Tesla is a car company that also works on self-driving technology. In this segment, they’re discussed as one of the companies testing driverless systems rather than selling them as fully legal driverless cars.
“Test mode” means the self-driving system is being run for trials, not as a normal product you can buy and use everywhere. The point here is that the tech hasn’t yet crossed the legal threshold for widespread driverless sales.
“Driverless mode” means the car drives itself without a person actively driving. The speaker says that, right now, you can’t buy a street-legal car that’s allowed to run fully driverless everywhere.
A “street legal vehicle” is a car that’s allowed to drive on public roads under the rules. The speaker’s saying that current driverless systems aren’t packaged as legal, sellable cars for that purpose.
Driverless operation means the car can drive itself without a person actively steering or controlling it. The episode is talking about the rules that governments are creating so this can be legal for regular customers, not just experiments.
WP 29 is a United Nations group that writes rules for vehicles. In this episode, they’re credited with building a regulation framework for testing self-driving cars without a human in the driver’s seat.
Type approval is the legal process that checks whether a car model meets safety and regulatory rules. Once it passes, the car can be sold more broadly under that approval framework.
NHTSA is the U.S. government agency that makes and enforces vehicle safety rules. In this episode, it’s mentioned as the body overseeing how these new rules apply in the U.S.
A testing agency is a group that performs the official tests on a car to see if it meets the required rules. The idea is that the test results can be used to approve the car for sale in multiple places.
Vehicle type approval is the “pass” that a car model gets after it’s tested against the rules. Once approved, it can be sold in other places that accept that approval.
FMVSS are the U.S. safety rules for cars. They tell manufacturers what tests to pass and what safety requirements the car has to meet before it can be sold.
Self-certification means the car maker does the compliance testing and paperwork itself. Instead of an outside lab certifying it, the manufacturer says, “we meet the rules.”
A recall is when cars are pulled back or fixed because they don’t meet safety requirements. In this context, it happens if the car is later found not to comply with the regulations.
The dynamic driving task is what the car has to do moment-to-moment to drive safely—things like steering and braking. The regulations being discussed aim to let the car do that without a human driver taking over.
The “2,500 cars” number is a cutoff in a U.S. rule. If a company sells fewer than that, it may not have to meet every regulation—though the car still has to be able to stop safely.
Driverless testing is when self-driving cars are allowed to operate without a human driver, but only in approved areas and under special rules. The podcast explains that this is different from normal commercial deployment.
“No driver” means the car is driving itself without a person sitting at the wheel. The podcast contrasts Waymo’s ability to do this under testing rules with Tesla’s earlier limitations for broader deployment.
San Francisco is one of the cities mentioned where Waymo is testing self-driving cars. The point is that the cars are only allowed to operate in specific approved places.
Phoenix is one of the cities mentioned where Waymo is testing self-driving cars. The podcast is emphasizing that these systems run in approved areas, not everywhere.
Austin is mentioned as a city where the host thinks Tesla is testing under updated rules. The takeaway is that these permissions can be limited to certain cities first.
These are U.S. safety rules that cars have to meet before they can be sold and used on public roads. The point is that the rules may be updated so driverless features can be approved.
A “cybercab” is Tesla’s idea of a self-driving taxi-like car. In this conversation, they emphasize it’s built without the usual steering wheel and pedals, but a person can still be inside.
“Cybercars” is basically Tesla’s name for the same kind of self-driving taxi concept. The big point is that it’s designed to not need the usual steering wheel and pedals.
Term
20 milliseconds stuff
“20 milliseconds stuff” refers to very short control-and-response timing in automated driving—how quickly the vehicle’s steering and control system reacts to perception and planning inputs. In practical terms, it’s about latency and control loop timing being tight enough for safe maneuvering.
This is the car’s “watch and decide” role. Instead of reacting instantly to every tiny change, it looks ahead for problems like road closures and chooses what to do next.
Fusion is a type of energy production that needs extreme conditions. Here it’s used as a joke/metaphor to say “that’s not going to happen anytime soon.”
Term
20 millisecond stuff
They mean the car can “think and react” extremely quickly—about 20 thousandths of a second. Faster reaction time can help avoid crashes when something unexpected happens.
Term
out of the loop
“Out of the loop” means the human isn’t making the driving decisions in real time. The idea is that the car’s system should handle the driving instead of waiting for a person to react.
The Tesla Cyber Cab is Tesla’s planned self-driving taxi concept. The point here is whether it can run with no human driver inside, while still being able to control the car safely.
A “mapped service area” is a specific region where self-driving cars are allowed to operate because the system knows the area well. The host is contrasting that with unpredictable situations outside the plan.
“Human in the loop” means a person is watching and can step in if the car gets confused. This part is about rules for cars that are supposed to drive without that human backup.
This is the “watching and stepping in” part of self-driving. Even if the car can drive quickly, there are situations where it needs a higher-level check or a human/backup response.
This is Tesla’s name for its advanced self-driving features. The point here is that even with it on, the system may still need a person ready to take over in tricky situations.
“Jaguars” is shorthand for cars from the Jaguar brand. The host is using it to illustrate that you can’t always deploy huge numbers of cars onto public roads, even if you could build them.
Zeekr is a car brand, and the host mentions it as a substitute when they can’t get enough of another brand onto the road. It’s tied to the bigger point about scaling driverless deployments.
A “business case” means there’s a real-world reason the idea can work financially. The host is saying safety alone isn’t enough—there has to be a practical way to make driverless cars work at scale.
Alphabet is the company that owns Google. The host is saying that even if Waymo (a driverless company) talks a lot about safety, the bigger company behind it may not have a clear business reason to push driverless cars at scale.
Elon Musk is the high-profile leader behind Tesla. The host is saying Musk’s bigger dream (going to Mars) is the reason Tesla’s driverless efforts are being pushed, because it helps generate the money needed for that goal.
Here, affordability means whether rides are priced low enough for more people to use them. The speaker says self-driving could help because it removes the need to pay a driver.
Driverless cars are cars that can drive themselves. Instead of a person doing the driving, the car uses sensors and software to handle the driving tasks.
New Jersey is mentioned because the speaker says the state’s politicians are making rules that affect testing self-driving cars. Those rules can change how quickly the technology gets proven and improved.
Concept
real testing of cars
The speaker is talking about doing actual real-world trials with self-driving cars. They’re arguing that rules can make it harder to test them properly, which slows down improvements.
