imap://alaink@exchangeimap.princeton.edu:993/fetch%3EUID%3E/INBOX%3E3022058?part=1.2&filename=hejedgabmgkdglfj.png

Saturday, November 23, 2019

http://SmartDrivingCar.com/7.48-Florida-112319

  Self-driving car capital? One senator thinks it can be Florid

R. Wile, Nov 22, "Sen. Jeff Brandes (R-St. Petersburg) had just finished serving in the Army, and was looking to make a name for himself in Tallahassee as a junior representative. He came across a talk given by the founder of Google’s driverless car project.

He quickly realized the potential of self-driving cars to transform many aspects of daily life. Ever since, he has made it his mission to turn Florida into what he calls “an angel investor” in automation policy. “We want to have policies in place for this technology to flourish,” Brandes said in an interview at the 7th Annual Florida Automated Vehicles conference in Miami, which concluded Friday.

Brandes has drawn headlines in the tech community for filing legislation allowing virtually any automated vehicle on Florida’s roads; this summer, he helped make Florida one of the first states to make AVs without a human back-up safety driver street legal.

Among the state’s advantages Brandes points to that he believes makes it ideal for AV companies: no snow, which makes lane markings more visible. That also means less road construction in general…."  Read more  Hmmmm…  Congratulations Jeff!!! It was a great AV Summit and congratulations on creating such a Welcoming environment and intelligently shaping the birth of this technology.  What you’ve done is enable Florida to begin to enhance mobility and the quality of life for all in Florida  and especially those who can most benefit from these mobility machines.  It was most impressive to witness the enthusiasm for nurturing the many aspects of this technology from Florida’s Governor, Miami’s Mayor, Fl DoT’s Commissioner, the heads of the toll road authorities, planning agencies and educational institutions.  Most impressive was Ford’s comment that their autonomousTaxi efforts are focused on developing driverless technology and intend to operate it to deliver Mobility-as-a-Service in Florida, rather than sell the technology to individual consumers.  I applaud that approach and hope that Ford will look to also bringing some of those vehicles to New Jersey so that we can begin to reap the benefits of this technology.  What you’ve accomplished in Florida is THE "best practice" for us to emulate in New Jersey.  Congratulations.   Alain

  Smart Driving Cars Podcast Episode 133

F. Fishkin, Nov.23, "Florida’s Autonomous Vehicle Summit shows what a state can do to create a welcoming atmosphere. That, plus, Tesla’s Cybertruck, Ford, Waymo and more in the latest Smart Driving Cars with Princeton’s Alain Kornhauser and co-host Fred Fishkin.."  Just say "Alexa, play the Smart Driving Cars podcast!".  Ditto with Siri, and GooglePlay …  Alain

  Tesla needs safeguards to prevent drivers from sleeping on ‘Autopilot’: U.S. senator

