Long ago, I was sitting on this very bench with my girl. We wanted to kiss but we were too far apart. So we agreed to move towards each other by halving the distance between us on each move. The mathematician blared ” You’re so stupid! If you did that, you never came together!” The engineer just smiled: “we got close enough!”. Alain
J. Lowy, AP, Nov 1, “he Trump administration has quietly set aside plans to require new cars to be able to wirelessly talk to each other, auto industry officials said, jeopardizing one of the most promising technologies for preventing traffic deaths….
…The administration has decided not to pursue a final V2V mandate, said two auto industry officials who have spoken with White House and Transportation Department officials and two others whose organizations have spoken to the administration. The industry officials spoke on condition of anonymity so as not to jeopardize their relations with the administration….” Read more Hmmmm… This basically completes Washington’s “540 degree turn” on Safety for cars. A complete spin from V2V and CV and on to AV.
One could sense this coming when it took until his 6th slide for Carl Andersen from US DoT to even mention the word ‘Connected’. Four years ago at the 2nd AV conference in Palo Alto, the US DoT folks from would only talk about Connected Vehicles even though it was an AV conference. More recently the pivot was underway when Washington coined CAVs (Connected & Automated Vehicles; I complained that this gave only lip service to AVs since an alphabetic ordering was passed over. Connected was obviously favored.) Secretary Foxx continued the pivot into a spin with his Federal AV Policy statement in September 2016. More spin a month ago with Automated Driving Systems v2.0 and now this to put the nail in the coffin. Jerome Lutin and I have begun to sit Shiva and Paul Brubaker has started covering the mirrors for DSRC.
This substantial change in Washington has profound implications because so many in State & Local government were following the Washington lead and will now need to pivot. Europe, Japan and others in transport planning around the world were also following Washington in the promotion of CVs. Much of the “Smart Cities” and ITS objectives are/were all about Connectivity and the implied control and orchestration of societies to achieve some optimized utopia. Always seemed too “1984” for my taste. Seeking some perfect “Best” when one can’t even reach a consensus on what “Best” is really the enemy of “good enough” . Give everyone a little room and let them individually work towards what they consider is best from their perspective. In my politics, this is my view of “Smart Cities” and SmartDrivingCars.
This change has also made obsolescent, if not completely obsolete, some recent reports such as much of the NCHRP 20-102 research which was initially motivated by CV and reports such as the recently released Future Cities: Navigating the New Era of Mobility.
The fundamental problem with V2V is that it doesn’t work unless there are other cars with which to communicate. That doesn’t happen until the adoption level is substantial. Assume that if the V2V communications works (is fast enough and the right data is communicated perfectly) it has no chance of improving safety unless BOTH vehicles that are about to crash have the technology. The chance that exists, ie the probability that BOTH cars have the technology, is the square of the adoption level. At the beginning it is zero times zero which is zero. But even at a 10 % penetration level, which would take an exhaustive mandate at least two year to achieve, it would only reach a 1% chance of being relevant. When half of the cars have it (and ‘it’ works in all of the cars that have it, whatever ‘it’ is) the chances are only 1 in 4 that it is relevant, It takes a 70% adoption (and working) level, which would take at least 10 years to achieve, before it become better than a coin flip. That’s a long time before those that bought the hardware can have a reasonable expectation of capturing some benefits. Alain
Draft for Public Comment, October 19, 2017, “This Strategic Plan establishes the strategic goals and objectives for the DOT for FY2018 through FY2022. The GPRA Modernization Act of 2010 (GPRA) aligns strategic planning with the beginning of each new term of an Administration, requiring every Federal agency to produce a new Strategic Plan by the first Monday in February following the year in which the term of the President commences. The Strategic Plan, therefore, presents the long-term objectives an agency hopes to accomplish at the beginning of each new term of an Administration by describing general and long-term goals the agency aims to achieve, what actions the agency will take to realize those goals, and how the agency will deal with challenges and risks that may hinder achieving results….” Read more Hmmmm… Nice…seems to be void of all references to Connected Vehicles (CV), V2V and V2I that were in the previous FY 2014-2018 strategic plan (Gone is any reference like footnote 13, p.24: “Transforming Transportation through Connectivity: ITS Strategic Research Plan, 2010-2014- Program Update, 2012 “ (even the link is 404)
This may finally be a realization by US DoT that the Connected Vehicle (CV) program was a fatally flawed concept, especially in light of having viable Automated Vehicles. CV was a grandiose plan to have the public sector (Washington, States & Municipalities) pay to deploy electronic Gizmos everywhere and have all of our vehicular mobility be centrally controlled (think 1984). It was part of the America’s National ITS Architecture which ITS America has been promoting for years in support of its Gizmo manufacturing members. Unfortunately, the fatal flaw in Architecture is that the benefits (Safety) would not begin to really kick-in until the Architecture was largely deployed throughout the highway infrastructure and installed in most vehicles. It is essentially all-or-nothing, and “all” need to be so ubiquitous, consequentially so expensive, that the public sector was was the only potential financier.
