Deborah, I just got punched by another guy who didn’t like me taking movies of him with my Google Glasses. It really disrupted my day.

Do you think they’d let me trade them in on a Super Duper Disruptive Self-Driving Car?

Some auto industry exec was on Bloomberg talking about self-driving cars and predicted that the first step would be cars that can park themselves. So, you pull up to the store, get out, and tell the car to go park. Then use the smart phone app to call it to pick you up. Valet parking everywhere you go.

You could imagine retail outlets like malls competing with auto-valet friendly parking lots that have parking spots with special facilities to accomodate the vehicles, such as reflective strips to follow, and electric chargers accessible without human hands.

My wife wants one. Usually, I drop her off at the door and go park at the far end of the lot and walk heroically through rain and howling wind. Guess she won’t need me anymore…

We havn’t even begun to understand the scale of disruption which will be achieved by autonomous self driving electric vehicles. Here are some of my predictions.

1. Eventually, most people will choose not to own a vehicle – prefering to order one when required.

2. The number of vehicles on the road will fall dramatically.

3. The average size of vehicles will fall significantly – many people buy a relatively large vehicle for the two or three occasions a year where the space will be used. The rational for this will disappear.

4. The amount of road space needed per vehicle will drop even more than the size of the vehicles as computers will be able to safely manage vehicles driving closer together than humans. As a result, traffic congestion will be greatly reduced.

5. More people will choose to car pool as all users will be identified and monitored at all times making travel with strangers much safer than it is now.

6. The number of parking spaces required will likewise drop dramatically resulting in far more space for pavement cafes, flower beds, play areas, social spaces etc. What parking spaces remain will for the most part be much smaller as autonomous vehicles can park a few centimeters apart in all directions – nobody has to get in or out of a parked vehicle (only one in a loading bay). It will not matter if a vehicle is blocked in as any other suitable vehicle can be dispatched to do the job, and even if a particular vehicle is essential, it can coordinate with other vehicles so they move to let it out.

7. A dramatic increase in the number of electric vehicles will carry with it a huge capacity to buffer the power grid – especially once vehicles can arrange their own charging and defer, or bring forward the charging process to help the grid balance supply and demand. In an emergency, any one of these vehicles will also have the capacity to help keep essential electrical services running e.g. lights, communications and refrigeration in critical facilities for several days – imagine how useful this could be after a hurricane, flood, earthquake, or winter storm.

8. Disabled and frail people, those who are drunk, on strong medication, and children (subject to parental control) will be able to access transport solutions without requiring a driver.

9. Eventually, all the sensors built into vehicles and attached to intelligent systems may be able to monitor our towns directing police and emergency services to trouble spots early so allowing rapid intervention before things get out of hand, gather evidence for any proscecution, or direct help to those who need it. (Or even verbally address trouble makers reminding them that they are being filmed, and will face the legal sanction if they do not stop making trouble).

In addition, it is likely that such vehicles will eventually contain biomedical sensors to identify medical emergencies in their passengers and obtain help without human intervention.

Gary, here’s my prediction for the scale of disruption of “autonomous, self-driving” cars, which have been around since the 1950s.

They will disrupt only the bank accounts of investors, and for a very simple reason: when my autonomous, self-driving car runs over my neighbor’s kid, whose fault is it? I have yet to receive an answer to this disruptive question from legions of enthusiastic supporters.

Bob,

Not such a difficult question. 

If you own the self-driving car, not sure why you would own it yourself, you would have to own insurance just like you do now and you would be liable.

However, since the self-driving car would be much safer vs. you behind the wheel the insurance would be cheaper.

In the first instance, it will be your fault. If you can prove that the car failed to perform as promised by the maker, you might be able to pass the buck to the maker, but since they have lawyers that work full time on liability and you likely don’t, the short odds bet would be that you will end up being liable.

“4. The amount of road space needed per vehicle will drop even more than the size of the vehicles as computers will be able to safely manage vehicles driving closer together than humans. As a result, traffic congestion will be greatly reduced.”

This mostly implies that traffic will flow more smoothly when not congested, but will not eliminate congestion. The maximum speed of a fully congested street network will not be substantially higher than it is with humans driving, though with less stop and go driving and more driving at a controlled low rate of speed, it would be more energy efficient.

