SmartDrivingCar.com/6.25-SoftBank-060318
25th
edition of the 6th year of SmartDrivingCars
F. Fishkin, June 3, "Softbank makes a multibillion dollar investment in GM's self driving company and Google's Waymo orders more than 60 thousand additional Chrysler minivans for a self driving fleet. Where does Uber fit in? Princeton University's Alain Kornhauser dives in along with co-host Fred Fishkin in Episode 42 of the Smart Driving Cars podcast. Listen and subscribe."
Hmmmm.... Now you can just say "Alexa, play the Smart Driving Cars podcast!" . Ditto with Siri, and GooglePlay. Alain 3rd
Annual
Princeton SmartDrivingCar
Summit
evening May 14 through May 16, 2019
Save the Date; Reserve
your Sponsorship
Photos
from 2nd Annual Princeton SmartDrivingCar
Summit
Program
& Links to slides from 2nd Annual
Princeton SmartDrivingCar Summit
F. Fishkin, May 31, "Artificial Intelligence may be able to drive better than humans most of the time....but is that good enough? Join Princeton University's Alain Kornhauser and Co-host Fred Fishkin for Episode 41 of the Smart Driving Cars Podcast. More on the latest from Uber, Tesla and Nuro. Listen and subscribe."
F. Fishkin, May 17, "How close is California to giving the green light to driverless testing on public roads? Deputy DMV Director Bernard Soriano joins Alain Kornhauser, Fred Fishkin and guest Michael Sena on Episode 39 of the Smart Driving Cars Podcast. And we review some highlights of the just concluded 2nd annual Princeton Smart Driving Car Summit. Listen and subscribe!"
F. Fishkin, May 10, "The continuing Uber crash investigation, Waymo and Ohio rolls out the welcome mat for the testing of self driving cars. All that and more in Episode 38 of the Smart Driving Cars podcast. This week Princeton's Alain Kornhauser and co-host Fred Fishkin are joined by Bryant Walker Smith of the University of South Carolina and Stanford. Tune in and subscribe!"
F. Fishkin, Apr 26, "Getting SmartDrivingCar companies to share their data on safety. It's a move that could benefit all says Princeton University Professor Alain Kornhauser in the latest Smart Driving Cars Podcast. He joins co-host Fred Fishkin...to chat about the move by Voyage. Also...Tesla, Waymo and more..
F. Fishkin, Apr 4, " Waymo is making it real! In Episode 33
of the Smart Driving Cars Podcast, hosts Fred
Fishkin and Princeton's Alain Kornhauser are
joined by Michael Sena,
publisher of The Dispatcher newsletter. Take a
deep dive into Waymo's
deals with Jaguar and talks with Honda..
Tesla, Volvo, Uber and Ambarella.
And the Princeton Smart Driving Car Summit is
coming up! "
F. Piekniewski, "Deep learning has been at the forefront of the so called AI revolution for quite a few years now, and many people had believed that it is the silver bullet that will take us to the world of wonders of technological singularity (general AI). ...We have now mid 2018 and things have changed. ..By far the biggest blow into deep learning fame is the domain of self driving vehicles ..
But by far the biggest prick punching through the AI bubble was the accident in which Uber self driving car killed a pedestrian in Arizona. From the preliminary report by the NTSB we can read some astonishing statements:...
Aside from general system design failure apparent in this report, it is striking that the system spent long seconds trying to decide what exactly is sees in front (whether that be a pedestrian, bike, vehicle or whatever else) rather than making the only logical decision in these circumstances, which was to make sure not to hit it. ...
In fact if there is anything at all we learned from the outburst of deep learning, is that (10k+ dimensional) image space has plenty enough spurious patterns in it, that they actually generalize across many images and make the impression like our classifiers actually understand what they are seeing. Nothing could be further from the truth, as admitted even by the top researchers who are heavily invested in this field....
the problem is that the input space is
incredibly high dimensional, while the action
space is very low dimensional. Hence the
"amount" of "label" (readout) is extremely small
compared to the amount of information coming
in..." Read
more Hmmmm....
Very interesting. We still have an
awful lot to do. See also,G.