However, along came the private sector and said…”maybe we can address this Safety thing by automating the vehicle so that it is much less likely to Crash while it shares the existing infrastructure without asking for any improvements (except maybe better paint and readable signs; both of which are really needed anyway for all existing users) “.
Sure, safe automation is hard and expensive, but nowhere near as expensive, especially in its early commitments, as the CV approach. And if successful, at least one vehicle and its occupants are safer. PLUS, the cost of the automated vehicle technology is likely get cheaper (it scales) as we replicate for the 2nd, 4th, 8th, 16th… vehiclesand likely to become very affordable very quickly (Moore’s Law) AND the cost of replicating the software is essentially zero and … So, with what amounts to a little bit of money we can get started with one vehicle and it is likely to scale very nicely to initiate viral adoption. Wow! (Deja vu all over again… Steve Jobs’ garage).
So.. it is very nice that US DoT has finally recognized through its strategic planning process that it is time to pivot from its fundamentally flawed CV concept to the SmartDrivingCar/Automated Vehicle (SDC/AV) concept, even though historically US DoT has been all about the infrastructure (roads) and not-so-much about vehicles (cars).
Moreover, this frees US DoT from an enormous future financial obligation. Congratulations for making the pivot. We are all anxious to help you succeed. Alain
Washington DC, Oct 20, “The U.S. Department of Transportation’s National Highway Traffic Safety Administration (NHTSA) is committed to the safe deployment of automated vehicles. NHTSA hosted a public workshop today to get feedback on the Voluntary Safety Self-Assessments discussed in the Automated Driving Systems 2.0: A Vision for Safety guidance released last month. …The workshop , overall, was a productive, open forum, where manufacturers, suppliers, safety advocates, and other entities shared the types of information that could be made available, and opportunities for making that information public. There were over 100 attendees present, and many more who participated via a LIVE Webcast. A full transcript will be available in the coming weeks. ” Read more Hmmmm… Congratulations Nat, I agree that it was productive. The comments the by Global Automakers, AAMVA, MEMA Waymo and AAM were positive and helpful.
The …opening remarks by NHTSA Acting Administrator Heidi King: “… At DOT and NHTSA, of course, our central focus is always on safety. NHTSA’s mission remains to help Americans drive, ride and walk safely…” Given that Safety is central, It is unfortunate that Automated Driving systems 2.0 skips over ‘Safe-driving‘ (ADAS or Level1/2 or whatever) and jumps right into Self-driving (Level 3/4/5 or whatever) to address Safety. Essentially all of the Automated Vehicle Safety achievements (crash avoidance, lane departure avoidance, etc..) will be achieved by Safe-driving vehicles that always over-ride our failures and do the right thing even if they don’t let us take our hands off the wheel or feet off the pedals. These systems are beginning to be made available today and it is not an understatement to say that they don’t work as well as they should/could and there is essentially total confusion in the marketplace/showroom about the capabilities/consumer-expectations about these systems. NHTSA’s 5-Star Safety Ratings program doesn’t even consider any of these systems. Since ‘Safe-driving‘ has the greatest and nearest term potential impact on Safety, why is it NOT part of this AV program? These systems are being tested; shouldn’t NHTSA be calling for a Self-assessment of these systems. Safe-driving systems are beginning to be here now and I contend the public is totally confused.