Term
turn on a dime
“Turn on a dime” means the car can change direction very quickly and very sharply. The point here is that a vehicle has to steer precisely when it’s needed, not slowly or unpredictably.
“Stochastic” just means things are a bit random and not perfectly predictable. So instead of a ride system running like a timetable, it has to handle changing demand and timing.
Concept
removing the driver from the task
This means the car does the driving work instead of a human. The speaker is saying that while it can lower costs, it also changes the whole way the ride service operates.
They’re talking about how much money the human driver costs, and then pointing out that there are other costs too. The idea is that driver pay isn’t the only expense involved in running a ride service.
“HR stuff” means the behind-the-scenes work of managing employees—like scheduling, benefits, and making sure there’s coverage. The speaker is saying that those costs add up too.
AV legislation is government rules for self-driving cars. It can include how companies test them and what safety standards they must follow before offering rides to the public.
Autonomous vehicles are self-driving vehicles that can navigate and drive without a human doing the driving. The point here is that they could help people who struggle to get rides or transportation.
Quality of life means how good life is day-to-day. The point here is that self-driving ride services should be judged by whether they help people get to school, doctors, and jobs reliably.
Latent demand means people want something but can’t get it right now. In this case, it’s the idea that many people would use better transportation options if they were available and dependable.
Value proposition is the reason people should care about something—what good it does for them. Here, it’s about how transportation can improve daily life, not just the self-driving tech itself.
“Driverless vehicles” are cars that can drive themselves instead of a person doing the driving. They still have to follow rules set by the government before they can be used on public roads.
“Testing to deployment” means taking a self-driving system from experiments and trials and getting it ready for everyday use. It’s not just about proving it works once—it's about meeting the rules and being ready for real roads.
Concept
proof of market test
A “proof of market test” is a real-world trial to see if something is actually practical and wanted. For self-driving, it’s about checking whether it can work in real life with real people and operations.
Mercedes is a well-known car brand. The speaker is saying that even if someone from Mercedes was there, car companies weren’t actually leading the discussion on the panel.
Volkswagen is a big car brand. The speaker is pointing out that car companies weren’t the ones presenting on the panel, even if someone from Volkswagen might have been in the audience.
A “legal framework” is the official rulebook the government uses to decide what’s allowed and who’s responsible. For self-driving cars, it covers safety requirements and what happens if something goes wrong.
Waze is a navigation app that uses real-time traffic info from drivers. The speaker brings it up as part of the group discussing how rules should work around driverless-related tech.
Concept
fallibility of all humans
The speaker is saying people aren’t perfect, so the rules should assume mistakes can happen. That mindset also applies to self-driving systems, which can also fail.
Concept
imperfection of all technologies
The point here is that technology can still make mistakes. So the rules for self-driving cars should be written assuming the system won’t always work flawlessly.
“Road fatalities” means people who die in crashes on public roads. The speaker is arguing that self-driving cars may reduce risk, but they won’t make deaths go to zero.
San Mateo police are the local police in the San Mateo area. The story is saying that when something unsafe happened in a self-driving ride, they called the police.
Concept
safety of the roadways
“Safety of the roadways” here refers to the broader real-world risk environment that autonomous systems must operate within. The host argues that rider misbehavior may be a limiting factor for how quickly autonomous driving improves safety outcomes.
The Dodge Ram is a large pickup truck made for carrying things and towing. Because it’s big and heavy, it can cause more damage if it crashes. That’s why it may be mentioned in safety or driving discussions.
“Cameras” are the car’s eyes. They help the system see what’s around it, and in this context they’re also used to review what happened if there’s a dispute.
“Insurance fraud” is cheating the insurance system—like lying about what happened to collect money. The speaker is saying that this kind of behavior can complicate how accidents and disputes are handled.
LIVE
Welcome back to the smart driving cars podcast. Thanks for spending time with us once again.
I'm Fred Fishkin along with the faculty chair of autonomous vehicle engineering at Princeton
University, Alan Kornhouser. Hi, Alan. Hi, Fred. Good evening. Good evening. Mid, middle
of summer here, a July Jersey edition, but not in New Jersey. Joining us once again
from Sweden is consultant and coauthor of the real case for driverless mobility, Michael
Senna. Thanks for joining us, Michael. Thank you. Nice to be here again. And Michael, it's
nice to have you because I actually were not in the same day. You're a day ahead of us.
That's right. I mean, you know, this is like weird, but you know, it's 153 in the morning
here in dark Sweden. It can't be dark. It should still be light. I mean, my goodness.
Anyway, thank you for getting up out of bed and dragging you and being a day ahead of
us and whatever. But you're always ahead of us. Nice to have you, Michael. We have important
things to talk about. We do. We do indeed. Well, there's been a lot going on when it
comes to regulations, when it comes to automated vehicles globally, really. And Michael, you're
here to catch us up more or less. I'll let you get started.
Yeah, it's, it's kind of hard to know where to start with with this whole issue because
the companies that have been operating in the in this sphere, primarily Waymo and Tesla,
there are other companies that have come and gone to trying to develop the technology
for removing a human driver from the task. And the principal reason that has been given
for this is that it improves the safety of the vehicles that because you are replacing
a human driver with a robot who will always do the right thing, there won't be any deaths
on the highways and the byways that and cars will be safer than they've ever been before.
The problem with that has been that the cars that are that are operating, whether they're
Waymo's or Tesla's or anyone else, they've been operating in test mode. There isn't,
there isn't any vehicle that can be sold today in any market today that can be sold as a
street legal vehicle that either an individual or a company could operate in driverless mode.
It's taken quite a long time since people began the efforts to establish the regulations that
would allow this allow vehicles to be operated driverlessly. But one thing actually has happened
and it was a month ago. The organization called WP 29, Working Party, it's a United Nations
function that has been responsible since the 1950s for creating regulations in countries
that adopt their regulations that are signatures to the this UN group that provide the requirements
for a vehicle to be legal, to be sold legally for operation commercially, not as either a test
vehicle or as a very short run specialized vehicle, but to be sold to consumers in any
any quantity, a 2000000 however many you want to sell. This organization after two years
of very intensive work created a structure for a regulation that could be used to test vehicles
without a human driver behind the wheel. One other thing occurred as part of this regulation
during the process of developing it, the countries who were not signatories to this
what's called the type approval process.