Marcus, below. Alain
KMay 24, "About 9:58 p.m., on Sunday, March 18,
2018, an Uber Technologies, Inc. test vehicle,
based on a modified 2017 Volvo XC90 and
operating with a self-driving system in computer
control mode, struck a pedestrian on northbound
Mill Avenue, in Tempe, Maricopa County, Arizona.
...The vehicle was factory equipped with
several advanced driver assistance functions by
Volvo Cars, the original manufacturer. The
systems included a collision avoidance function
with automatic emergency
braking, known as City Safety, as well as
functions for detecting driver alertness and
road sign information. All these Volvo functions
are disabled when the test vehicle is operated
in computer control..."
Read more Hmmmm....
Uber must believe that its systems are
better at avoiding Collisions and
Automated Emergency Braking than Volvo's. At least
this gets Volvo "off the hook".
"...According
to data obtained from the self-driving
system, the system first registered radar
and LIDAR observations of the pedestrian
about 6 seconds before impact, when the
vehicle was traveling at 43 mph..." (= 63
feet/second) So the system started
"seeing an obstacle when it was 63 x
6 = 378 feet away... more than a
football field, including end zones!
"...As the
vehicle and pedestrian paths converged, the
self-driving system software classified the
pedestrian as an unknown object, as a
vehicle, and then as a bicycle with varying
expectations of future travel path..." (NTSB:
Please tell us precisely when it
classified this "object' as a vehicle
and be explicit about the
expected "future
travel paths." Forget
the path, please just tell us the
precise velocity vector that Uber's
system attached to the "object", then
the "vehicle". Why didn't the the Uber
system instruct the Volvo to begin to
slow down (or speed up) to avoid a
collision? If these paths (or velocity
vectors) were not accurate, then why
weren't they accurate? Why was the
object classified as a
"Vehicle" ?? When
did it finally classify the object as a
"bicycle"? Why did it change
classifications? How often was the
classification of this object done.
Please divulge the time and the outcome
of each classification of this object. In the tests that
Uber has done, how often has the
system mis-classified an object as a
"pedestrian"when the object
was actually an overpass, or an
overhead sign or overhead
branches/leaves that the car
could safely pass under, or was
nothing at all?? (Basically,
what are the false alarm
characteristics of Uber's
Self-driving sensor/software
system as a function of vehicle
speed and time-of-day?)
"...At 1.3 seconds before impact, (impact speed was 39mph = 57.2 ft/sec) the self-driving system determined that an emergency braking maneuver was needed to mitigate a collision" (1.3 x 57.2 = 74.4 ft. which is about equal to the braking distance. So it still could have stopped short.
"...According to
Uber, emergency braking
maneuvers are not enabled while
the vehicle is under computer
control, to reduce (eradicate??)
the potential for erratic
vehicle behavior. ..."
NTSB: Please describe/define potential
and erratic vehicle
behavior Also
please uncover
and divulge
the design
& decision
process that
Uber went
through to
decide that
this risk
(disabling the
AEB) was worth
the reward of
eradicating "
"erratic vehicle behavior". This
is
fundamentally
BAD design.
If the Uber
system's false
alarm rate is
so large that
the best way
to deal with
false alarms
is to turn off
the AEB, then
the system
should never
have been
permitted on
public
roadways.
"...The vehicle operator
is relied on
to intervene
and take
action. " Wow! If Uber's
system
fundamentally
relies on a
human to
intervene,
then Uber is
nowhere near
creating a
Driverless
vehicle.
Without its
own Driverless
vehicle Uber
is past "Peak
valuation".
K. Pyle, May 9, "Safety and, as importantly, the perception of safety could be the pin that pricks the expectations surrounding the autonomous vehicle future. Recognizing the importance of safety to the success of this still nascent industry, autonomous taxi start-up, Voyage, recently placed their testing and reporting procedures in an open source framework. ...Oliver Cameron, Voyage Co-Founder and CEO, is excited to see participation and says, “We can’t wait to have all of these contributions from companies from around the world; contribute to build the actual standard in autonomous safety.” Read more, Hmmmm.... See the video that was played at the Princeton SDC Summit which generated substantial positive discussion at the Summit. See also full length video. Alain
A. Efrati, May 7, "Uber has determined that the
likely cause of a fatal collision involving one
of its prototype self-driving cars in Arizona in
March was a problem with the software that
decides how the car should react to objects it
detects, according to two people briefed about
the matter." Read
more Hmmmm....Uber
is "leaking" this??? Is this Spin?