“…Public trust is essential to the advancement of automated technology….” I wholeheartedly agree!! That trust needs to be earned and its first exposure is mixed. Anti-lock brakes and Electronic Stability Control are automated systems that have earned public trust event though they automatically detect erroneous driver behavior and automatically over-ride those actions in order to do the best that they can to keep the driver safe. But what about these Safe-driving (Level 1/2, …) systems. These are automated systems focused on Safety, yet NHTSA hasn’t even bothered to include any of these systems in its 5-Star Safety Ratings program. The public is totally confused about what is being offered and there seems to be no public trust evernthough these systems are the very foundations of Self-driving and Driverless systems. It is necessary that Safety and public trust be established first in Safe-driving systems. This forms the basis on which to expand that public trust to the downstream systems that deliver other societal benefits, comfort & convenience for Self-driving and affordable mobility for all for Driverless, while providing very little, if any incremental Safety benefits over Safe-driving technology. So… NHTSA’s 1st order of business should be to ensure that Safe-driving technology actually works and is valued by car buyers.
A substantial part of the problem here is that the terminology that is being used is totally confusing. NHTSA’s decision to give up on its original 4-Level nomenclature was good, they just chose to adopt an even worse one, SAE’s. It focuses entirely on the details of the technology, rather than on the value that is to be derived from the technology. The Levels invoke no fundamental cognitive relationships; nothing that would inspire…”tell me more”. Thus, engineers might eventually pay attention long enough to absorb the more than 7+/-2 chunks of cognitive information needed to understand the differences in the “Levels”. Unfortunately, corporate buyers, journalists, planning, policy and/or legislative officials and the general public/consumers remain totally confused.
I’ve suggested three categories: Safe-driving…, Self-driving… and Driverless… Not necessarily perfect, because the leader of Driverless chose long ago (~8 years) to call itself Self-driving. Unfortunately, the term Self-driving with human supervision, reinforces the auto industry’s 100-year old business model of selling personal comfort and convenience to consumers. The auto industry doesn’t bother emphasizing the partial nature of its Self-driving. Waymo has chosen to add the prefix “Fully” in an effort to differentiate itself as really Driverless that is fundamentally attractive to a different business model focused on Fleets delivering mobility services to a public that doesn’t own cars. But few are aware of the enormous difference implied by the the existence of the prefix.
In its efforts to engender public trust, NHTSA needs to rethink what it calls these things. An opportunity exists in the re-framing of its Star Ratings, Or maybe, this crash-avoidance technology is so different from the crash-mitigation technology that is NHTSA’s sweet-spot, that a new agency or a new division of NHTSA should be created to provide the crash-avoidance safety oversight. Alain
Rulemaking Actions, Oct 1The following 3 PDFs are important:
1. Autonomous Vehicles Notice of Modification (PDF) Act
3. Autonomous Vehicles 15 Day Express Terms (PDF) Act Hmmmm..This is all about Driverless! Thank you California, and especially Dr. Bernard Soriano, for leading this noble effort and for continuing to distinguish this technology from Self-driving and all of the various other names seemingly meant to confuse. Alain
Press Release, Oct 4, “Federal Highway Administration (FHWA) officials today awarded a $4 million Advanced Transportation and Congestion Management Technologies Deployment (ATCMTD) grant to South Carolina’s Greenville County for its automated taxis.
“Technology is the future of U.S. transportation,” said Acting Federal Highway Administrator Brandye L. Hendrickson. “These funds will help Greenville County lead the nation into a future with more driverless vehicles, which will improve mobility for some and reduce traffic congestion for all.”