Primarily the United States and China, they took part actively in all of the work that was done
and they agreed to take this regulation and to put it into their own country requirements. So in
overseeing by NHTSA, the National Highway Traffic Safety Administration,
NHTSA is responsible for creating the regulations and they're responsible for ensuring that they're
followed. Many of the people who are listening right now know the difference between the process
that's used in the so-called type approval countries principally Europe as well as Japan
and South Korea and a number of other countries where there is a testing agency. There are many
agencies but they all perform the same thing and the car is tested by this agency in let's say it's
in Germany, it could be in France, it could be in the Netherlands, it could be in Sweden but let's
say it's in Germany. They perform the tests that have been established by WP29 as a regulation.
They determine if the vehicle, if the entire vehicle that's tested meets those regulations
and then that car can be sold anywhere where type approval is in effect. So if you test it in
Germany it can be sold everywhere in within the European Union and also in Switzerland,
Norway, the United Kingdom and various other countries that are part of the type approval process.
In the past when according to other regulations in the United States there is no testing agency.
The FMVSS provides the regulations and the requirements for determining whether a car is
in Europe type approval countries are performed by organizations, agencies, testing agencies
and these and that each company tests its own vehicles and provides a certification,
self-certification. It says our vehicle complies.
Obviously the FMVSS has to be followed, everything has to be according to those regulations
and if it's determined after the car is put on the market that it doesn't then there's a recall
process that goes through and things happen that in order to get those cars really according to
the regulations but that's the main difference between the two. They're essentially the same
types of requirements, the same types of regulations, it's just that in the United States
it's the companies who go through the certification process rather than an independent testing agency
and so far it's worked out just fine. Now it has taken quite a long time to get to the point of
having a regulation in place which could be followed in order to remove a driver from the
what's called a dynamic driving task, to remove that driver completely from the dynamic driving task.
There's the possibility if you take Tesla right now with their full self-driving.
Tesla has said that our full self-driving is not, does not meet the requirements today
for removing a driver and the reason that they've done that, the reason that they've said that,
is that if they said that our car is driverless, even though they call it full self-driving,
but if they said our car is driverless, there would be no regulation which would cover that car.
So all of the companies that are out there today who are providing driverless cars,
Zooks and Waymo, they're doing that under a, in the United States, they're doing that under a
special regulation which NHTSA has that says if you sell fewer than 2,500 cars,
you do not have to meet certain regulations. Obviously you must have breaks and you know
the car must be able to stop, but you will give you a pass on certain regulations
within a limited area for testing purposes. And this is what Waymo has been doing for
well almost the last 10 years essentially. They've been able to remove the driver
and allow cars to be driven around in special areas where the states and the cities have also
allowed this to be happening, but they fall under this requirement that says you can go out there,
you don't have to meet all of the requirements of the federal motor vehicle safety standards,
but you can't go out there and start selling and deploying a vehicle, a vehicle in greater numbers
than 2,500. And Tesla is part of that group as well. Now it is of late. Of late it is, of
but in a very special way. They've been allowed to do what Waymo has been doing
with the same requirements and the same sort of capabilities. They haven't been able to do it
with the no driver, no the cyber cab, because the cyber cab doesn't have the minimum
requirement for being able to put the car into a market. Now they've been allowed to do that,
and I think they're doing it. I'm not 100% sure. I think they're doing it in some areas
in the Austin area. But in the case of Waymo, Waymo has got their vehicles operating without a
driver, without a backup driver, without somebody sitting behind the wheel. There are wheels in
their car. There are all the cars that they're using for testing. There are controls so those
cars can be driven in a case of emergency. But they're operating those cars in San Francisco,
in Phoenix, and in Austin, and in various other places, Atlanta now, without having anybody
in the car when the car is being driven around. But these cars are operating in a test mode.
There isn't any commercial operation. Now Waymo has been authorized to charge for rides.
How do you get around that situation where they look like they actually have a business
which they don't? Waymo says their mission is to be the world's most trusted driver.
That's Waymo's mission. I heard it. I was in Geneva at a conference listening to the
representative from Waymo, and Waymo said our mission is to be the world's most trusted driver.
They didn't say our mission is to deliver the most affordable rides. Our mission is to deliver
the safest and most affordable rides for people who can't afford it like the mission of handy rides.
They just said our mission is to be the world's most trusted driver. Their business
is to help companies to get the technology that's necessary that would comply with the
regulations that have now been signed and they're now going to be put into the regulations in every
country with a type of approval process. There's no timetable that's been given for this,
but they will eventually be put into the Federal Motor Vehicle Safety Standards
so that cars can implement driverless technology, can be tested so that they can be approved for
street legal operation, and people can begin to create businesses without having drivers.
Now, that's where we are today. That's as far as we've come with driverless technology today.
And let me just ask you this. Tesla has said, and they've said a lot of things over the years,
of course, that have taken much longer, that they intend to begin selling to consumers
their cybercabs, cybercars. Those are vehicles without the steering wheel or pedals
to consumers next year is what they're saying. And to help them along that path,
the NHTSA has said that they are in the process of creating regulations that would go into their
Federal Motor Vehicle Safety Standards that would allow vehicles to be driven without having pedals
and steering wheels. Now, time frame on that, as far as I know, none. Allen put his hand up,
he wants to interrupt me. Go ahead, Allen. No, I don't want to, I mean that, but Fred, you opened
you opened up another can of worms here. Okay, cybercab is not a driverless vehicle.
Cybercab is a vehicle without pedals and steering wheels. But it has a person in it.
Task or, you know, the dynamic driving task, assume it here, a Venn diagram,
boom, we're going to chop it in two pieces. Okay. One is what I like to call nobody else's
calling. Well, you know, I didn't like the levels. I'd like to make up my own names. That's why I
call smart driving cars. Give me a break. You know, I'm supposedly a professor, so I get to
give names. The one piece over here is I'd like to call it the 20 milliseconds stuff.
You know, what angle should the steering wheels, steering wheel, the wheels on the ground be? I
don't care what, what, which way should the wheels that are going to guide the vehicle,
which directions be? And what does the rattle brake combination look like? Or what's really the
acceleration, deceleration, piece of the vehicle? And, and traditionally, that's been done by a
person grabbing a wheel and hitting the pedal or putting the pedal to the metal or, you know,
whatever, you know, that's what's been done every 20 milliseconds or so as you go from A to B.
That's that guy. And then the other guy is more of a supervisory function.
One that says, I'm stuck here. Turn around. Or I got to find new route. Or the road is closed ahead.
I mean, that supervisory function, the road is closed ahead. I don't only need to,
the supervisory function needs not only the road closed ahead, but needs to know
which side of the closing you're on. Are you going into a closed road that's closed ahead?