Fake News?? I guess Uber doesn't
believe in transparency here. Where is
the official public statement of
reassurance???
"The car’s sensors detected the pedestrian, who
was crossing the street with a bicycle,
Hmmmm....Pretty much what I wrote on
March 24, the sensors "Saw something"
... but
Uber’s software decided it didn’t need to react
right away. ..."right
away" is Fake News. It never
reacted. Uber has not released any data
indicating that the software ever
reacted. "That’s
a result of how the software was tuned." ...That was a
major "tuning" faux pas. What is being
divulged here is that Uber's software
never became confident enough that what
it was seeing was something that it
should not hit and, at least, begin to
apply the brakes (or swerve, or ???).
Even the driver in the video recognized
that the object should not be hit a
split second before the crash. So the Problem
is not "tuning" it is outright "fuhgeddaboudit"
Like other autonomous vehicle systems, Uber’s
software has the ability to ignore “false
positives,” or objects in its path that wouldn’t
actually be a problem for the vehicle, such as a
plastic bag floating over a road.... Is Uber
suggesting that its software can't tell
the difference between a plastic bag
floating over the road and a pedestrian
with a bicycle, even after seeing the
object 30 to 60 or more times over the 3
or more seconds that the object was in
view? If this isn't Fake News then
Uber is hopelessly far behind... In
this case, Uber executives believe the company’s
system was tuned so that it reacted less to such
objects." It didn't
react at all!...
But the tuning went too far, and the car didn’t
react fast enough, one of these people said.... ... It
didn't react at all! If this wasn't so
important I'd put it in C'mon man.
"False positives" are
the symptom, not the problem. The problem
is Uber's system design and operational
policy. Uber system designers knew that the
sensors under certain conditions reported
"false positives" (were "spooked"). One of
those conditions was possibly the
combination of "is the closing speed = car's
current speed" AND "is the car's current
speed greater than 30mph." In situations in
which both are true, then Uber's "tuning"
is outright
"fuhgeddaboudit".
This "tuning" effectively turns-off Uber's
sensors to detecting anything that is
stationary or moving across its lane ahead.
If Uber has understood this, then Uber
would/should have ...
1. limited the operation of its cars to speeds under 30 mph,
2. limited the
operation of its cars at speeds greater than
30 mph only to roadways where
pedestrians are extremely unlikely to cross,
and
3. focus on
substantially improving its ability to
interpret its sensor data so that the false
alarm rate becomes so small that false
alarms are tolerated throughout Uber's
operational domain.
..."Meanwhile, the human driver behind the
wheel, who is meant to take over and prevent an
accident, wasn't paying attention in the seconds
before the car hit..." ...I
think that this is a cheap shot against the
driver. I suspect that this car had a
screen that displayed the real-time status
of the automated driving system. I would
not be surprised if that screen was mounted
below the radio and that the driver was
actually monitoring the operation of the
automated driving system prior to the
crash. Why this display wasn't on the dash
so that the driver's peripheral vision could
remain on the road ahead when the driver was
monitoring the performance of the system is
a question Uber should answer,... if it had
any interest in being transparent.
Another question that
Uber could be asked: Why didn't the
monitoring system warn the driver that it
was "seeing something" and ask the driver
to look to see if it should be "saying/doing
something".
Since it doesn't look
like Uber is going to really divulge
anything, it is incumbent on the NTSB to dig
deeply into this "false alarm" issue.
Disregarding "false positives" in order to
circumvent a little passenger/customer
discomfort enables "false negatives" which
kill people. Not pretty!
A. Madrigal, Mar 28, "On Tuesday,
Waymo announced
they’d purchase 20,000 sporty, electric
self-driving vehicles from Jaguar for the
company’s forthcoming ride-hailing service....
But the company embedded a much more
significant milestone inside this supposed
announcement about a fancy car. With orders
now in for more than 20,000 of these vehicles
and thousands of minivans that Chrysler
announced earlier this year, Waymo will be capable
of doing vast numbers of trips per day. They
estimate that the Jaguar fleet alone will be
capable of doing a million trips each day in
2020.