Joe, in my opinion you vastly underestimate, as does Google, the complexity of the decision making process you undertake when you get behind the wheel. Here’s what will happen: the cars will perform wonderfully in their idealistic test situation until one of the fifteen million “unpredictable” situations which Google engineers somehow never anticipated will happen and result in tragedy. Then another will happen. Then several more.

After paying out many $millions in damages, Google will make it safer by increasing the car’s sensitivity to objects in the car’s radar field. The car will slam on the brakes for a baseball rolling out into the street, or a balloon. You will be rear-ended by a human driver who never anticipated your stupid car would slam on the brakes for a balloon (yes, it’s their fault, but you’re the one in the hospital).  It will  trail the car in front by several car lengths farther than you would, and take twice as long as you would getting to where you need to go. You will say &?!* this, give me back my 2007 Acura, please.

This whole episode is comically reminiscent of a period in the 1960s when a segment of the population was equally certain everyone would be piloting personal helicopters by the 1980s. And no, they will never be safer than me, nor you, nor anyone else on this board.

“And no, they will never be safer than me, nor you, nor anyone else on this board.”

Bob – You guys must have much better drivers in SoCal.  When I drive to work in the morning about 1/4 of the folks are driving distracted. Talking on phones(not hands free), texting, playing games, doing makeup, eating breakfast with one hand , drinking coffee with the other.  

Also – although Google has done some interesting work in this area they are not who I would look to for leadership in this area. 

It might be Volvo- see here.

• “Volvo made its self-driving ambitions clear Thursday with the announcement of a pilot program that will put 100 autonomousc cars on Swedish roads by 2017.”

It might be BMW- see here.

It could be Ford – see here.

• “You can go into a dealership and get a Ford Focus that can park itself right now,” Raj Nair, Ford’s product development chief, said of the automaker’s compact car that starts at $16,810. “If you want to go to the full extreme — full autonomy — literally a vehicle that has no steering wheel and has no pedals, that’s a tremendous technical challenge, but one that we believe that in the next five years will be possible.”

However, I would bet on Tesla. see here.

• “Mr. Musk said on Thursday that Tesla had been testing its autopilot on a route from San Francisco to Seattle, with company drivers letting the car navigate the West Coast largely unassisted.
• After the software update this summer, the cars can also be summoned by the driver via smartphone and can park themselves in a garage or elsewhere, he said. That feature, though, will be allowed only on private property for now, he said.”

I do think that people talking about fully autonomous cars by 2020 are being way too optimistic. I also agree that there are definite legal and perception issues that will need to be overcome.

However, just like solar energy  – self-driving cars are improving very rapidly and offer many benefits.  I am betting that by 2040 we will have areas – possibly even entire towns/cities(maybe Singapore?) – where human driving will be illegal.  

Joe, one more observation.

You and other posters are quick to equate fault with liability, i.e., moral with financial responsibility. But say your car, with its incredibly-low insurance premiums, strikes and kills a child on a bicycle. Google acknowledges a software glitch, announces they will pick up the tab, and showers the child’s distraught parents with money. You’re not out a dime.

But who is at fault – Google, or you for trusting someone else’s safety to an automaton?

Bob,

Again you have the focus of “your car”. Maybe that is your SoCal coming through – you guys do like your freeways.

 I recently bought a new car and I hope it is my last purchase. I don’t want to want to own a car. Spent a long weekend in SanFrancisco recently and took Lyft/Uber everywhere. That is the future. Just get me from Point A to Point B.  I’m done with driving.

By the way, eventually there will be many neighborhoods that will be designed to not allow cars. So that child on the bicycle won’t need to worry about getting hit two blocks from her home.

Does this include 3:00am trips to the ER?

Paul,

Not at all clear what you are asking here…

– Are you saying that you would want to have your own car so that you could drive yourself(or family member) to the ER at 3:00am?   Personally, I would much prefer an ambulance service or Med-Uber.

– or are you saying would I want to be driven to ER by an Autonomous car? Of course, it will be much safer.

Joe, whether the operator of the vehicle is you or someone else is irrelevant – I want to know who’s going to accept responsibilty when someone’s hurt.

btw, I don’t like driving on the freeway at all, but I do ride my bicycle quite a bit – about 4,000 miles/year. In GoogleUtopia, will I have to ride around in circles in your safe zone?