County officials will use the funds to deploy an integrated system of “taxi-shuttles,” known locally as “A-Taxis,” on public roads. These are driverless taxis providing shuttle service to and from employment centers–expected to improve access to transportation for disadvantaged and mobility-impaired residents…” Read more Hmmmm… Wow!! FHWA is actually going to fund aTaxis!!! Congratulations, Fred Payne! This is a non-trivial achievement. Alain technology, much as trucks and airlines have with their own ‘Administration’. Alain
M Burns, Aug 3, “Cadillac is about to start selling vehicles with an autonomous driving mode …Once the light bar on top of the steering wheel turns green, the driver can let go…
“Wait for the green light and let go,” the Cadillac engineer instructed. That’s it. The car was driving itself. I, the person behind the steering wheel, was no longer the driver. Cadillac’s Super Cruise system was driving. The 2018 Cadillac CT6 sped along US-23 under the direction of Super Cruise. Traffic was light and the weather was perfect. The system held the Cadillac sedan in lane and responded appropriately to traffic. I spent an hour on the expressway and touched the steering wheel and pedals only a few times. Super Cruise made the drive boring. I think that’s the point….
When active, Super Cruise controls the steering and speed, but again, only on an expressway. This is done through on board sensors and using GPS and mapping data. GM employed GeoDigital, a startup in GM Venture’s portfolio, to map 160,000 miles of expressways in the U.S. and Canada. The car company then used Super Cruise-equipped vehicles to test each mile.
Cadillac’s system also lacks several autonomous features found on Autopilot including the ability to pull the car out of a garage and change lanes by using the turn signals. Hmmmm… fluff features with little value.
Super Cruise’s IR sensors tracks eye location and head movements. As long as the driver looks at the road every seven to 20 seconds, the system works as expected. Hmmmm… Fantastic!
General Motors will have to rely on independently owned dealerships to correctly position this product and train buyers on its capabilities. Hmmmm… Yup!
For better or worse, Super Cruise is built into the CT6 like a standard system and not something a driver must use every time they’re on an expressway. This should help timid buyers. Super Cruise feels like a feature ready for the masses. The system is deeply integrated into the vehicle and using it is akin to using cruise control or turning on the lights. There’s a button for Super Cruise on the steering wheel. Press the button when the system is available and it works. It’s that easy to turn a driver into a passenger. Read more Hmmmm… Over the air updates? See also Motor Trend’s view: “… a stand-alone option (as yet unpriced) on CT6 models with the premium luxury trim package and as standard equipment on top Platinum models (the price of which went up $500 for 2018, if that’s any indication)….” Finally, I guess that I’ll have to go test drive one. Alain
The docket material is available at: https://go.usa.gov/xNvaE” Read more Hmmmm… A few comments…
1. Since lateral control (swerving) couldn’t have avoided this crash (the truck is almost 70 ft long (6 lanes wide) stretching broadside across the highway) , it doesn’t matter if Josh Brown ever had his hands on the steering wheel. That’s totally irrelevant.
2. Why didn’t autobrake kick in when the tractor part of the tractor-trailer passed in front of the Tesla?
3. How fast was the truck going when it cut off the Tesla. I couldn’t find the answer in 500 pages.
4. With sight distances of greater than 1,000 feet, why didn’t the truck driver see the Tesla? Was it the drugs?
5. This intersection invites “left-turn run-throughs” (no stop or yield and a 53 foot median and turn lane need to be crossed before one slips through a gap in two traffic lanes. So you certainly roll into it, (plenty of room to stop if you see something coming) and if you don’t see anything, you hit it. If you’re in the Tesla, you think you’ve been clearly seem, you expect the truck to stop, it doesn’t, you can’t believe it, BAM! All in probably a second or so.
6. The head injury description (Table 1 p2 of 3) certainly suggests that Joshua Brown was seated upright facing forward at impact. The bilateral lacerations on the lower arm from the elbow to the wrist may indicate that he saw it coming in the last second and raised his arms in an attempt to protect his head. The evidence reported doesn’t seem to suggest he saw this early enough to bend toward the passenger seat and try to pass underneath.