You could quibble with their math (will it
really be that many daily trips per car?) or
their overall utilization rate (how many cars
will be lost to maintenance per day?), but if
Waymo is even
within 50 percent of that number in two years,
the United States will have entered an
entirely new phase in robotics and technology.
The company’s autonomous vehicles have driven
5 million miles since Alphabet began the
program back in 2009. The first million miles
took roughly six years. The next million took
about a year. The third million took less than
eight months. The fourth million took six
months. And the fifth million took just under
three months. Today, that suggests a rate on
the order of 10,000 miles per day. If Waymo hits their
marks, they’ll be driving at a rate that’s
three orders of magnitude faster in 2020.
We’re talking about covering each million
miles in hours.
But the qualitative impact will be even
bigger. Right now, maybe 10,000 or 20,000
people have ever ridden in a self-driving car,
in any context. Far fewer have been in a
vehicle that is truly absent a driver. Up to a
million people could have that experience
every day in 2020.
2020 is not some distant number. It’s hardly
even a projection. By laying out this time
line yesterday, Waymo
is telling the world: Get ready, this is
really happening. This is autonomous driving
at scale, and not in five years or 10 years or
50 years, but in two years or less...." Read
more Hmmmm...Yup!!
This is HUGE! It will change the city
and the key to making it so it doesn't
make thing worse is Ride-sharing. If we
ride-share we'll reduce energy,
pollution & GHG by more than 50% and
provide high-quality, affordable
mobility indiscriminately for all. It
becomes the new high-quality, low-cost
mass transit. If it's kept/operated as
another alternative for the 1%ers to be
chauffeured alone, then the outcome is
UGLY. Ride-sharing is KEY! Alain
R. Mitchell, Mar 22, "Police late
Wednesday released a video that shows an Uber
robot car running straight into a woman who
was walking her bicycle across a highway in
Tempe, Ariz. The woman was taken to a
hospital, where she died Sunday night.
The video, shot from the car, is sure to raise
debate over who's to blame for the accident.
In the video, the victim, Elaine Herzberg, 49,
appears to be illegally jaywalking from a
median strip across two lanes of traffic on a
dark road. But she was more than halfway
across the street when the car — traveling
about 40 mph, according to police — hit her.
The car did not appear to brake or take any
other evasive action....
Bryant Walker Smith, a law professor and
driverless specialist at the University of
South Carolina, said: "Although this appalling
video isn't the full picture, it strongly
suggests a failure by Uber's automated driving
system and a lack of due care by Uber's driver
as well as by the victim."..." Read more Hmmmm...
"..."What we now need is for the
release of the radar and lidar
data," Princeton's Kornhauser said in an
email. (Lidar is a sensing technology that
uses light from a laser.) "Obviously, the
video of the driver is extremely bad for Uber
and probably implies that Uber should suspend
all of its 'self-driving' efforts for a while
if not for a very long while.
"The 'self-driving' systems are supposed to
have 'professional' overseers who are really
supposed to be paying attention during these
'tests'. Apparently Uber didn't make it clear
in this case."
Kornhauser questioned the police description
of a situation that would have been difficult
to avoid. He said Uber should reveal what its
collision-avoidance software was doing during
the couple of seconds before impact.
"The front-facing video suggests that this
person was crossing the lane at a slow speed
and should have been noticed by the system in
time to at least apply the brakes, if not stop
the vehicle completely," he said. "While a
human may not have been able to avoid this
crash, a well-designed, well-working collision
avoidance system should have at least begun to
apply the brakes."..."
" ... Again,
my sincerest condolences to Elaine
Herzberg's family and friends.
The simple
arithmetic is: She crossed more than a
lane and a half before being struck or
more than 15 feet. Average walking
speed is about 4.6 ft/sec
which means that she was "visible" on
this stretch of road for more than 3
seconds. Uber's speed of 38 mph = 55.7
ft/sec
means: Uber was 150 ft away when she
began crossing the left-hand lane and
could have been visible by an alert
driver. The car's lidar
and radar surely must have "seen" her
beginning at about that time. Car
stopping distance including
"thinking time used in The Highway Code"
@ 38mph is 110 feet. The driver should
have been able to stop 40 feet short.
