2014-03-15

March 15, 2014Newsletter archive →

March 14, 2014

303 Deaths Seen in G.M. Cars With Failed Air Bags

By DANIELLE IVORY and HILARY STOUT MARCH 13, 2014 "As lawmakers press General Motors and regulators over their decade-long failure to correct a defective ignition switch, a new review of federal crash data shows that 303 people died after the air bags failed to deploy on two of the models that were recalled last month...

The Center for Auto Safety, a private watchdog group in Washington, commissioned the study, and, in a letter to the National Highway Traffic Safety Administration, criticized the agency for not detecting the air bag failures, as well as the defective ignition switch... The Center for Auto Safety's letter said that the 303 victims were in the front seat, where air bags are situated, and had died in non-rear-impact crashes of Cobalts and Ions, in which the air bags did not deploy..." (emphasis added) Read more

Hmmm... This is tragic for many reasons; however, we in the SmartDrivingCars world need to learn very clear lessons. A simple one is not to muddy the focus by making C'mon man! statements like "Then, the computer determines whether to deploy the air bag with full force or at a lower level. The goal is a balancing act to protect the occupant from the impact of the crash, while keeping the air bag itself from causing an injury." The good intentions of the computer to select between full and lower are irrelevant because "...air bags did not deploy..." Unless "lower" is the politically correct way to say "not at all", in which case this achieves a C'mon Man! squared.

More importantly (and as is obvious to the most casual observer) we can't wait for "...303 victims..." of the "we didn't think of that" problem to occur before we fix it.

As SmartDrivingCars evolve, we are going to trip over many "cracks" that we never anticipated. Humans simply aren't very good at anticipating extremely rare events. We'll fare no better at anticipating all of the "where'd that come from?" scenarios that SmartDrivingCars will encounter. What is critically important is that we pay attention to these "cracks" as we encounter them and are diligent in filling them as soon as possible after the "stumble". In order to be diligent about this we must first make sure that the characteristics of the "crack" and the "stumble" are thoroughly recorded and available. Google has driven over 500,000 miles, must have encountered some cracks, but have shared none with the SDC community. That's wrong! As I reported last week, I suspect that my new S-Class does not record when the 997 package is working well or when it is challenged. I understand that my privacy is important and that MB needs to not expose its behind; however, if this is coming at the expense of not recording the characteristics of the unexpected and the corresponding "stumble", then my position is that Mercedes has struck the wrong balance. I'll go out on a limb here and assert that essentially all other manufacturers have struck the same inappropriate balance as MB. Hopefully I am way off base here and the industry will correct me. They know my email address. If they write, I'll be certain to correct the pages of SmartDrivingCars.com (however, the silence may be deafening).

In case I am correct here, one of the things that SmartDrivingCars.com is going to dedicate itself to is creating an environment for the collection and distribution/exposure of the "cracks" and "stumbles" encountered in the evolution of SmartDrivingCars. At first it is not going to be very pretty or easy. As I pointed out last week, we have created an email address for the readers of this newsletter who own one of the early forms of SDC to manually share their "cracks & where CARMAKE is.

We will also be setting up Twitter and Facebook means of crowd sourcing this most valuable information. We'll also try to create an AppForThat. If you have ideas on how this can be done really well, please share them with me. This is really important. We have to first know, then we can fix it. Alain

How a System for Shared Taxi Rides Could Transform New York City

Emily Badger, Mar 12, 2014 "The Senseable City Lab submitted a public records request to the New York City Taxi and Limousine Commission for all its data on those 150 million trips from 2011. They've mapped it in gorgeous, interactive detail here, as part of the HubCab project unveiled today. Even more usefully, though, they've made a startling calculation with all that data: nearly 80 percent of all of those trips could have been shared if passengers were willing to travel no more than three minutes out of their way.

If we assume that each trip carries one person (the data doesn't specify this), and that two trips can reasonably be combined in the back seat of one cab, that means a totally optimized taxi network in New York would produce 40 percent fewer trips, a similar drop in emissions, and far less traffic. That's 40 percent of all those trips that would just never need to be made." Read more

Hmm.. This is very interesting. I'm surprised that it is only 80%/40% fewer trips (the correct question is not fewer trips, but fewer cab miles, but that's a detail). What my students and I have done is to synthesize the equivalent of the NYC taxi data for a typical day in New Jersey and then simulated, as they did, the ride sharing potential. This is VERY interesting! Alain

How to Stop Worrying and Love the Robot That Drives You to Work

Based on the research of Adam Waytz "...Demonstrating competence and reliability will be crucial for manufacturers of self-driving cars. But Waytz and his colleagues were not merely interested in showing how much the occupant of an autonomous vehicle could be compelled to trust it. They wanted to examine something subtler: Would making the self-driving car more "person-like"—that is, giving it a voice, a name, and other psychologically individualized characteristics—make the rider trust it more or less than a self-driving car that was equally reliable, but offered no pretenses of being anything more than a mindless machine?.." Read more.