7. About 40 feet of tractor and trailer passed directly in front of the Tesla prior to impact. Depending on how fast the truck was traveling, that takes some time. Has NTSB run Virtual Reality simulations of various truck turn trajectories and analyzed what the truck driver and the Tesla driver could/should have seen? Seems like a relatively simple thing to do. We know what the Tesla was doing prior to the crash (going 74 mph straight down the road.) and we know where it hit the truck. How fast the truck was traveling doesn’t seem to be known.
8. Why wasn’t there any video captured from the Tesla. Didn’t that version of the MobilEye system store the video; I guess not, 🙁
Anyway, lots to read in the 500 pages, but there is also a lot missing. I’m not linking the many articles reporting on this because I disagree with many of their interpretations of the facts reported by NTSB. Please reach your own conclusions. Alain
R. Abrams, June 16, “Shares of Walmart, Target, Kroger and Costco, the largest grocery retailers, all tumbled on Friday. And no wonder.. Grocery stores have spent the last several years fighting against online and overseas entrants. But now, with its $13.4 billion purchase of Whole Foods, Amazon has effectively started a supermarket war. Armed with giant warehouses, shopper data, the latest technology and nearly endless funds — and now with Whole Foods’ hundreds of physical stores — Amazon is poised to reshape an $800 billion grocery market that is already undergoing many changes….” Read more Hmmmm… Since Jeff Bezos doesn’t need to have you impulse buy on your walk through the store while you get a quart of milk, he simply has to get you click on organic milk and he’ll present you with everything you absolutely can’t checkout without. All he then needs is to get all those impulse buys (and the quart of organic milk) to your home from the hundreds of physical stores. That’s where low speed driverless local delivery vans come in (operating initially in the early morning hours when the streets connecting those stores to our houses are completely empty and simply drop off everything you’ll need for the day ahead in your “Amazon Box” that’s replaced your 20th Century mailbox). So in the end it will be Jeff Bezos’86 battling Eric Schmit’76 for deploying the first fleets of driverless vehicles sharing our neighborhood streets. If they should decide to join forces and have these vehicles providing mobility whenever anyone wants to travel and moving groceries and other goods the rest of the time, watch-out!!! Then everybody wins!! (except Walmart, Target, Kroger and Costco) See also..Amazon and Whole Foods and Self-Driving Cars Alain
May 18, Enormously successful inaugural Summit starting with the Adam Jonas video and finishing with Fred Fishkin’s live interview with Wm. C Ford III. In between, serious engagement among over 150 leaders from Communities at the bleeding edge of deployment, Insurance struggling with how to properly promote the adoption of technology that may well force them to re-invent themselves and AI (Artificial Intelligence) and the various technologies that are rapidly advancing so that we can actually deliver the safety, environmental, mobility and quality of life opportunities envisioned by these “Ultimate Shared-Riding Machines”.
Save the Date for the 2nd Annual… May 16 & 17, 2018, Princeton NJ Read Inaugural Program with links to Slides. Fishkin Interview of Summit Summary and Interview of Yann LeCun. Read Inaugural Program with links to Slides. Hmmmm… Enormous thank you to all who participated. Well done! Alain
D. Hall, Apr 17, “In the race to the autonomous revolution, developers have realized there aren’t enough hours in a day to clock the real-world miles needed to teach cars how to drive themselves. Which is why Grand Theft Auto V is in the mix.