Any Automated Emergency Braking (AEB)
system should have been able to stop the
car in little more than the stopping
distance of 72 feet, half way to
Elaine. This simple arithmetic suggests
that there may be a very fundamental
fatal flaw in Uber's AEB.
And the driver
was not paying attention. At 3 seconds
prior to impact, Elaine was within a 12
degree field of view when she began to
cross the left lane. While outside the
fovea, this is well within a normal gaze
had the operator been looking out the
window.
The released
video is from a "dash cam" and is
unlikely to be the video captured by
Uber's "Self-driving" system (or
whatever Uber calls it). That video may
well be at a much higher resolution and
frame rate. Uber MUST release that
video (not just the dash-cam video) as
well as the radar and lidar data that
was being used by their "Self-driving"
system. Uber was testing its system at
the time of the crash and therefore MUST
have been logging those data in case
something went wrong. Uber needs those
recorded data in order to have a chance
to learn what went wrong and fix it.
Something did go wrong, very wrong.
Uber and everyone else MUST also have
the opportunity to learn from this
tragedy. So Uber MUST release all of
the data. Alain
R. Mitchell, Mar 21, "As long as
robot cars roam public streets and highways,
they will occasionally kill people. That's an
ugly truth that no one in the driverless
vehicle industry can deny.
Will those robot cars kill people at
significantly lower rates than drunk, stoned,
tired or distracted human drivers do now?
Automakers, technology companies, politicians
and regulators are betting they will, as
driverless vehicles are rolling out faster
than almost anyone expected as recently as a
year ago. But the Sunday night incident in
Tempe, Ariz., in which an Uber robot car hit
and killed a woman walking her bicycle across
the street, makes clear the industry is much
further behind in making its case to the
public.
"It's likely there will be far fewer deaths
with driverless cars," said Marlene Towns, a
professor at Georgetown University's McDonough
School of Business. "But getting to the point
where people will be convinced of that will be
tough."
Speculation by Tempe's police chief that the
robot may not be at fault in the crash may
temper any public or political backlash.
Uber was testing the robot car in autonomous
mode with a human engineer, who was behind the
wheel but not driving. Elaine Herzberg, 49,
walking a bicycle, stepped in front of the car
from a center median, according to video
evidence, police said...." Read more Hmmmm... "...Carmakers and
technology companies need to be far more
transparent as they push forward, experts
said. "It's important that we all learn from
this accident and we make these technologies
even better, said Alain Kornhauser, a
professor at Princeton University and a
leading authority on driverless cars. "To
that end Uber must release all of the data
leading up to this crash. All of the video,
radar, lidar
and logic trails for the three or so seconds
leading up to the crash. If this releases
some of Uber's intellectual property, so be
it."..."
" ... My
sincerest condolences to Elaine
Herzberg's family and friends. I hope
that Uber with its "$60"B valuation
will make a very generous contribution
to homeless charities and think even
more seriously about "buying" (by
partnering) rather than "making" this
technology. Alain
G.
Kumparak, Mar 13, "...." Read more Hmmmm... This is REALLY
big news.This
marks the real beginning of on-demand
mobility provided by vehicles without a
driver or an attendant on-board, only
the passengers and the vehicles used
normal public roadways that operated in
normal everyday manner and used by
conventional cars and trucks. Ng Waymo to their
o police escorts, no warning signs, just
normal everyday operating conditions.
Except for the one trip given to Steve
Mahan in November 2015 in Austin Texas,
this is the First time that it kind of
mobility service has been delivered
anywhere in the world. Waymo has
achieved 5 million vehicle miles of
Self-driving (automated driving on
normally operating public roadway;
however, with a driver/attendant in the
car ready to take over should the
automated system begin to fail. Many
others including Uber, Lyft/Aptiv,
GM/Cruise, nVIDIA,
Apple, Tesla, Nissan and many others
have also done many miles of
Self-driving on normal roads but each an
everyone had a driver/attendant in the
vehicle ready to "save the day" should
something go bad. Nobody else anywhere
in the world is doing what Waymo is now
doing in Chandler AZ. Now that the first
one has been done, any community that is
similar to Chandler AZ can now think
seriously about inviting Waymo to provide
affordable on-demand mobility to
everyone in their city.