Or are you coming out of it? Because the supervisory function is going to be quite different.
You know, if I'm outside of it, then the supervisory function says to go around it.
But that, that doesn't take place at the 20 millisecond level. That takes place at the
second, five second, 10 second, 20 second, sit there and think about it.
You know, there's, I see water ahead on the road. I don't know that there's any
sensor that tells me how deep that sucker is. I can't tell because I'm driving my 55 Chevy.
Okay. And so, you know, I do a supervisory function that says, whew, maybe I, I should
do a K-turn right here or you would right here because otherwise, you know, I'm done.
Ba-ba-boom. Those are supervisors. So that's the second piece.
Both of those are covered. Both of those have to be covered and are covered in the, in the
regulations. Of course, the dynamic driving task needs to be. Okay. But it may be that a technology
can do the 20 millisecond stuff. Well, I'm better than I can.
And even though the folks in the AI business think, oh my goodness, generative AI,
blah, cut it out.
Cut it out. Stop right here. You know, that's, that's 50 years from now. You know,
that's one of these thermonuclear things. Hey, when are we going to have thermonuclear fusions
always 50 months, 50 years ahead? That's cut it out. So in fact, you know, I would look at the
RoboTaxi as a vehicle that it probably can do the 20 millisecond stuff, not only better than I
can do it now, better than maybe I've ever done it. Are you talking about the cyber cab or about
Waymo? Because they're both called RoboTaxi. No, I'm talking about the cyber cab without,
I don't need a steering wheel and a throttle. That's just because it has to interact with
me as a human. And actually, I'm the safety problem in the human. I'm the safety problem.
Mindless behavior is the problem. And that's why that's why this whole chase chase of, you know,
oh my goodness, I'm going to solve the safety problem by taking the person out of the vehicle
and out of the loop. If I found, if I'm one of these guys that have, man, I love to drive.
You're going to be again, 50 years in the future before you get that before you make
you in a headway on the safety problem. Okay, so the fact that there's a throttle,
something for you to put your foot on or something to boost, irrelevant in some sense.
If a darn thing sees 360 and interprets properly and reacts every 30th of a second,
it doesn't, it doesn't need this thing. It doesn't need that thing, but it does need the wheels.
And that's need a motor that go or provide some friction one way or another on that,
on that surface. So that the taking out of the, of the steering wheel, but guess what that thing
still needs. I think it needs supervision.
Yeah, people can do that. Okay. And so that's the thing that
that in fact, the Tesla cyber cab, I don't, I don't know if we've really,
have we really, have they sent that sucker out without anybody in it yet?
Let me answer that. What they're saying as of yesterday.
Yeah, tell me what they're saying, but you know, what they're saying is that yesterday employee
rides at giga, Texas have started without anybody else in the vehicle. Anybody else.
what's a basin or has is reporting employee rides at giga, Texas are the first time real
passengers on a repeatable schedule are stepping into a production cyber cab and being moved from
point A to point B by software alone. No steering wheel to grab. No, there's no steering wheel.
And once you're in your in the software's hands, there is not who does a supervision piece. I've
broken the task up into two pieces for my purposes, not regulation purpose, you know,
dynamic driving. Let me have a choice. Two pieces. Okay. Who's doing the supervision piece?
Can we, I'm not sure where you're going with the cyber cab, but if you sit behind the wheel
in a Tesla with full self driving and you put full self driving active.
I do it every day. That's the only way I go around anymore. Okay. Yes, go ahead.
And are you supervising it? Absolutely. Okay. And for you supervising it means if it starts
to do something that you don't think it should be doing that you take over and put your hands on
the wheel and put your feet on the pedals and you drive the car. Yeah, but I could just as easily
say, hey, yo, slow down or turn left 22.6287432 degrees. I could I could give it voice
input. Yeah, you can do that. Sure. Yeah, sure. Right. But I'm supervising. When I came up against
the barricade. Yeah, I supervised it and told it to take the gap. Yeah. Okay. Okay. So it didn't
know it didn't what I knew that I was inside the closed area. Yeah. But Waymo
says our cars are being driven around in a in a specific area. We've mapped it. They're being
driven around mapped it. There is no, there is. Look, Alan, look, they mean it's a bunch of crap
because they go into water. They go into crowds. Their supervisory function, my opinion. Whoa. I
mean, otherwise, what are we reading about? What are the people in San Francisco complaining about?
They go into crowds, they go into places with firecrackers and and celebrating Chinese New
Years. Right. So what what to return to where I was before the regular what the regulation does it
it addresses both of these issues. It says if you're going to have a car
that's going to be operated without a human in the loop that the vehicle must be able to meet
all of both of these requirements. Yeah, that's why I said that. But Fred took us into the
cybercar without steering wheel and without gas pedal. And I said, I don't know that doesn't
make it into a driverless vehicle. What makes it into a driverless vehicle is being able to handle
the whole regulation. As you said, it's got to handle the 20 millisecond stuff. And it's got to
handle the supervisory stuff, which is not 20. Nobody, no, no person can intervene
in 20 millisecond. I mean, my cognitive cycles, I mean, I don't know. I do maybe 20 cognitive
cycles per second. You mean, if you're not, you mean, if you're not driving the car, like with
full self driving, if you're not driving the car and the car says to you something, you need to do
something now, because I can't do it, that you don't have enough time to do that. Right?
I mean, if I'm expected to operate at the 20 millisecond level, I can't.
Yeah, I can't. I mean, out of the whatever. And it's always been, it's been a, it's been a level
three problem forever. You know, somebody's who knows doing what in the backseat. Oh, hey, yo,
you, I mean, you know, whatever. Okay. So, so, so, and I just, you know, I'm taken because,
because I guess I'm of the opinion, just my personal opinion that right now, my experience,
my personal experiences, this stuff at the 20 millisecond level, I can't
compete with it. It's so damn good. Okay, that's good. It's my butt. Yeah, that's good. It does,
the 22nd millisecond, the supervisory piece. Cut it out.
Yeah, so when you're driving, when you're driving your Tesla and you have it in full
self driving, you find that it does the things that it needs to do to stay on the road and not
hit things and get you to where you want to go. But there are situations people and not kill people.
Oh, okay. Yeah, not hit things, not kill people. But there's something missing.
Well, you know, I'm not ready. I'm not ready to, to let that sucker go to
go to Fred's house and pick them up. Without having someone behind the wheel
in order to be able to supervise it. Behind a wheel, I don't care. I don't need a wheel.