The blockbuster video game is one of the simulation platforms researchers and engineers increasingly rely on to test and train the machines being primed to take control of the family sedan. Companies from Ford Motor Co. to Alphabet Inc.’s Waymo may boast about putting no-hands models on the market in three years, but there’s a lot still to learn about drilling algorithms in how to respond when, say, a mattress falls off a truck on the freeway….The idea isn’t that the highways and byways of the fictional city of Los Santos would ever be a substitute for bona fide asphalt. But the game “is the richest virtual environment that we could extract data from,” said Alain Kornhauser…” Read More Hmmmm... Well…we have a slightly different view of history wrt to GTA5. The ‘Alain view’ is that Chenyi Chen*16 independently started investigating the use of virtual environments as a source of Image – Affordances data sets to use as the training sets in a ‘Direct Perception’ approach to creating a self-driving algorithm. Images of the road ahead are converted into the instantaneous geometry that is implied by those image. An optimal controller then determines the the steering, brake and throttle values to best drive the car. The critical element in that process are the Image – Affordances data sets which need to be pristine. Chenyi demonstrated in his PhD dissertation , summarized in the ICCV2015 paper, that by using the pristine Image – Affordances data sets from an open-source game TORCS one could have a virtual car drive a virtual race course without crashing. More importantly, when tested on images from real driving situations, the computed affordances were close to correct.
This encouraged us to look for more appropriate virtual environments. For many reasons, including: “wouldn’t it be amazing if ‘Grand Theft Auto 5’ actually generated some positive ‘redeeming social value’ by contributing to the development of algorithms that actually made cars safer; saving grief, injuries and lives”. Consequently, in the Fall of 2015, Artur Filipowicz’17 began to investigate using GTA5 to train Convolutional Neural Networks to perform some of the Direct Perception aspects of automated driving. With Jeremiah Liu, he continued his efforts in this direction last summer which were presented at TRB in January. Yesterday, he and Nyan Bhat’17 turned in their Senior Theses focused on this topic.
Indeed, GTA5 is a rich virtual environment that begins to efficiently and effective address the data needs of Deep Learning approaches to safe driving. Alain
With the upward trend showing no sign of subsiding, NSC is calling for immediate implementation of life-saving measures that would set the nation on a road to zero deaths:…” Read more Hmmm…”Automated Collision Avoidance” or anything having to do with ‘Safe-driving Cars‘ is not mentioned anywhere in the Press Release. One of us is missing something very fundamental here!! So depressing!! 🙁 Alain
Serving the Nation’s Personal Mobility Needs with the Casual Sharing of autonomousTaxis & Today’s Urban Rail, Amtrak and Air Transport Systems
A. Kornhauser, Jan 14, “Orf467F16 Final Project Symposium quantifying implications of such a Nation-wide mobility system on Average Vehicle Occupancy (AVO), energy, environment and congestion, including estimates of fleet size, needed empty vehicle repositioning, and ridership implications on existing rail transit systems (west, east, NYC) and Amtrak of a system that would efficiently and effectively perform their ‘1st mile’/’last-mile’ mobility needs. Read more Hmmm… Now linked are 1st Drafts of the chapters and the powerPoint summaries of these elements. Final Report should be available by early February. The major finding is, nationwide there exists sufficient casual ridesharing potential that a well–managed Nationwide Fleet of about 30M aTaxis (in conjunction with the existing air, Amtrak and Urban fixed-rail systems) could serve the vehicular mobility needs of the whole nation with VMT 40% less than today’s automobiles while providing a Level-of-Service (LoS) largely equivalent and in many ways superior than is delivered by the personal automobile today. Also interesting are the findings as to the substantial increased patronage opportunities available to Amtrak and each of the fixed rail transit systems around the country because the aTaxis solve the ‘1st and last mile’ problem. While all of this is extremely good news, the challenging news is that since all of these fixed rail systems currently lose money on each passenger served, the additional patronage would likely mean that they’ll lose even more money in the future. 🙁 Alain
Hmmm…What we know now (and don’t know):
U.S. DOT and IIHS announce historic commitment of 20 automakers to make automatic emergency braking standard on new vehicles
Video similar to part of Adam’s Luncheon talk @ 2015 Florida Automated Vehicle Symposium on Dec 1. Hmmm … Watch Video especially at the 13:12 mark. Compelling; especially after the 60 Minutes segment above! Also see his TipRanks. Alain