Be sure to see
the video. Congratulations Waymo!!!!!
Alain
D. Etherington,
Feb 27, "California’s Department of Motor
Vehicles established new rules announced Monday
that will allow tech companies and others
working on driverless vehicle systems to begin trialling their cars
without a safety driver at the wheel. The new
rules go into effect starting April 2.
Until now, the DMV has allowed companies
approved for autonomous vehicle testing to run
their cars on the roads, with autonomous driving
systems engaged, provided that there’s a trained
safety driver behind the wheel ready and able to
take over manual control. Now, the regulators
are updating their rules to allow for fully
driverless test, which is a key step along the
route towards actually deploying self-driving
vehicles in a commercial capacity.
This doesn’t mean test vehicles will be out
there on the roads without any kind of human
intervention backup – the DMV will require that
those testing autonomous cars without a driver
present have a dedicated communications
channel that ties the car to a remote
operator, who can take over if
needed. ..." Read more Hmmmm... Even though we
have been expecting this, it is a major
hurdle for it to actually have
occurred. How long after April 2 will Waymo take to
begin this type of testing. Again this
is only testing and deployment, but NOT
commercial service, which may happen
first in Arizona, but it is a major step
in this r-evolution. Commercial
services are regulated by other agencies
in California, not CA DMV. It is those
other agencies that will need to
grant/award the licenses for the various
commercial operations where these
driverless vehicles would be used. This
regulation allows properly licensed
commercial operations using CA DMV
certified driverless vehicles to have
those vehicles use California public
roadways in delivering the otherwise
licensed commercial activity. Note: CA
DMV does not license the commercial
transport of people or goods. That is
the purview of other CA regulatory
agencies. Alain
Andrew Hawkins, Jan 30, “Waymo, the self-driving unit of Google parent Alphabet, has reached a deal with one of Detroit’s Big Three automakers to dramatically expand its fleet of autonomous vehicles. Fiat Chrysler Automobiles announced today that it would supply “thousands” of additional Chrysler Pacifica minivans to Waymo, with the first deliveries starting at the end of 2018.
Waymo currently has 600 of FCA’s
minivans in its fleet, some of which are used
to shuttle real people around for its Early
Rider program in Arizona. The first 100 were
delivered when the partnership was announced
in May 2016, and an additional 500 were
delivered in 2017. The minivans are plug-in
hybrid variants with Waymo’s
self-driving hardware and software built in.
The companies co-staff a facility in Michigan,
near FCA’s US headquarters, to engineer the
vehicles. The company also owns a fleet of
self-driving Lexus RX SUVs that is has been
phasing out in favor of the new minivans. (The
cute “Firefly” prototypes were also phased out
last year.)…” Read
more Hmmmm...
We’ve all been wondering” Who’s going to
make the cars? How will that evolve?Will they
magically appear???
Well….Looks like
it is FCA for now. We've gone from a
handful 5 years ago, 2 years ago added
100, added 500 last year, “thousands”
this/next year, … Beginning to look like
exponential growth! (A Bit Coin Bubble??)
What is also most interesting: no
parallel announcement that Waymo was hiring
“thousands of attendants” to ride around
as "drivers" in these “thousands of
minivans”. Guess what that means… The Kornhauser
Scale is going to start really going
up!!! J
While ultimately
they’ll need about 35 million of these to
provide affordable mobility to all in the
US, this is a real start at making this
into a business as opposed to an NSF-style
study that collects dust on a shelf or,
worse yet, a digital manuscript that is
never downloaded by anyone outside a
"group of three". This is a major
announcement!
From Stan Young: It will be interesting to watch. It probably has the OEMs, Uber and Lyft scared out of their wits. Based on any objective comparison of accomplishment with automated vehicles, there is not a close second to Waymo, despite all the claims to the contrary by trade rags – and the competition knows it. Still a huge unknown concerning the ‘social side’ of riding in an un-attended vehicle, but we will likely get over it like we did with elevators. ‘Thousands’ of vehicles if deployed in one city will put it on scale of Uber and Lyft – an interesting study when/if it comes to that.