I live about 5 miles (max) from the Hospital. My son once stabbed his his wrist accidentally. Lots of blood, screaming, fear and anxiety. Fortunately I have a car.

If I had waited for an ambulance, not knowing if any nerves were damaged, I would not have forgiven myself. Further, the cost of the car ride to the Hospital was maybe 5 cents, an Ambulance would have cost several Hundreds. If I had some hundreds to spare, I’d rather use it for the $150 ER Co-Pay.

Having your own ride is handy for many other reasons. I could do grocceries whenever the need arises. I could take my wife out for dinner on the spurr of the moment, we could decide, last minute to watch a movie and  still get there on time. I also have more confidence in my own driving skills when there is an emergency. I certainly don’t want to be driven at 25mph in the midst of an emergency.

In North Dakota where I live, the Red River used to flood, and some years past, folks had to evacuate. I wonder if Lyft/Uber ride has room for some 300 or so rsidents.

I would need to wait and see how the Google ride driverless vehicles work first before I decide if I like it or not, although there are situations where it could be a God Send.  But I doubt I’d ever trust my family transportation needs to an outside vendor.

Bob, The entire premise of autonomous cars is based on them being safer.  Because much of the general public is fearful of any new technology they will probably have to be a LOT safer.  Autonomous cars should be able to dramatically reduce deaths and injuries. If they can’t, they won’t happen – its that simple.

Same here on the bicycle front-  I ride – about 3,000 miles a year. In the self-driving car utopia bicycling will be way more prevalent.There will be thousands of more miles of dedicated biking “streets”. Lanes that are actually physically separated from any vehicles. Entire streets that are now used by cars will be switched over to bicycles.

People will be much more likely to ride their bicycle to work and on other small trips – where they now take a car. If they want – they can always “call” for an Uber/Lyft/Tesla car to pick them and their bicycle up for the ride back home. It will cost a little extra to get the car/van that can carry a bicycle. No worries about rain, a flat tire, needing to be home earlier… etc.

Deborah, I’m curious how it’s known they were “caused by humans”.

Perhaps humans were legally responsible for an accident which would have been easily avoidable by a human driver in the place of the Google-automaton.

In any case, racking up an accident every 217 miles is an abysmal record. Any fatalities?

Bruce, if my autonomous car runs over my neighbor’s kid, it’s fairly obvious it has failed to perform as promised.

Or possibly there will be a disclaimer: “May cause occasional fatalities. By trading others’ safety for the ability to check email/text messages/voicemails on the way to work/school/the airport, Driver accepts full responsibility for any injuries sustained to himself/herself or third parties, including cats, dogs, and the occasional pigeon.”

I’m starting to warm up to this concept. Those suffering from paranoia could benefit from a Personal Security Assistant – a robotic gun mounted on a helmet – that would “neutralize” any passerby it determines to be a threat. Muggings, rapes, and robberies would be things of the past.

Really, the possibilities are endless.

Just as a fun fact, as of 2014, Google’s autonomous cars had completed 700,000 miles without an accident.

The other factoid is that this feat requires around $150K of fancy electronics and sensors.

This is presumably why the first applications will be specially adapted environments like parking lots and garages, where you can make one investment in reflective strips, transponders, etc and simplify the vehicle.

Self-driving car

“I’m curious how it’s known they were “caused by humans”.”

From the wiki:

In 2010, an incident involved a Google driverless car being rear-ended while stopped at a traffic light; Google says that this incident was caused by a human-operated car.[28] In August 2011, a Google driverless car was involved in a crash near Google headquarters in Mountain View, California; Google has stated that the car was being driven manually at the time of the accident.[29]

“…every 217 miles is an abysmal …”

The Google car logged not 700 but 700,000 miles.   

I think you identify valid concerns with autonomous vehicles: who’s liable, and that they sooner or later will have a lethal accident due to a software bug or the like.  These concerns should be put in context: the autonmous vehicle will never fall asleep at the wheell, will never get distracted in an intersection, will never feel the need for some extra speed on slippery roads. 