Wheels and pedals, that's old school. You don't need them anymore. Okay, you really don't.
Because probably the best way for me to react is if something starts to go wrong
and they need supervisory, I hope I can talk to it. Okay, yeah, because because if I grab a wheel,
I'm gonna do that. Yeah, holy hell, that's the worst thing that could happen. I'm gonna flip it.
I'm gonna end up in this roof and it breaks my neck. It doesn't have it. anything
to protect my neck. I mean, a pillar, if you trust them, I don't. And that's what needs to,
I guess that's what needs is, is counting on by saying that we're gonna, we're gonna remove the
pedals and the steering wheel, but you still have to be able to maneuver the vehicle in case
the software in the car doesn't do it. No, no, I don't maneuver. I supervise them. It's,
supervising is different than the 20 millisecond. I do, we, my opinion, personal opinion,
we can no longer compete as humans at the 20 millisecond level. So the perform and maneuver,
park this, do a K turn, pull over, hey, you're blocking traffic, pull over to it.
It kicks our butt. We're done as humans doing that. I mean, if we put up a competition,
we lose. I mean, my money's on the, on the machine. I think I win. I get rich. Okay.
Let's do it right now. Okay. Just to put it into, it's put it into the real world context.
This is what companies are saying that they're able, they will be able to do this. The regulations
are such that they can be tested. Do what? Do what? The 20 millisecond or the supervisory?
I got to do both. I, if I don't do both, I'm not driverless. I can't put it out there.
The regulation is in place right now. Addresses both of those. Okay. Driving is not just steering
the wheel. Everything, everything that's, that's part of the driving task has. And I just chose
to break them into two pieces. It's the most complex activity that humans do today.
You can shake your head. You can shake your head, but all of the,
everything that's been done to analyze that problem. It's the most difficult and complex human.
More complicated and ballet dancing. I doubt it. I mean, come on. I mean, you know,
that's, that's a bunch of, oh, I don't know. I don't know anything about ballet. So, you know,
I mean, sure, sure, but come on. I mean, you know, some of these, some of these guys out here,
I think you're usually on point. I think you're usually on point. So, yeah.
No, but, but, but this,
I guess the reason why I bring all this out
is that in fact, even though I've been one of the biggest champions of driverless,
for why? Because the only way that I see you provide any societal good out of this technology
anytime early in its deployment. I mean, in other words, you go from nothing
to full. You've got to be delivering value all along that evolution curve. And it doesn't go
like that as we've noticed. I mean, my goodness. I mean, otherwise we would have done this in
2014. Okay. And here it is 2026. And we're sold. You got nothing. And you've delivered essentially
zero that societal value out of this. And you've not made, you've not made rounding error on safety
yet. And you're not about to. And you've done nothing for societal value. And you spent 320
billion or whatever McKinsey says that we've all spent. And I spent 50 plus years of my life
trying to do automation to try to provide some good. And guess what? The scoreboard says
we got zero. We've provided 00 or as close to zero, I mean, 10 to the minus who knows
what. And what the hell are we doing here? And now we have regulations that finally say this.
And that these regulations supposedly are going to chase safety. It's goofy.
It's look, it's the only thing that the regulations can do. They can't, they can't
determine a business. This is the problem. Alan, look, you and I know this. Let's communicate
this to the people who are listening to us. The reason that we're in the situation we're in right
now is because we've had a company principally, and I'm going to put this the primary burden
on Waymo. We have a company out there that has established itself. You can cut this out if you
feel it's so too controversial. It's established itself. It's a mission to be the world's most
trusted driver. Now, they're out there giving rides. What is the world's most trusted driver
doing? What's the reason that's doing this? What's its motive? What's the mission that
actually is trying to accomplish? It says it wants to be the most trusted driver,
and how is it going to do that? Is it going to be the taxi for the world?
We know that that's not the case. We know that the reason that they have, they don't have 80,000
cars out there. They could have bought 80,000 Jaguars. The reason they don't have them out
there is they couldn't put them on the road, right? They couldn't put that many cars. They can put
2,000 or 2,500 Jaguars on the road, but they can't, so they move to Zeekr. They haven't told,
that's the point. They haven't told us anything, and the people who are trying to produce regulations,
the people who eventually are going to have to pass laws, are out there saying,
wait a minute, what should we be doing? We have to focus on safety, and Waymo is telling us that
the safest cars, and they've got all this technology in there, but there's no business case.
There's no, why are we actually doing this? Waymo can talk to everything they want about
being the world's most trusted driver, but they're operating inside a company called Alphabet.
Alphabet has nothing in the world related to driving, or being trusted, or anything else.
If you look at, what is Google's or Alphabet's mission in life? To be one click away from
information. Don't do evil. Well, actually they left that whole business. What is Tesla doing in
driverless vehicles? Is it because they want to sell robots, and that this is a way of selling
robots? What's the motive behind Tesla? The motive behind Tesla has to be the motive behind Elon Musk,
because he's the guy that's calling all the shots. What's his motive? Does he want to be
the richest man in the world? Well, he's done that. No, he wants to be the richest man in the
world so he can go to Mars. That's his motive. So the motive of Tesla through Elon Musk is
to go to Mars. Wait, I want to go to Mars too, okay? You want to go to Mars too, but you know,
in order to go to Mars, you need a lot of money. So he says, in order to get a lot of money,
I got to convince enough people to buy my cars at very high prices in order for me to earn as
much money as I possibly can, in order for me to fulfill my dream, which is to go to Mars.
I mean, we all have motives. How we get there, what we do, if we want to deliver affordable
rides, if our mission is to deliver affordable rides, what's the motive for that? We want to be
everyone. We want to have something nice on our tombstones saying, this guy really did a great
job in life. Or is our motive to have as much money as possible so that when we retire, we can
do anything that we want to give it to our kids? Wait a minute, Michael. I mean, every once in a
while, Elizabeth, and I wonder why we're doing this, okay? Well, I think the short answer why
we're doing it is because in the past, more than a few times, I've needed a ride. And it was a shame
I couldn't get one because it might have made my life a little bit better. And unfortunately,
you know, I found myself meeting a lot of people whose lives could
really be improved, who would be a little bit happier if they could just go get their hair
done. Yeah, so do simple things. And so, you know, unfortunately, there's the motive. I guess,
I'm a good Marxist or something. I don't know. Yeah, fine. So you've got a motive.
Yeah, I have motive. How do you do that? How do you actually get affordable? This is the thing.