...An issue is: where will Waymo choose to deploy (and for Waymo, the word "deploy" is the right word... they make the decision where to place these, in some sense take it or leave it... as opposed to waiting for people to show up at a dealership to buy or have it stay on the lot or have some governmental agency thinking that it actually has a role/power/where-with-all to “deploy”) where, when and how many. They could "flood/concentrate" on Chandler/Phoenix/Tuscon area with scale to be really relevant and substantively demonstrate the evolution of mobility, or they could sprinkle them out nationwide and remain irrelevant everywhere. I like the "flood/concentrate" approach in a state (Arizona) where they seem to be truly welcomed and whose climate, topography and road network are "easy". More importantly it would demonstrate the viability/challenges of the at-scale approach. From our simulations we uncovered that at-scale, one might need to be managing as many as 20,000 aTaxis in a 2.5x2.5 mile area (the extreme in Manhattan, which may be the last place that you want to try this) but it can be large. We’ll drill down in our data and take a look at Chandler/Phoenix and report back as to what we think it would take to provide mobility for all. Alain
Jan. 9, T. Papandreou & E. Casson. "... Waymo driverless service..." Read more Hmmmm... Tim and Ellie made presentation at the Transportation Research Board's Vehicle-Highway Automation (AHB30) Committee meeting on Tuesday in which they gave an update on Waymo's progress to launch "Waymo's driverless service" (slide 11), an app-based ride hailing service to the general public in a geo-fenced area of Arizona. To date Waymo has been testing such a service using volunteer riders in their driverless vehicles in various areas around the country (slide 7): however, to date, except for one ride given to Steve Mahan in Austin, TX, rides on normally operating public streets have always had trained Waymo-authorized personnel (an attendant) in the vehicle capable to intervene in the driving of the vehicle should the need arise. Since October, in Arizona, those personnel no longer sit behind the wheel, but are in the back seat so that Waymo can observe the response of the volunteer riders to riding in a vehicle on normal public streets under normal conditions without anyone in the front seats of the vehicle.
Tim said, without providing a specific date, that Waymo will soon launch "Waymo's driverless service" providing mobility to the general public on public roads in a geo-fenced area of Arizona. I asked Tim "Will that service be offered with vehicles that have an attendant in the vehicle?". Tim's answer was "No!". I asked a follow-up question: "Will these vehicle's have telemetry capabilities that enable these vehicles to be closely monitored from a "situation room" or "control center" that would enable remote operation of the vehicle, should the need arise?". Tim's answer was "No!". Another questioner asked if the geo-fenced area included special "connected vehicle" road infrastructure improvement that Waymo's system will be relying on?" Tim's answer was "No!".
While the definition of "soon" was not given, I've taken this as a really big pronouncement that Waymo is actually going to go to launch commercially-viable on-demand mobility to the general public on conventional public roads. This is really big news because this is finally going to enable us to begin to evolve on the "Kornhauser Scale" ( log of (world-wide VMT of Driverless (VMT-D) vehicles without a human attendant/driver on board accumulated while providing mobility to the general public on conventional roadways). So far we are beyond the "undefined value" associated with VMT-D = 0 and are at KS = 1 only by virtue of the one Steve Mahan ride in Austin). :-) 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
Rulemaking
Actions, Oct 1The following 3 PDFs
are important:
1.
Autonomous Vehicles Notice of Modification
(PDF) Act
2.
Autonomous Vehicles Statement of Reasons (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
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
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
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
This list is maintained by Alain
Kornhauser and hosted by the Princeton
University
Leave
|Re-enter
[log in to unmask]" alt="imap:[log in to unmask]:993/fetch%3EUID%3E/INBOX%3E3022058?part=1.39&filename=dhbhaandkmfbffia.png" class="" height="88" width="106" border="0"> [log in to unmask]" alt="imap:[log in to unmask]:993/fetch%3EUID%3E/INBOX%3E3022058?part=1.40&filename=lglcejopfgfnajaj.png" class="" height="92" width="238" border="0">[log in to unmask]">Mailto:[log in to unmask]