Hmmm... This is all very fundamental and very important. I would only add that this is only one aspect. The machine must earn our respect and confidence before we will trust it to "drive for us". That respect will need to be earned through the demonstration of competence. This places an enormous burden on the user interface. My initial reaction to the user interface in the S-Class is that they haven't even begun to scratch the surface. I need to be reassured that the 997 anti-collision, object and pedestrian recognition and lane centering are working properly. The current version cries out to me..."Hey, we got you to buy this stuff, but it really doesn't work. Try it if you dare, but don't rely on it."

I got the message. However, I think that the system is actually better than what they've divulged to me. (Maybe that's been their intention???). It is interesting that MB had no hesitation to superimpose on the rear camera image the lines delineating where the car might go in reverse at the current steering angle. But, they've refused to superimpose the objects seen by the object recognition system on the front facing camera's images. A rendering of the car at its current location, a simple depiction of the road edges that it is aware of and superimposing the objects that it is focused on avoiding would go a long way to earning the respect of the user. The data are there, the display environment is available, MB shows similar images in its advertising and demos, yet it chose not to make it available to its customers. That is truly unfortunate. Such a display would help the 997 earn some respect. It doesn't need to be on all the time, but it should be selectable at the user's discretion. Alain

Google's Self-Driving Car: What We've Done and What We Need

"...Andrew Chatham, principal software engineer for the Self-Driving Car Project at Google, Inc., led today's plenary session. In an energetic and dynamic presentation, Andrew explained Google's approach to developing its self-driving car and how ITE members can provide the input and information needed to help make the innovation a reality in the marketplace. Google's self-driving car relies on the company's extensive mapping of areas that the car is then equipped to navigate. Andrew explained how traffic engineers have already put much of the infrastructure in place that allows Google to create these maps, from medians to clearly marked lanes. Knowing where traffic cones are being placed in construction zones is an important element to enabling the self-driving vehicle to operate—an area in which Google could use ITE's assistance..."

Hmm... Andrew made a very compelling presentation followed by an extensive Q&A. The takeaway was an appeal to the conventional highway community to do some of the simple straightforward things to enable SmartDrivingCars to coexist harmoniously with conventional cars. Things such as making available reliable real-time data on the location and characteristics of construction zones and other temporary changes to the road infrastructure and its operation such as road closures and designated detours.

There's even more that ITE can and should do. ITE should take the lead at re-examining the rules of the road, roadway operational policies and traffic laws. Up to today, those rules, policies and laws were designed to serve human drivers having a range of cognitive and response capabilities. We are about to be joined on our roads with cars that have enhanced cognitive and response capabilities. Do they really need to adhere to some of the human-oriented rules? For example, the only reason cars must come to a complete stop at a stop sign is because there is insufficient opportunity for normal drivers to determine that it is safe to enter the intersection without coming to a complete stop. However, because SmartDrivingCars look both ways all the time, stopping may not be necessary. Also, what is the speed limit anyway, or maybe we should finally do away with the term and simply say, "you go above this and you get a ticket" then everyone including SDCs will operate on the same level playing field. This suggests that we need "truth in transportation". Rules should be rules and guidelines, guidelines etc. all in a context that SmartDrivingCar users have joined us on our roadway system. ITE is the ideal group to address the technical aspects of these issues. Alain

Jinan Foreign Language School's Transportation Club

During the opening session, ITE presented a certificate of recognition to the Jinan Foreign Language School in Jinan City, Shandong Province, China for its Transportation Club, which is focused on Smart Driving Cars. The Jinan Foreign Language School's Transportation Club is the first high school club chartered by ITE. Here to accept the award was the school's International Relations Representative, Jeremiah Liu, with a delegation of the club's members. The students will be discussing the club during today's student paper presentations Video

Autonomous Vehicles in California

California DMV has put up this go-to site in an effort to keep everyone informed on the issue of SmartDrivingCars in California. The video and/or transcript of recent (3/11/14) Public Hearing on Autonomous Vehicles Regulations, should appear shortly. Alain

California crafting driverless car regulations

By: Justin Pritchard, The Associated Press "...Among the complex questions officials wanted to unravel:

How will the state know the cars are safe?

Does a driver even need to be behind the wheel?