D. Shepardson, Nov 20, "Democratic U.S. Senator Ed Markey asked Tesla Inc (TSLA.O) on Wednesday to disable its “Autopilot” driver-assistance system until it installs new safeguards to prevent drivers from evading system limits that could let them fall asleep…  Markey cited a local news report that said a driver had fallen asleep behind the wheel as a Tesla drove 14 miles on Autopilot. Other unconfirmed videos on social media appear to show drivers sleeping behind the wheel of Tesla vehicles.  … Democrats keep falling for fake news…  “…  That’s not safe. Somebody is going to die because they can go to YouTube as a driver – find a way to (get around safety requirements),” Markey said. “We can’t entrust the lives of our drivers and everyone else on the road to a water bottle.”  … During an average day about 100 people die on our roadways, largely because of misbehavior by drivers of cars other than Tesla.  Is the good Senator going to ask the makers of those cars to install new safeguards to prevent misbehavior such as speeding, drinking, tailgating, …,  in addition to falling asleep?  Certainly it is trivial for automakers to limit speeds, to limit tailgating, …  Why hasn’t the NTSB stepped up to demand those limits?  "… A series of crashes involving Autopilot has prompted U.S. investigations and criticism from the National Transportation Safety Board (NTSB).   In September, the NTSB said the Autopilot design was a key factor in a January 2018 crash of a Model S into a parked fire truck on a highway in California. …Excuse me NTSB… Avoidance of that crash required the driver, or the Automated Emergency Braking system, to have have its foot on the brake, not hands on the wheel… "The system’s design “permitted the driver to disengage from the driving task” in the 2018 crash and allowed him to remove his hands from the wheel for nearly all of the last 14 minutes of the trip, it said.…" Read more  Hmmmm…. Shame on you NTSB for not recognizing the designed-to-fail aspects of Automated Emergency Braking system in these crashes.  These AEB are explicitly designed to disregard stationary object in the lane ahead because their inability to reliably discern a real stationary object ahead rather than a phantom stationary object.  Consequently, rather than responding to phantom stationary objects, they simply assume that a stationary object ahead can either be readily passed under if it is real or is a phantom object.  Consequently, that object can be, and is,  be disregarded.   But, what if a real object in the lane ahead can’t be passed under?  Ouch!!! See also Fred Lambert’s take. Alain

  A call for clarity on self-driving terminology

Blog, Nov 15, "Driverless or self-driving — autonomous or automated — automation or autonomy — these terms are often used interchangeably. That’s regrettable because they don’t all mean the same thing.

Automation, for example, describes the presence of automatic equipment that’s used as one part in an overall process. Autonomy, on the other hand, confers the idea that a system is governing itself and all actions. Self-driving implies that a vehicle is being driven without a human involved, while driverless might indicate no one or nothing is in control at all.

Those terms are just the start. Add the likes of semi-autonomous, partially self-driving and driver-assist feature, and the language landscape becomes more cluttered. Throw in industry jargon such as "Level 2 Plus" and Elon Musk’s special dictionary entries for "Autopilot," "feature complete" and "fully self driving," and this confounding mix is complete.

It’s time to clean up the mess…." Read more  Hmmmm…  Amen!!! For:

  • "Do it yourself mobility":  Safe-driving (Hands on wheel, feet on pedals all the time (anti-lock brakes and electronic stability control on steroids)), & Self-driving (Hands off wheel, Feet off pedals some of the time BUT Brain on road all the time.  (Today’s Tesla give or take))
  • "Done for you mobility":  Driverless (Today’s elevator ( the human-machine experience aspect of elevators)). 

Alain

  Sustainability Prospects for Autonomous Vehicles

G. Martin, Nov  2019, "The Autonomous Vehicle (AV) has been strongly heralded as the most exciting innovation in automobility for decades. Autonomous Vehicles are no longer an innovation of the future (seen only in science fiction) but are now being road-tested for use. And yet while the technical and economic success and possibilities of the AV have been widely debated, there has been a notable lack of discussion around the social, behavioural, and environmental implications. This book is the first to address these issues and to deeply consider the environmental and social sustainability outlook for the AV and how it will impact on communities. Environmental and social sustainability are goals unlike those of technical development (a new tool) and economic development (a new investment). The goal of sustainability is development of societies that live well and equitably within their ecological limits. Is it reasonable and desirable that only technical and economic success comprise the swelling AV parade, or should we be looking at the wider impacts on personal well-being, wider society, and the environment?

The uptake for AVs looks to be lengthy, disjointed, and episodic, in large measure because it faces a range of known unknown risks. This book assesses the environmental and social sustainability potential for AVs based on their prospective energy use and their impacts on climate change, urban landscapes, public health, mobility inequalities, and individual and social well-being. It examines public attitudes about AV use and its risk of fostering a rebound effect that compromises potential sustainability gains. The book concludes with a discussion of critical issues involved in sustainable AV diffusion." Read more  Hmmmm…  Very worth reading because there has been…"… a notable lack of discussion around the social, behavioural, and environmental implications…."  Alain

Is Majoring in English Worth It?