I helped design one the vehicles competing in Darpa’s 100 mile autonomous vehicle competition through the Mojave.   Even in 2005 the state of the art had progressed much further than even most technical professionals imagine. 

Deborah, flying has a fraction of the interaction which upon which safe driving depends (at least in an urban context). How often have you changed lanes to avoid a driver who, from experience, you know is either drunk or distracted by his actions? How many times have you stopped in mid-right turn because you know the guy on the skateboard, on the sidewalk next to you with earbuds in his ears, is going to cross the street without looking? How many times have you slowed to let someone into your traffic lane because they stick their hand out of the car?

These just a few of the subtle visual and aural cues which, like handwriting, computers are incapable of analyzing to the accuracy of a human being. Much less, the ability to make snap value judgements:

“Consider this scenario, spelled out by Jason Millar in the September issue of Wired:

You are travelling along a single-lane mountain road in an autonomous car that is fast approaching a narrow tunnel. Just before entering the tunnel a child errantly runs into the road and trips in the centre of the lane, effectively blocking the entrance to the tunnel. The car is unable to brake in time to avoid a crash. It has but two options: hit and kill the child, or swerve into the wall on either side of the tunnel, thus killing you. Now ask yourself, Who should decide whether the car goes straight or swerves? Manufacturers? Users? Legislators?

Google is as yet unprepared to answer that question.

“People are philosophizing about it,” Ron Medford, the director of safety on Google’s self-driving car project, told The Associated Press last month, “but the question about real-world capability and real-world events that can affect us, we really haven’t studied that issue.”

Possibly that should be the first issue of study.

Secondly, over-reliance on automation has resulted in many flying accidents including the 2013 crash of Asiana Flight 214, in which three people were killed and 187 injured:

The final report into the crash was released on June 24, 2014. The NTSB found that the “Mismanagement of Approach and Inadequate Monitoring of Airspeed Led to Crash of Asiana flight 214”. The NTSB determined that the flight crew mismanaged the initial approach and that the airplane was well above the desired glidepath….Over-reliance on automation and lack of systems understanding by the pilots were cited as major factors contributing to the accident.

Automation of cars will serve as a helpful convenience for simple chores like parking, or long distance highway driving (it already does). What you should be fearful of is the growing tendency to use automation not as a beneficial tool but as a responsibility dodge, or to save money. As the above accident shows, that’s already permitting airlines to put undertrained pilots in control of the fate of hundreds of people, sometimes to disastrous result.

Thanks for the correction, Mark, I mistakenly interpreted 700k as ‘700km’.

An eyewitness account of the 2011 accident, which involved five cars:

Google’s Prius struck another Prius, which then struck her Honda Accord that her brother was driving. That Accord then struck another Honda Accord, and the second Accord hit a separate, non-Google-owned Prius.

“Google has stated that the car was being driven manually at the time of the accident” – although the company also stated they only hired drivers with spotless driving records.

What a coincidence.

“These just a few of the subtle visual and aural cues which, like handwriting, computers are incapable of analyzing to the accuracy of a human being.“

With respect to driving, that’s no longer the case.  

https://youtu.be/PXQlpu8Y4fI?t=1192

https://youtu.be/eXeUu_Y6WOw?t=48

http://www.telegraph.co.uk/news/science/science-news/11410261/Driverless...

https://youtu.be/lPG0hzoKIto?t=147

Mark, where specifically do your videos demonstrate a vehicle – not on a race track, but in traffic – analyzing its surroundings with greater accuracy than a human being?

I’m not going to sit through an hour of Sebastian Thrun justifying his salary at Stanford with hyperinflated promises.

Bob – Apologies for the overflow of links.  See traffic example here with pedestrian in crossing at 48 secs into the clip.

https://m.youtube.com/watch?feature=youtu.be&v=eXeUu_Y6WOw&t=48

I would be okay with self-driving cars if they had steering  wheels, brakes and accellerator pedals, and if they didn’t overide me when I overide them.

I would hate for a car to drag me at 20 mph when my wife is in the back seat hoping to get to the ER for emergency surgery on a burst appendix.

As I recall Google’s vision of self-driving cars don’t include stearing wheels.

Comments on this thread are perfect examples of why self-driving cars will have to be at least 5x safer vs human drivers before they are allowed. Plenty of Luddites out there.