Can manufacturers mine data from onboard computers to make product pitches based on where the car goes or set insurance rates on how it is driven?..." Read more

Is This Ride-Sharing Smart Minivan the Future of Self-Driving Cars?

Alissa Walker, Gawker Media Mar 13, 2014,"...

Released this week at the Geneva International Motor Show, the Akka Link & Go 2.0 is an autonomous car that's designed to be social..." Read more

Half-baked stuff that probably doesn't deserve your time:

Volvo Car Group Tests Road Magnets For Accurate Positioning Of Self-Driving Cars

Hmmm.. This is less than half-baked. Volvo couldn't possibly be going back to this bad idea. Alain

Calendar of Upcoming Events:

2014 IEEE Intelligent Vehicles Symposium

June 8 - 11, 2014, Dearborn, Michigan, USA

Sponsored by the IEEE Intelligent Transportation Systems Society

Recent Versions of:

March 7, 2014

http://www.apple.com/ios/carplay/

Hmmm. Just what we need in our cars, an easier way to be distracted from driving. Maybe this should be called "Apple CarCrash".

Alain

New Jersey Rail Ridership Opportunities If a Taxis Are Available...

A.L. Kornhauser, et al March, 2014 ".... Examined are the rail ridership opportunities that NJ Transit (NJT) might enjoy if autonomous Taxis (aTaxis) were available to readily bring customers to or from its rail stations. Such collection and distribution services would not only shower NJT with new customers but they would enable NJT to redevelop its parking facilities to more profitable "Transit Village" uses. Read more

February 28, 2014

Automated cars are coming, but has anyone grasped their implications?

Issue 641 Feb 2014 Scott Le Vine, Centre for Transport Studies: "...Autonomous cars, driverless cars, automated cars – whatever you call them (and yes, it does make a difference) they are the hot topic in transport today. Carmakers are investing millions of pounds in research & development, mock towns are being built to test automation concepts, fact-finding hearings are taking place, laws are being passed – even international treaties are being re-opened..." Read more This is a lead/summary of the next item. Alain

February 21, 2014

Google's Project Tango is a smartphone with sensors to map the world around you

BY Sarah Silbert @sarahsilbert February 20th, 2014 at 2:25 PM ET "Google's just announced Project Tango, a 5-inch Android smartphone prototype equipped with Kinect-like 3D sensors and other components to track motion and map your surroundings. ..." Read more

February 14, 2014

Imagine: A World Where Nobody Owns Their Own Car

Eric Jaffe Jan 13, 2014 "...The problem with buying the drink today, says Kornhauser, is that the labor cost of on-demand taxi service is enormous. As a result, we buy the bottle just in case we want a drink. Driverless cars change the whole equation. Read more

February 9, 2014

Volvo-world first public pilot for driverless cars

I prefer to call this: Volvo's "Texting Machine" concept video. It demonstrates very clearly that it is the "NHTSA Level 3" feature that will be needed to make the car as good as transit; otherwise everyone is simply going to take transit. Who cares about far the walk, how long the wait and the ride and the crowded conditions. I can be in my own world doing what I want to do during that whole process. Without Level 3 I have to be off the grid focused on keeping the hulk between two white lines and not running into things. Transit wins. Play video Alain

February 4, 2014

U.S. Department of Transportation Announces Decision to Move Forward with Vehicle-to-Vehicle Communication Technology for Light Vehicles

European Commission agreed upon the demonstration schedule Adrianon Alessandrini

By far the BEST Super Bowl Commercial Play video Alain

January 26, 2014

What Will Happen to Public Transit in a World Full of Autonomous Cars?

by Emily Badger, Jan 17, 2014 "...The great promise of autonomous cars is not that we could each own one in our own driveway – the 21st century's version of owning your own Model T, or your own color TV, or your own bulky Macintosh – but that no one would need to own one at all..." Read more In fact read the whole thing as well as the 266 Comments.

January 17, 2014

Autonomous Vehicle Technology: A Guide for Policymakers

by James M. Anderson, Nidhi Kalra, Karlyn D. Stanley, Paul Sorensen, Constantine Samaras, Oluwatobi Oluwatola

This report is excellent

January 6, 2014

Self-Driving Cars Moving into the Industry's Driver's Seat

Jan. 2, 2014 "Accident rates will plunge to near zero for SDCs, although other cars will crash into SDCs, but as the market share of SDCs on the highway grows, overall accident rates will decline steadily". Self-driving cars (SDC) that include driver control are expected to hit highways around the globe before 2025 and self-driving "only" cars are anticipated around 2030, according to an emerging technologies study on Autonomous Cars from IHS Automotive, driven by Polk. Read more

December 27, 2013