I get my butt kicked. I'm getting my butt kicked. Exactly. I am getting my butt kicked.
When I try to make the case. I am painted in right now. I am painted in the corner.
Because no one is making the connection that you've made, Alan, that I've made, that Fred has made,
that there's a connection between affordability and how you actually deliver that ride. Absolutely.
People are saying, we're going to have cars that aren't going to crash. That has absolutely
nothing to do with affordability because what we've seen over the course of the last 50 years
is the more crash, crash, avoidability, the higher the cost of the car. And in the end,
that directly affects the cost of the ride. If you're trying to deliver rides to people who
can't afford to drive a car, can't afford to operate a car, or even can't drive a car for
whatever reason, the benefit of driverless cars is that you've taken out the cost of the driver.
This sounds so logical, but no one is getting it, including the politicians in New Jersey
who are trying to put laws into place that prevent real testing of cars that would deliver
affordable rides. They don't care about affordability. Wait, wait, wait, wait, wait. I do want to,
in my opinion, and it may not be the opinion of you or the opinion of Fred, in my opinion,
the cost of the driver is the driver itself.
The cost is really not that much. If you can get that driver to really be productive.
And if you can tell that driver to turn on a dime, the problem is, is to get the driver to turn on
a dime and be there when you need it and really deliver that. Because what the giving of the ride
function is, it is uncertain. It is not planned. something that you can set up a timetable
three years ago and run a schedule. Giving rides is an extremely
called stochastic thing. And to be able to do it efficiently, you've got to be able to turn on
a dime. And unfortunately, to get people to turn on a dime, you have to beat the crap out of them.
They can do it for a while, but eventually it's brutal. It's inhuman. It is just what?
The beauty of a machine, as you can tell it, turn on a dime and it will, because it follows orders.
Boom. And that's the real value here. But it sounds like instead of promoting
promoting driverless because it's affordable, because you've taken the driver out of the task,
what you're saying is, well, if we replace this car with a machine that we can direct,
you've decoupled it. And I don't want to do that. I want to make it very, very clear that there are
multiple advantages of removing the driver from the task, but they all come with the same result.
It's less costly. We agree. And we agree. If you come back at me and say, well,
wait a minute, it's really the cost of the driver. The cost of the driver, just the cost of the driver
is 60 to 70% of delivering a ride. That's not insignificant. The cost of having a human being
being paying the person, we know that because we're doing this today with handy rides.
We're paying people to be there. But the cost, the cost of paying the person to be there
is only one of the small costs of having that. You've got to have a whole HR stuff.
You have to have a whole benefits package. You've got to be able to have somebody ready to go in
case that person isn't there to do it. That's all built into the cost of the driver. I'm not
saying it's just the person. It's everything. I know. But drivers will tell you, oh my goodness,
I don't get that much. How could I be that? But what is not realized is to deliver the ride.
You don't need one person. You actually probably need three.
If not more. Whereas, holy hell, if you have some code and have really efficient code,
and of course it's not, boom, the code that does it and it doesn't bother turning a wheel. It just
goes and gives you the little old thing that takes the motors that turn the wheels
and takes the motors that do this to the brakes. I mean, boom.
One of the things that we have had to learn, and I think we're getting closer because it seems like
the politicians are at least beginning to pay attention to this. You've got to make it as
simple as possible. If you complicate it to the extent, if I say removing the driver reduces the
cost by approximately 60 percent, I'm not saying that it is just that particular driver.
And if you take that particular driver out, everything else falls in place. I'm saying
that the driver has cost. If you remove the driver, those costs, which include the HR,
which include the time off for having breakfast and lunch and dinner and everything else that
you need to do, all of that comes out. Yes.
So that's the beauty. Well, sure. I don't think that it's realized
how much more productive a vehicle can be if you have a machine, an algorithm telling you what to
and an algorithm that is sensing the demand and is responding to the demand in real time.
Okay. So that it gives the rides, the people, the rides that they want, that they want now
to go there. Yes. Not because I happen to have a bus route and it's going to go downtown.
Who now wants to go downtown? Nobody. If a senator or a representative state
federal listening to this conversation, it sounds like we're not, it sounds like we're
not in agreement. We are
We're in two ways. I just want to make sure that people go keep going. Waymo is not saying that.
What Waymo is saying and the reason where we are, where we are right now is that Waymo hasn't
defined its motive. It hasn't given us its real mission and it certainly hasn't established for
us what its method is because it says, I want to go out there and get money for delivering rides.
But then when you ask them, what are you actually doing? We want to be the most trusted driver.
So this came out in the presentation and in the seminar. If you think your job
is to deliver rides, why aren't you approaching this problem from a ride delivery
point of view as opposed to from a technology point of view? Why aren't you out there talking
with people and saying, this isn't, you're not going to eliminate all of the deaths on the
highways. And by saying that, coming out and saying, we're 95% safer, 99% safer, 92% safer than
human drivers. And all you get back from the people is, well, okay, but what am I going to do?
Am I going to buy a car or am I going to depend on giving rides? Are you giving me rides?
Are you going to be doing this for me everywhere? They don't answer that question
because they know the answer. The answer is, we're going to sell our technology to companies
who will go through the type of approval process, who will get their cars on the
market, who will sell those cars either to individuals or to companies who want to buy
1,000 or 2,000 or 10,000 cars, who will deliver rides. That's not our job. Our job, Waymo,
our job is to make the technology work so that you, car company, will be able to sell those cars.
We're not in the car company business. We're not in the car selling business. We're in the
most trusted driver business. Now, if only they would come out clearly and say, this is what
we're doing, here's our motive. What our motive is, why are we doing this? Because we want to
have our technology in every car so that we can sell you more advertising. That's the real motive.
We can't get cynical here because... I'm not cynical. This is not cynicism. This is exactly
what they're doing. Why won't they admit it? Because they're a private entity. They don't
care about us. They don't have to tell us anything. I mean, what do they care about us? They don't
care about us. I mean, they've just been crushes, trivially. Okay? Well, I think in the interest
of Michael getting some sleep tonight, first of all, you guys have given a brilliant synopsis
of your book, The Real Case for Driverless Mobility. All they have to do is listen to
what you had to say here. Well, look, Fred, guess what? That's what we're doing. We're selling our
book. We expect... Let's see. Some revenue to me over that time frame is... Let's see. What's the
number? In my class, I always tell them, if I ask you a question, I'm going to frame the question
so that you know the answer, everybody. So everybody in that class knows the answer to
all my questions. It's zero. So I always frame a question so that zero is the right answer,
so that they don't have to cheat. They know the answer even before I ask a question.