Meanwhile:

Old drivers

http://www.hngn.com/articles/70209/20150217/wisconsin-driver-92-crashes-...

Drunk Drivers:

http://www.9news.com/story/news/local/2015/01/26/4-killed-in-weld-county...

No kidding Joe. Certainly the creators of Google Glass couldn’t be wrong about something as foolproof as cars driven by robots, which will somehow prevent drunk and old drivers from killing other people. Google hasn’t studied that issue yet, but you can bet they will.

I put a deposit down on a personal helicopter in 2002 and now the company’s phone is disconnected, thanks to Luddites like the ones on this thread.

http://www.newscientist.com/article/dn1793-personal-helicopter-gets-off-...

Mark,

Thanks for the links. The progress since 2007 has been nothing short of amazing.

I like the racing idea.  

I’m guessing that by 2020 we will have a full-blown race with 15-20 autonomous cars racing around Laguna Seca.  

It’s also interesting to see the parts suppliers getting into the game on the testing front:

http://www.autoblog.com/2015/04/03/delphi-thrilled-with-results-from-aut...

and making things rapidly cheaper.

http://blogs.wsj.com/digits/2015/04/21/quanergy-says-it-will-offer-low-c...

Assuming this comment is deemed to be a reply to Deborahs concern about automated flight, i’ll try to do a better job than Bob to explain the issue:

On the road, lateral error tolerance il mostly less than the vehicle’s width and the consequences of greater deviation to either side are most dramatic, especially to the left in bidirectional traffic. Another critical aspect is the necessity to drive along in files, with high risk of telescoping. Last but not least, level crossings are a frequent source of lateral collisions with extreme risks due to poor lateral protection offered by the car body.

In the airspace, neither is deviation from the flight path an immediate risk because there are ample separation distances both horizontally and vertically, nor is there any necessity for flying in files, i.e. each aircraft can claime its own virtual highway — as to level crossings: flying above or beneath a crossing aircraft doesn’t need the construction of bridges or tunnels, just software to structure the airspace…

Minimizing the consequences of road collisions means ever increasing vehicle weight and size. Mid-air collisions, while being very rare, are mostly fatal, thus the aim in aviation is not to diminish the consequences of a collision but to avoid it… because you can easily avoid collisions either by pulling up or diving, or pulling to the left or right — which his is why the exact contrary is true in the air: not ever heavier and larger, but ever lighter and smaller vehicles because the lightest and the smallest will be the swiftest and the most ractive to avoid collisions.

Think of all the savings in terms of road building and maintenance costs, as well as material costs of the vehicle, and please don’t forget the apalling costs of the annual road toll… 

Bob,

Too bad about that investment.  Could have told you that personal helicopters weren’t gonna work out if you had asked.

Here’s some more recent news on that other “fantasy market” – autonomous vehicles. You might want to stay away from the highways in Nevada.

http://www.freightlinerinspiration.com/newsroom/press/inspiration-truck-...

“The Freightliner Inspiration Truck is the first licensed autonomous commercial truck to operate on an open public highway in the United States. Developed by engineers at DTNA, it promises to unlock autonomous vehicle advancements that reduce accidents, improve fuel consumption, cut highway congestion, and safeguard the environment.

The Freightliner Inspiration Truck underwent extensive testing before the Nevada Department of Motor Vehicles granted it a license to operate on public roads in the state. Earlier on Tuesday, Nevada Gov. Brian Sandoval formally granted the license to operate the vehicle in the state, affixing a license plate to the truck and taking part in the ceremonial first drive of the truck in autonomous mode.

The Freightliner Inspiration Truck operates on highways at what the National Highway Traffic Safety Administration defines as Level 3 of autonomous vehicle capabilities, enabling the driver to cede full control of all safety-critical functions under certain traffic or environmental conditions. The autonomous vehicle system is responsible for maintaining legal speed, staying in the selected lane, keeping a safe braking distance from other vehicles, and slowing or stopping the vehicle based on traffic and road conditions. The vehicle monitors changes in conditions that require transition back to driver control when necessary in highway settings. The driver is in control of the vehicle for exiting the highway, on local roads and in docking for making deliveries.”

More here.  Funny – no mention of Google.   Just some company called Daimler.