Opinion: W. McGurn, Sept 9, "… A just-released Bankrate study won’t make you feel any better. In a ranking of 162 college majors by median income and unemployment rate, English majors landed among the bottom dwellers, at 132. At $47,800 in median income, they did better than those in drama ($35,500) or fine arts ($37,000), but they earned less than half as much as someone who majored in, say, electrical engineering ($99,000)…. A 2017 MarketWatch story was blunt: It called English “the most regretted college major in America.”…

“The English major was once a guarantor of effective, formal writing skills and the ability to comprehend and analyze the complex thoughts found within centuries of brilliant and challenging poetry and prose,” he told Campus Reform. “Its decline into the epiphenomena of popular culture and identity politics is a self-inflicted wound that has rocked its credibility.”

In other words, what’s on offer today isn’t your father’s English degree. An ACTA study of English programs reports that 48 of 52 top schools (as ranked by U.S. News & World Report) allow English majors to graduate without ever having taken a course on Shakespeare. In the past ACTA has also highlighted studies showing that the average grad, even those from prestigious flagship universities, shows little or no improvement in critical thinking for having gone to college….

So why have the sciences kept their integrity while the humanities haven’t? Mr. Pidluzny suggests it’s because the costs of a dumbed-down STEM degree can be both more obvious and more consequential….

“The university can’t get away with not teaching engineering students differential equations because we’d then have collapsing bridges all over the place,” he says.   “But for an English major who studies Harry Potter instead of Chaucer, or spends his time on gender theory instead of reading great literature, the costs aren’t as obvious—except to the graduate who only later realizes he never developed the keen analytical mind and precise style of writing college was supposed to cultivate.”"  Read more  Hmmmm… From the previous post we are learning/realizing that the easy parts of this "mobility revolution" are essentially solved,  the technology parts.   Now comes the hard part…  the social, behavioural, and environmental implications…".  To address those aspects and actually deliver value to society we are going to need the critical thinking and effective writing of "the English major".  Alain

  Uber plans to start audio-recording rides in the U.S. for safety

F. Diddiqui, Nov 20, "Uber plans to record audio during rides in the United States as part of a new security feature, in its latest push to protect riders and drivers amid rising safety concerns.

The feature, which is first to be piloted in some Latin American cities next month, allows users to opt in to activate an audio recording on any trip or all trips, according to internal communications viewed by The Washington Post and confirmed by Uber. In markets where it’s available, users would probably be given a blanket warning that trips are subject to recording — and that the feature will be active in their market. Riders and drivers will not be able to listen back.

“When the trip ends, the user will be asked if everything is okay and be able to report a safety incident and submit the audio recording to Uber with a few taps,” according to an email written by an Uber executive and obtained by The Post. “The encrypted audio file is sent to Uber’s customer support agents who will use it to better understand an incident and take the appropriate action.”…

“We have taken a position that whenever you are in an Uber, the feeling that we want both parties to have is ‘the lights are on.’ " he added. “That leads to safer interaction on the platform.”… "  Read more  Hmmmm… This is very tricky.  Privacy issues come into play and well as other legal issues, but dealing with the sociology of the ride is very challenging.  Transit companies have installed cameras in buses to address this issue.  Lots of data suggesting that it works well.  Cameras are in many/most? elevators.  Dealing with misbehavior is simply not easy, but must be addressed.  Copuld use some good "English Majors" here.  Alain

  Board Meeting : Collision Between Vehicle Controlled by Developmental Automated Driving System and Pedestrian