As I was saying, in the interest of Michael getting back to bed at some point before the
sun really comes up again. Or before we show up on the same day as he.
In the latest newsletter, and Michael was already alluding to this, you link to a recent
article from Electric by Fred Lambert that was headlined, New Bill would ban Tesla's camera-only
RoboTaxi, not Waymo. And he is writing about New Jersey, but it seems a bit late.
Well, I mean, my goodness. I mean, first of all, I mean, to relate it to what Michael just said,
he implied... If what he said is true, that must mean that Waymo is selling
the lidar that they developed and making sure that's sold in New Jersey.
I mean, really, Waymo, that's what you're doing. I hope that's not what you're doing.
And in fact, not what you're doing because he wrote that one on July 6th.
The bills that he's claiming that were adopted in New Jersey, or I don't even know what he claims,
those things, one was pulled by the Senator who was writing it on what, the 22nd of June,
and the other one was pulled by the Speaker of the Assembly without comment on the Assembly version.
So what it... Fred, I thought, Fred, I thought you were good. I mean, where do you get your
information? I mean... I want to clarify your buggy back, Fred Lambert, correct? Oh, yeah. Oh,
yes, Fred. I'm not talking about Fred Fishkin. No, this is Fred. Fred, I thought you were good.
I used to quote you all the time and so on as a source of real information. Where'd you get this?
Come on. I mean, what... I mean, you got to talk to the folks in Jersey here. I mean, yikes.
I mean, what do they think we are? I mean, they think we're Greenland or something,
or Venezuela, or I mean, whatever, the 55th state, we were part of the original 13. I mean,
I mean, we even had the... Congress here in a couple places. I mean, we were actually,
we were the capital, a couple places in Jersey with capital, the USA at one point.
I mean, cut it out. We aren't just stepping stones to New York. I mean,
we might be their bread baskets or their garden state, but come on, we're not. They're stepping
stones. Hopefully, there will be some new legislation coming forward, but we'll see.
Yeah. And I think, and we... If we could only survive, which we're probably not going to be able
to survive, but it does look like that finally, there is interest in maybe writing some sensible
AV legislation in New Jersey that looks at the real value, at least what Michael and I and you,
Fred, have been talking about for what, I mean, 14 years, who knows, is quality of life aspects
of this thing. The opportunity of this technology to provide rides to the latent demand of our
society and let those folks who end up having to stay at home because they can't get a high quality
affordable ride to go get their hair done, go to school and learn, go get some decent food,
go to a doctor, go to work and punch in on time every day, even on the midnight shift
and not get fired because they couldn't get there. They couldn't bum a ride. They couldn't,
I mean, the value proposition to these folks that allow them to elevate themselves
because of the value proposition of mobility, that's what transportation's all about.
And when art goes and says, oh my goodness, mobility and transportation worldwide is worse,
what did she say recently? $10 trillion, where the substantial percentage of global GDP,
it's because of the quality of life enhancement that this mobility could do over what people already
have today. And some people, if you have a Lambo or a voodoo doodoo
you can, I mean, you know, and have 18 different ways to get to the airport already, I mean to have a
19th. The people we've been given rides to, they don't care about the technology.
They care that, oh my goodness, it took them for me to be the time T when they wanted to go.
And most importantly, the thing was going to be there to bring them home.
Why do people use their cars today? Not to get there.
So they can get home.
Why did Hansel and Gretel put the crumbs down to Grandma's house
so they can get home?
Now with these things sitting around, just waiting, begging, doing nothing,
doing nothing. That must be a Pittsburgh term or something, doing nothing.
I think New Jersey has a not for nothing going on.
Maybe New Jersey has nothing too, doing nothing and bring them home.
That's societal value. And hopefully the Trump administration will see it. I think Duffy said it.
Saval Oz has said it.
The Speaker of the Assembly in New Jersey saw our letter,
pulled it. I think he sees it. Maybe some people are finally seeing it.
And, you know, if we can get through the regulations and we get the damn technology
to work and we can actually do it, first in the small for a few, why not do it for a few?
Only hell, we could do this in South Jersey and rural New Jersey.
South Jersey, you know, they talk about buses in North Jersey and complain about them and they
say, oh my goodness, the buses in North Jersey are late. That's what people in North Jersey say.
People in South Jersey complain about buses. They say, we can buy them. Oh, that's right.
We don't have any buses. I mean, in rural areas, there's an awful lot of there's not even a
challenge, but the safety is not a challenge. Stuff works. And why we're not putting it to work
to provide, I mean, I mean, come on, folks. But how long have we preached this, Michael?
Yeah, well, here's here's the last the preaching that I did when I was in Geneva at the AI,
this AI for good conference, the question. Yeah, yes, right. That's I didn't make up.
testing to deployment. And I took it seriously. As one of the four panel members,
I took that seriously. And I said, based on what we've learned during a proof of market test,
what deployment will take as the following one regulations that allow driverless vehicles on
the road, the people establishing the regulations for what is street legal need to deliver the
requirements, which the vehicle builders must follow. WP 29 is working on the standards.
Governments have to have begun to focus on creating the right types of laws and regulations.
They need to complete their work. To vehicles that are street legal. The people building
driverless cars need to deliver driverless cars. People building driverless cars,
not the technology for making cars driverless. The people building driverless cars need to deliver
on their promises of safe operation above all following the regulations that are established
for their operation. There was not one car company at this conference speaking. There may have been
some somebody from Mercedes or Volkswagen hiding in the audience someplace, but there was not one
person from a car company speaking. And I made sure that I made that point, especially to the
panel on which they were discussing how you apply regulations and the only people on that
panel were from companies like Waymo and Waze. Third, a legal framework that is based on the
fallibility of all humans and the imperfection of all technologies. Driverless vehicles will
not eliminate all road fatalities and is a dangerous dream to believe they will. It is
irresponsible of companies promoting driverless technology to claim that their robotic systems
are perfect and infallible. Driverless vehicles can help to provide mobility to a large number of
people who otherwise would not be able to fulfill their realizable dreams of improving their own
lives and the lives of the members of their families. Now, when this session finished,
I was approached by one person and that one person was so enthusiastic, but it was only one
person. I wasn't mobbed by everybody sitting there because I think this presentation hit them
between the eyes and they weren't paying attention because it didn't talk about safety. It didn't
say how much money they were going to make by removing drivers. It didn't tell them they were
going to get rich. It said if you remove the driver from the driving task, you're going to be able to
make the drive the ride for people who can afford mobility affordable. But this one person
convinced me that I should give him a copy of the book. He was prepared to buy it, but I said
you're so enthusiastic. Here's a copy. Please read it and tell me what you think. We need to put our
words that we're talking about. We need to put them in the hands of people
and get them to take the message, to understand the message. People who are putting, promoting and
proposing legislation in states and countries that are focused on something that isn't really the
main benefit of having driverless cars is just moving this, moving the bar further and further
and further into the future. We wanted to move it further and closer into
the present. That's what we need to be doing. Well, Michael, the only thing I'd add is that
we also have an audio version of the book and the audio book is talk about affordable. Much
less expensive. Oh, my goodness. I mean, the other version, never mind Elsevier, we have nothing
good to say about you, or at least I have nothing good to say about you. A new publisher is in the
wings, right? Well, for the new book, thank you. But boy, do we have an audio version. Who did the
audio version? Fred. I mean, Fred, a mollifluous voice to good thing. I didn't do it because
nobody I mean, I had COVID when I read that, Alan. So well, Fred, Fred, he did a marvelous job.