Nov 19,   "Abstract  …This may well be the most important document in this proceeding.  While it states…"The ADS detected the pedestrian 5.6 seconds before impact. Although the ADS continued to track the pedestrian until the crash, it never accurately classified her as a pedestrian or predicted her path.  … So what??? Why does it need to "…accurately classify" objects that are in its lane ahead????  Why does it need anything more that a couple of position values to "predict a path"???  Kinematics are way better than Duhhhh.  The object is in the lane ahead.  In fact it tracked the object and did have a positions over time that readily gave it a kinematic path.  Why didn’t it slow down?  The code is/was terrible.  The framework is/was terrible.  Everything else is/was irrelevant.  The fundamental safety issue is that it was terrible code!  Not slowing down when objects are detected in the lane ahead, or kinematically to be in the lane ahead, is just being totally irresponsible. The findings call out the pedestrian’s unsafe behavior (# 3) before the system’s design that "precluded braking in emergency situations" (#5)… That is totally backwards.  These systems need to be cautious and not even allow the vehicle’s to evolve into an "emergency situation". 

 

The fault of this crash is totally on Uber’s fundamentally flawed code.  Recommendations should have been ordered by most important.  #1 is Uber’s fundamentally flawed code.  Listing that as #6 seems totally irresponsible.  C’mon NTSB

 

Opening Statement – Robert L. Sumwalt, Chairman

Crash Overview – David Pereira, Investigator-in-Charge

Pedestrian and Vehicle Operator – Rafael Marshall, PhD … Sorry, but this is really bad!  Totally irrelevant.  These systems are supposed to be designed so that they can operate without an attendant and not hit anybody/anything in their lane ahead.  If this is only "for information" and "not relevant" (which it is "not relevant"), then this overview should state that in the beginning!

 

Managing Risk of ADS Testing – Ensar Becic, PhD

Uber ATG Operations – Michael Fox … Sorry, but this is also bad.  It is tracking an object that is clearly moving across is path, yet it doesn’t even bother to slow down a little.  What if the abject was a moose?  Would it have been OK to hit it? (Hitting moose @ 40 mph likely kills occupant of car.  Why doesn’t the summary point out that the code was bad?  I guess that it is so obvious that there is no need to make it explicit???

Testing of Automated Vehicles – Ensar Becic, PhD  … Last Summary is "Lessons learned from this investigation" would,be a good last summary item if the focus was on the faults in the code and what led to the code being so poorly written.  Why is it that the system disregarded objects unless they were classified?  What led to that being acceptable logic in the code?  Those questions don’t seem to have been pursued?  Alain
Closing Statement – Robert L. Sumwalt, Chairman 

Port Authority Plans to Test Self-Driving Buses

P. Berger, Nov 18, "Transportation officials are looking to self-driving technologies to help New Jersey commuters reach New York City in one of the nation’s busiest bus lanes.  … It is the busiest bus-lane in the world… 

The Port Authority of New York and New Jersey plans to test the technologies over the next year with a view to easing traffic in its exclusive bus lane, which runs for 2.5 miles between the New Jersey Turnpike and the Lincoln Tunnel.

The eastbound bus lane occupies one westbound lane of New Jersey Route 495 each weekday between 6 a.m. and 10 a.m. for buses heading to the Port Authority Bus Terminal in Manhattan. It currently runs at capacity, carrying about 1,850 buses each morning.

The autonomous technology push isn’t aimed at replacing drivers, Port Authority officials say. Instead, it will help drivers merge into the bus lane, better space out vehicles and reduce or avoid collisions…."  Read more  Hmmmm… Since there have been essentially zero collisions in the XBL since its inception by NJ DoT Commissioner John C. Kohl in 1971, it is unlikely that this technology "will reduce or avoid collisions" .  Human drivers have done a really good job over the years. 

The real reason to do this is to substantially increase the capacity of the XBL, which today is limited to about 700 buses per hour.  That capacity limit is due to the relatively large variation in the spacing of buses as they traverse the XBL due to variations in human driver behavior.  Implementing adaptive cruise control driver assistance to each bus would substantially reduce the variability in the headway (separation) between buses.  This would yield a substantial increase in the nominal throughput of the XBL.  Increases of 50% to 1,050 buses per hour, and possibly more are realistically achievable.   Such throughput increases would enable serving an additional 17,500 commuters during the peak hour.  This is essentially the same amount that would be served if two new rail tunnels were to be built at a now estimated cost of $12.7B as part of a $30B Gateway project.  Cost of retrofitting the 3,000 buses that regularly use the XBL at what couldn’t be more than $10k/bus is $30M,.  Even if it was 10x @$300M it would be dirt cheap.  (Of course, the mid-town bus terminal needs to be able to "eat" 1050 buses per hour and not just the 700 that it currently struggles to serve.But a new bus terminal is being designed.  Will the Port Authority design it to serve 1,050+ buses per hour or do what is done today?  Unfortunately, that may well be an open question? 

By the way, Lou Pignataro and I proposed such an operational test in 1995.  Needless to say, we were a little early.  Better late than never?  🙁  Alain

Self-driving cars: Promise or Peril?

H. Fairman, Nov 22, "Many have heard of the promise, potential, and peril of self-driving cars, and trucks and buses. They can free us from the tedium of commuting, and of deliveries, and bus routes. They can reduce the use of fossil fuels and reduce traffic. Occasionally some technological error or inadequacy has led to crashes, but solutions are within reach. What then is the real outlook for autonomous vehicles? Are they a dream intriguing mostly engineers or might they offer real life improvements to the rest of us?

Someone, who has investigated, imagined, and tested the advances that autonomous vehicles may offer, spoke to the recent local Sierra Club gathering at Mercer County Community College…."  Read more  Hmmmm…  🙂   While we are at it Ken Pyle taped my presentation at the PodCarCity Conference in San Jose earlier this month. Also, Ken Pyle has been active trying to improve mobility in the Bay Area.   Alain

  FCC Spectrum Proposal Spotlights Automaker Technological Divide

T. Shields, Nov 20, "U.S. Federal Communications Commission Chairman Ajit Pai proposed reallocating to mobile devices airwaves long assigned to vehicle safety while preserving some of the spectrum for carmakers planning to deploy new technology.

“We want to move on from something we’ve tried for a long time that wasn’t working, and open the door to new and exciting opportunities,” Pai said in a speech at a Washington event. “After 20 years of seeing these prime airwaves go largely unused, the time has come for the FCC to take a fresh look.”…

Pai set a Dec. 12 vote on his proposal, which would commence a months-long comment period on giving Wi-Fi gadgets access to 60% of the airwaves reserved for auto safety in 1999. Automakers would retain use of the remainder…."  Read more  Hmmmm…  This may finally get resolved.  Alain

  Tesla already received 146,000 Cybertruck pre-orders worth over $8 billion

F. Lambert, Nov. 23, "Elon Musk announced that Tesla already received 146,000 Cybertruck pre-orders just over a day after they unveiled the electric pickup truck to mixed reviews.

 
The Tesla Cybertruck has been polarizing, to say the least.  Many people didn’t like the radical design of the electric pickup truck, but most agree that the specs and price are impressive.

Right after the unveiling on Thursday night, Tesla started taking reservations for the truck with a $100 deposit.  Today on Twitter, CEO Elon Musk confirmed that the automaker received 146,000 Cybertrucks reservations after the first day:…."  Read more  Hmmmm… No comment.   Alain

  Two Peas in a PodCar – Brief Overview of Podcar City 2019

K. Pyle, Nov. 13, "It is important to build autonomous transportation systems that meet the needs of individuals while reducing congestion and pollution by sharing vehicles is the message from Princeton Professor Dr. Alain Kornhauser. Reflecting on day 1 of the Podcar City 2019 Conference, Kornhauser emphasized that whether these systems are on existing roadways or, in some cases, guideways, automating the driver promises to reduce the cost of mobility and improve safety. …"  Read more   Hmmm… Thank you Ken.  Alain

 