So yes, look, look, I want to wrap up with one thing because we've talked a lot about
Waymo today. And in the newsletter, you have something positive to say. I do. Yeah, I think
you do. A TV report from San Francisco that said Waymo called San Mateo police on teens for drinking
and shooting toy guns from a driverless vehicle. It stopped the ride and called the police. And I
one of the things that we haven't said loud enough and have maybe said is that that if these
systems driverless vehicles end up giving high quality affordable rides, there has to be a
pro quo by those getting the rides. They have to behave. What is ruined driving? It's the misbehavors.
Why do we do? Well, I think it'll be a long time before we make any advancements in safety
of the roadways is because the misbehavors may well be the last people to give it up.
And therefore, they're the ones that are causing the problem.
They're the ones that rear end you go through red lights, go down a cul-de-sac at 73 miles an hour
and ram into a house, drop their phone and reach down to go grab their phone. And let's say, hey,
the dog ate my homework. The misbehavors kids riding in a vehicle that Waymo is trying to give them
at least a good ride. And they're out there using it to become influencers on, you know,
the inner tubes. I mean, come on, kids.
New York City subway has had a problem for years with kids surfing the subway
to be able to take a selfie of themselves.
I don't know. Maybe if I was, you know, 13 or 14 or 15, maybe I'd probably be, I don't know.
My behavior was not really very good. But my goodness, you can have kids hanging out of windows.
And that's why, back to my supervisory function, there has to be a supervisory function in there.
It's able to tell that the kids are hanging out the windows, taking selfies of themselves to
become influencers on, you know, what has become. Where's my phone? I mean, I really wonder whether
or not the phone is net positive. I mean, it's getting to be a point where,
because all this stuff has run to the lowest common denominator. You got to. Okay. And basically,
with respect, I mean, in terms of the way we approach, and if we end up still staying trying to
thrive, right, you're probably going to get the misbehaved once and probably will say,
okay, dog ate your homework, you'll get the misbehaved a second time. You're not going to get the
misbehaved three times. Sorry. As I tell my students, United Airlines would not let me
anywhere near Newark Airport. If they thought when I picked, oh, I think I'll take the exit
row right next to the window so I can help people in case and open the window and help people out
if there's a crash and I can help. If they have any thoughts that I might at 30,000 feet, oh,
does this open? Oh, let me take a selfie. They wouldn't let me anywhere near that airplane,
anywhere near the airport. When I get on a plane, I got to behave.
Otherwise, you don't get on the airplane. They just say, sir, come with me.
And they should,
because it's so easy to misbehave. It's so easy to think you're going to become,
or you're going to get your 15 seconds of pain. And then you can tell your friends about it.
And you're ruining it for everybody. The tale of the behavioral distribution,
that tale of misbehavours is the reason for 40,000 people dying. It's the tale of the distribution.
It's not the bulk. It's the few. And the few can easily ruin this. Because,
I mean, you better have a pretty darn good supervisor in there
who supervises the behavior. And we do. There'll be cameras all of them. There'll be data.
I mean, I don't know. There was one legal case in which, even with all the data shown,
I mean, the jury said, well, give the people the money.
So the ambulance chasers in this thing do a pretty darn good job of making sure they get
their 40%. And they're going to be there. We've got to figure out how to deal with them.
And hopefully they won't ruin it. But I mean, insurance fraud in the current,
automobile thing, plus the misbehavours, makes you want to stay home the hell with it. I'm not going.
So, yes, thank you, Waymo, for, you know, and my goodness, you know, I'd say prosecute to parents.
Parents have a responsibility here, too. And of course, my kids did nothing but
exemplary behavior because they're such a good parent, of course.
And on that note, I think it's time to tuck in Michael, let him get back to sleep.
Thank you for being with us again, especially in the middle of the night for you, Michael,
there in Sweden. The place to go to read more is MichaelElsenna.com. Thanks, Michael.
Pleasure to be here. Thank you. And Viva France on Bastille Day.
Yeah, I'm neutral on that one. What do you mean neutral? You got to go, I mean, come on.
They've got two more games. If they make it to the second round, they've got first,
they've got to get past Spain. But we'll see. And for England, things could get messy. Can we
say that? Yes, exactly. Things will finally, I mean, come on. I mean, how is that gold disallowed?
Get out of here. I was, for Sweden, that wasn't, was that Sweden or is there a difference between
Sweden and Norway? We're gonna stop and close on that note. Stop them right here. Okay.
You can find more at the newsletter at smartdrivingcar.com. You can find my
tech reports at techstination.com. Thank you again for taking the time to watch or listen.
And please stay safe.
About this episode
Regulations for driverless cars are finally taking shape, but the hosts argue the gap between “test mode” autonomy and street-legal, commercially deployable vehicles is still huge. They walk through how UN WP 29 and U.S. NHTSA/FMVSS frameworks differ, including type approval and special rules tied to low-volume sales. The discussion also splits fast 20-millisecond control from slower supervision, questions safety claims and business motives, and highlights real-world issues like rider misbehavior.
What will be the impact of the latest WP 29 regulations on driverless mobility? Consultant and author Michael Sena joins Princeton's Alain Kornhauser and co-host Fred Fishkin for episode 416 of Smart Driving Cars. Plus the path forward in NJ for driverless, Waymo, Tesla and more. Tune in and subscribe!