Half-Baked

  Swiveling seats and sound ‘bubbles’: What the inside of cars will soon be like

A. Poliak, Nov 22, "Ever since cars first hit the road, the basic interior layout has remained the same, with everything built around the driver: a driver’s seat, driver’s steering wheel and driver-controlled power and brake mechanisms. But that’s only until autonomous vehicle technology completely eliminates the need for a driver. Cars can now be designed for a new generation of driverless vehicle transportation.   …OK…

It’s true that safety concerns about autonomous cars are an issue in the industry right now. But the presumption is that, thanks to an array of emerging and advanced technologies, an autonomous vehicle will eventually prove to be safer than a vehicle with a human driver. …OK…

The overall idea is that an autonomous vehicle cabin can become a mobile living room…."  Read more   Hmmmm…  Now wait a minute.  Apparently you didn’t hear what John Rich, CEO, Ford Autonomous Vehicles, said at the Florida AV Summit a couple of days ago…. Ford will operate the driverless vehicles that it makes.  It will be a very long time before Ford sells driverless cars to individuals…

What is being realized is that it is one thing to make a car that safely drive itself.  It is another thing altogether to ensure that a high enough percentage of individuals who purchase these vehicles operate them in exclusively safe enough "Operational Domains"  that they remain safe enough to not ruin it for everyone else.  Just look at how individuals abuse the use of what is a fundamentally safe entity… the conventional car.  94% of the crashes are caused by human misbehavior, or so it is said.  These fundamentally safe devices become incredibly safe devices if you take human misbehavior out of the loop.  Look above at what one Congressman wants to do to AutoPilot.   It is not AutoPilot that is unsafe; it is AutoPilot that is put into the hands of misbehaving individuals.  Just wait to see what happens with StupidSummon.  It only takes just a very very few to ruin it for everyone else. 

I suspect that all Driverless car developers have begun to understand that it is the removal of individual ownership of this technology that has led us to this Gartner Trough of Disillusionment.  What remains is the opportunity for this technology to be operated at scale by entities that will ensure, if only for their own survival, that these vehicles will be operated without misbehavior essentially all the time.

That means that the business case for Driverless is as Mobility Machines, or, I hesitate to say, Transit (defined as the provision of "done-for-you-mobility" to anyone.   This business case has as its objective to get as many people as possible from A to B safely and comfortably between as many As and Bs as possible.  Given that objective, then some of these vehicles may be designed to attract patronage by the 1%ers, as is discussed in this article.  However, most of the others will be designed to appeal to and be affordable by the 99%ers. They are the customers that bring scale. Serving the 99%ers and especially the 80 to 20%ers requires a substantially different mind-set than is presented here.  Alain

  Andrew Yang: Yes, Robots Are Stealing Your Job

A. Yang, Nov. 11, "My vision calls for new top-line measurements that take into account indicators like: health and life expectancy, mental health, substance abuse, childhood success rates, average income, environmental quality, retirement savings, labor force participation and engagement, infrastructure and homelessness…."  Read more   Hmmm… I agree.   But machines have all the jobs that we don’t want. They do the dirty, boring, monotonous parts. When it comes to self-driving trucks the automation does the simple tedious parts  but still needs adult/professional-driver supervision.  That partial automation substantially enhances the safety, while reducing stress and anxiety associated with truck driving.  It may also enable an increase in hours-of-service regulations.  Thus truckers will be healthier, happier and better able to feed their families.  Driverless trucks, at best, will slowly gain traction in niche markets primarily to address driver shortages.   Alain


 C’mon Man!  (These folks didn’t get/read the memo)


Simply ClickBait


StupidSummon


Corrections


Calendar of Upcoming Events:

 

4th Annual Princeton SmartDrivingCar Summit

evening May 19 through May 21, 2020

Princeton University

Princeton, NJ


On the More Technical Side

http://orfe.princeton.edu/~alaink/SmartDrivingCars/Papers/


%d bloggers like this: