Google has shown the world that they have a car that can do a pretty good job of driving by itself in simple situations. Stanford has shown us a car that can race through the Pike’s Peak race course autonomously. In the next 10-20 years, what can we expect from autonomous vehicles? How will the law deal with accidents caused by autonomous vehicles? What countries or states will pioneer this field?
Written By: Keith Kleiner
Posted: 05/5/11 9:36 PM
Cars are already driving themselves and yes Automated Vehicles (AV) will catch on like wild fire within the next 10 years. I do not feel that the legal aspect will be big of an obstacle as others believe for one simple reason, this tech will be a huge benefit to business. Once the big money realize the cost savings benefits that this tech represents, doors will be kicked open. The almighty dollar rules all in this government of the company for the company.
I feel that Standford’s efforts will generate a vehicle that will become comparable to the average human driver within the next 5 years (if they haven’t already reach that goal) and within 10 they will be comparable with the very best drivers in the world. This tech will remove the need for any job that controls a machine and will irreversibly change the landscape of society for all time. The software that is being developed will allow machines to control themselves in the physical world. The car is just a convenient tool to test this new software, but it will be applicable to all new new machines from here on out.
Second to all of the benefits that AV will bring to our economy and society as a whole is the fact that we are releasing robots into the world. I cannot over emphasizes the tremendous impact this will have. One small rotation of a wheel, but the introduction of intelligent machines that will eventually rule the world.
Yes. This was exactly what I was going to say, well the first part about they are already. The dollar rules all. No where is AI taking off quicker then in the Financial district. Small start-ups are making money purely by making powerful AI’s and relying on micro-trading. However unlike humans that must be retrained, AI’s keep their training and loyalty to the company forever, and as computers become cheaper, the cost/benefit, will gradually become so lopsided, its simply a non-issue. The bigger question is do we feel sorry for all the truck drivers that have been replaced? Personally the more bad drivers that get off the road the better, getting rid of those with 20 ton rigs is just a start. Now what about every other dipshit given a drivers liscence? This change couldn’t come quickly enough for me. Plus robotic taxis could allow poor people to get their own ride. I’ve given people rides to and from work due to their financial situation and I think its really sad that they can’t afford they can’t have personal transportation around the city, without owning a car and all the hassles that come with that. In the South having a car is a necessity. And what’s more people who really need one can’t afford one, or atleast the newer ones that will be electric and fuel efficient. There is a huge market for someone willing to service that part of the population with this technology. All the economic essentives tell me we are just waiting for the technology to mature and I think the argument against it will be so weak, I doubt the debate lasting very long. Unless repeated disasters of Hindenburg-like proportionality occur, which can completely halt investment in technology for 60 years.
I think that the autonomous advancements can proceed gradually enough to evade a lot of resistance to “robotic vehicles”. This is about law and the law is all about definitions. Number one: Don’t call them robotic cars. Call them cars with “intelligent assist capabilities” and “enhanced safety features”. That’s truthful. As long as someone is sitting behind a control panel of some sort, and that person is licensed and insured, the legal fiction can be maintained that he or she is the operator and the car can be autonomous in fact. And when everybody is doing this safely, the ridiculousness of having to plant a human behind the control panel will become overwhelming and that requirement will go away.
I personally believe that vehicle ownership will begin fade away as well as robotic vehicles can pick you up and drop you off whenever and wherever, giving people the opportunity to forgo all the costs and responsibilities of vehicle ownership while keeping the convenience.
I agree with both of your points: this will be a very gradual process which will accelerate once it proven (and accepted by society) that Automated Vehicles are better at driving than 99% of human drivers out there. Im competent in my own driving skill, but most of you out there scare the hell out me.
Today taxi’s are expensive because their is a person that has to feed his family involved, once the car can drive itself taxi services will become so cheap that only the wealthy will own their own vehicle.
Yes there will be autonomous car.
Lidar prices could be $20, and a complete kit for 200 to $1000
The biggest problem is political : and is about norms, and rules
The autonomous car can talk to each others
Ibm has build systems for regulating cars this will be necessary
In fact i don’t think it will be easy to have autonous and non autonomous cars
BY THE WAY :
by the time we speak about autonomous cars, we will take more seriously space …
Not just on the ground but also in the air, using field propulsion.
Have you seen how primates drive? Bring on the freak’n robot cars.
Recently I had an idea that could be a good topic within the Timing / Roadmap of autonomous cars.
I’ve posted it on my blog, but I think this community could give me a smarter feedback.
I was thinking that Google could develop something more legally and
wide acceptable before introducing a complete automated car.
For example in order to use and gain a massive amount of driving data
G could use the hardware and software created to monitor and make
decision to drive the car in a different way.
In my opinion this way could be big external airbags for car-car collisions.
The great problem of external airbags was the impossibility to foresee
a crash and its entity, but Google automated driving system does it
quite well.
- External airbags would be a great way to reduce injuries and deaths.
- There are no legal issues. You keep driving the car.
- Nobody would say: “I don’t want it cause I love to drive”.
- Could be sold in many different versions: -side only protection to
complete 360 protection (number and position of airbags) – little
add-on for older car – car maker collaboration to sell built-in
system in new cars…
- And last but not least: Google would have millions and more cars
testing (internal simulation) G’s self-driving system and possibly
sending a huge amount of driving data from all over the world.
Then
- Software and Hardware could be easy updated, step by step,
to a complete autonomous car.
So, while legal and political issues are fading away, G could have just
gained a wide beta test for his cars and a huge market share
that just need a little “unlock”.
In my opinion it could be a great deal and a step forward within the
self-driving car roadmap.
What do you think?
Yes, but only if all components can be manufactured in large volume, such as flux capacitors:
http://www.youtube.com/watch?v=lULl63ERek0
Yes within the next 10-20 years ant then taxi driver will become unemployed.
If you could make it safe and economically realistic sure. That would entail a huge change in infrastructure, I believe. To implement a driver-less scenario we would need several safeguards and regulatory procedures. No matter how advance Artificial Intelligence may become, humans will never leave the responsibility of proper operation of a car in the hands of a machine without several precautionary measures.
There’s no way that driving automation would solely depend on the car itself to get the job done, especially considering the factors involved in everyday driving. Weather, traffic density, speed, who is and isn’t automated and a more immediate factor being accidents. All of this would need to be made available centrally among all automated vehicles and regulated by a central governing “server” (if you will).
This brings about a massive change in the already changing infrastructure. You would need smart roads and highways which would be sensor-heavy connected to an overall network which would also need to be dense to handle a specific amount of cars (as well as meet a future capacitance which is always integrated into gov’t construction). Would this be for all roads? Only major highways?
This could open a new job market depending on how large a scale this would be implemented on. That would be monitoring. You would require several teams and locations to monitor and regulate as the human fail-safe.
You would also need to implement a manual mode which I need not explain.
The big gap in adoption is government acceptance. Generally they require products and services to be on the market or available for 5-10 years prior to integration. They are NOT big on state-of-the-art when it comes to public safety. I would assume that this would first need to be introduced privately, such as an automated parking service or other small scale services not on public road. The problem with this is implementing this into the common consumer’s vehicle without raising costs astronomically but at the same time having the availability to use this option wherever that may be available. This would pave the way for government adoption.
Independent trials (such as Google’s automated driving) may show promise as well as benefits, but eventually it will be made illegal somehow as public safety is the major issue here (the automated cars are already getting into accidents).
The last issue would be population density vs. future roadway renovations. With cities and roadways becoming a nightmare (and in some cases already being a nightmare), regardless of lane additions, it may not be responsible to add automated driving, as it is more of a luxury than a necessity. The trade-off of automation may be less automobile related deaths and smoother traffic (hopefully eliminate it), but it’s still a luxury to drive, not a necessity.
All in all, I say it’s minimum 20 years out.
I think there’s no need of infrastructure changes. Almost every self driving car projects are very non-infrastructure-oriented.
When (probably just now) self driving cars will be safer than human drivers, it will become just matter of money to drive an automated car… because car insurance will be more expensive for human drivers. (Commercials will do the rest)
I think all in all it needs just some beta-times to test and debug the system, and a wider lobbing by Google.
Maybe the most likely forecast could be 2018, for a wide consumer automated cars adoption.
A self-driving car will have to be at least one or two orders of magnitude safer than current drivers before it’s allowed on the road. That shouldn’t be hard, but injury or death-causing accidents involving self-driving cars, even when a human driver in another car is clearly at fault, are bound to bring lawsuits against the manufacturer and every company involved (those that make the sensors, the computers, the software, etc.) People will blame these things for deaths even while they are demonstrably much safer than human drivers. Manufacturers know this and they’ll be looking for how to defend against and pay for the suits they know will come.
There will be little or no need of added infrastructure. Self-driving cars will communicate with one another continuously. When the first car in a row of 50 cars has to slow down, ALL the cars in that lane will immediately slow down as well. Every self-driven car will know which cars are NOT self-driving, and if I have anything to do with the programming, it will stay WELL AWAY from any and all human-driven cars.
I’d have to disagree with the no need of added infrastructure, strictly because I understand the transportation industry. Sure if cars can communicate to one another that would be a satisfactory scenario for a perfect world. What about outside interference such as an object that is not a motor vehicle with communications capability is obstructing the roadway. Sure the care should be equipped to sense this object and avoid it, but governing jurisdictions will not have the public depend on just a vehicle sensor as a means of safety. There will always be the necessity of redundancy. What is that first car fails to communicate with the other cars and a collision results. You would also depend on the cars with sensors to communicate with each other in the event that a car without sensors or automation is in proximity or about to crash. Again I urge the necessity of redundancy. Like I said, I work in transportation. Trains mainly. This is a method of transportation which is isolated from outside interaction and public tampering. Teams of people are on stand-by in the event that ANYTHING happens to those trains. Teams of people monitor and know the location of every train in their control. They have the ability to remotely stop every single train by the flip of a switch as well as locally at substations. As I said these trains are isolated from public interaction besides people riding on them, and they have sensors and controls globally on all tracks to tell them everything they need to know and give them control of everything in redundancy in the event of an emergency. And they no immediately when something goes wrong. That being said, automatic vehicles in comparison to a train in terms of scale of infrastructure and public safety are much larger in proportion. There are several guidelines that are to be followed to the tee when dealing with the implementation of new technology or concepts into transportation systems, the top priority being redundant safety measures. You mention something about programming which leads me to assume you may have personal insight as to how automated cars may drive. I’m only integrating personal experience with this concept, it’d be interesting to know how car companies would try and jump the safety hurdle.
Please read something about self driving cars here:
http://www.templetons.com/brad/robocars/
Avoid new infrastructure is a priority, to ensure fast and easy development of this tech.
Sure, the lack of redundancy could not lead to a 100% safe system but I think that just cutting by 50% (I think 80%-90% could be more likely) the 1.000.000 deaths per year would be a great results.
Maybe in a narrow-capitalistic-insurance way of thinking, just a 10% safer could worth the risk, but probably Google will wait until the system will reach far more than 50% more safety, until sell a consumer product.
PS: Just joking (but not at all): Probably to avoid any problem on “who’s the blame for this crash?”, Google will force to sign a release that delegate responsibility to the owner.
Having taken Sebastian Thrun and Peter Norvig’s online class on AI, I think you’re woefully mistaken about the kinds of infrastructure necessary to have successfully automated cars. The cars don’t need special infrastructure. I think Stanley’s win of the DARPA Grand Challenge solidifies this as there was no infrastructure whatsoever. The cars were autonomous and successfully navigated the course. Google has also demonstrated autonomous cars navigating rush hour traffic and driving down Lombard street in San Francisco with no additional infrastructure. While it’d be great if cars can communicate with each other, just driving around each other is communication enough.
Google Maps is nothing more than Google building the maps necessary for their AIs particle filters to be successful. It’s great for humans right now, but I guarantee the long term goal is for AI navigation.
These vehicles are very safe in fact. My guess is that the main problem is going to come from human caused collisions. Of course, there’s potential for software bugs to cause issue, but that’s already happening (See Toyota Prius Acceleration issues).
Automated vehicles are on their way. In the short term I’m guessing we’ll need human fallback systems, but we already use this model with air transportation.
I think it’s funny people mention the jobs lost to automation such as cab drivers. What is the percentage of the population that are cab drivers? 2%? While certainly not insignificant, it’s not like it’s going to have a massive impact on society to have less cab drivers.
How about truckers? That’s a much bigger bucket, but truck drivers also afford trucks with much needed security. Maybe that’s easy to overcome, but I think humans will want to travel with freight for security alone.
As for congestion, automated vehicles should reduce, if not eliminate most heavy congestion because AI response times are much faster than humans. The compression waves caused by human reaction times would probably go away. Once the AIs see cars moving, they begin to move immediately. Sense slowdowns, slow down and space out. This of course only works if all the AIs “play nicely” with each other (non-adversarial).
I think there is about a 50/50 chance because of the money to be made by new industries and lost by others, on top of regulatory issues. Aside from the idea that people could relax in the car and watch a movie or do work, This could create entire industries. For example all restaurants, supermarkets and stores like Target could deliver food and other things to homes for cheap. On the other side car makers might suffer because many people would be able to summon a rental vehicle from any internet connected device. It would be like Zipcar but the car would come to you. hence cars would be shared by many people.
Perhaps the first legal driverless cars will be passengerless delivery drones; programmed never to run into anything and designed to be virtually harmless if they did.
1. the technology for this exists
2. it\’s already safer than human drivers
3. we can safely expect the cost for the hardware to come down at a rate close to Moore\’s law
4. Nevada is already implementing this, giving other states a legal framework
5. we\’ve already seen a robot driven car that can drive up Pike\’s Peak far faster than 90%+ of the people reading this article
6. we have an aging population that needs cars to maintain their independence
There\’s NO WAY we WON\’T see this in the next 10 years.
Agreed. My guess is that fully-automomous cars will be available in 2017 and mainstream in 2019.
Safer, more efficient, however self driving cars will lead to technological obsolescence of driving-based jobs. The number of people involved in transportation globally as their main source of employment is staggering. According to the Research and Innovative Technology Administration (RITA) Bureau of Transportation statistics, the total labour force in transportation was over 13 million in 2009. Of this around one and a half million are truck drivers. Jobs involved with driver training, and licencing will eventually be greatly reduced or eliminated altogether. Police department revenues from traffic violations may vanish into thin air leaving municipal and state coffers dry.
In a wider scale, once self-driving cars are tied into other systems, there will be effects in areas not necessarily related to driving alone. With cloud systems like Google’s ROS, the object recognition database world-wide will be populated exponentially. Google maps will be updated in real-time, with much greater levels of detail. Once other robotic systems (like android assistants, flying robots and smaller long-distance autonomous vehicles) come on-line, they too will be able to navigate with safety and precision. Logistics and supply chains may be filled with even greater speed and accuracy than they are today, even down to a personal level.
Transportation design itself may be put in the hands of artificial intelligence agents to allow for maximum throughput and efficiency for traffic corridors. With the amount of people that will have their jobs displaced by the technology, it may be necessary to develop with rampant homelessness a consideration.
Questions remain such as how drivers’ privacy will be affected, whether current vehicles can be retrofitted and how many vehicles would be needing the systems to develop an effective network. Many drivers are already uncomfortable with “event data recorders” like OnStar in most new cars that record everything a car was doing immediately before an impact. What will the reaction be to cars that continuously broadcast everything a car does?
There are also issues of trust. Some drivers will always put ultimate faith in their own driving skillls no matter what. Moreover, many will simply miss the pleasure of driving a car. Will there be special tracks or areas set aside for human-driven vehicles?
With concerns rising over “distracted driving,” however, it’s clear that most people would rather be doing something else while they’re behind wheel. According to GM’s Burns, “We’ve concluded that, for a lot of people, driving has become the distraction.”
Like any other revolution in technology, the predictions do not often match the results, but in the case of self driving cars we won’t have long to wait
Excellent points, to delve deeper into potential job lose due to automated vehicles consider the travel industry. I can see the hotels losing large percentage of revenues. Why pay for a hotel when sleeping in your vehicle saves the cost and gets you to where you need to be that much quicker. Putting the family to bed in the van and waking up in the parking lot at Disneyland saves the need to pack that overnight bag.
You mentioned municipalities losing traffic ticket revenue, that would be easily paid for since the cops that hand out the tickets wouldn’t really be necessary either.
Insurance companies will certainly feel this tech, everyone is worried about who is responsible when an accident occurs and therefore do car companies or drivers needs to carry insurance, I’m thinking more along lines of who needs insurance when this tech has the potential to make accidents virtually a thing of the past. I won’t miss all of those commercials but broadcasters certainly will.
Don’t mistake me for a Luddite however, the economical benefits will far outstrip the job loss. An immediate impact will be the energy savings, AV will save huge amounts of gas through efficiency. Getting anything from point A to B will be much cheaper since trucking company won’t have to pay incomes, reducing the cost of living for everyone.
More later
Well said
It’s important to keep in mind that an autonomous system isn’t necessarily and “either/or” proposition.
For instance, I can see a sort of super cruise control engagable on command by the driver for, say, lengthy highway driving. Truckers might jump on something like that to reduce fatigue as might the drivers of passanger cars who regularly make lengthy drives. As the systems improve/prove themselves the range of their conditions of use increases.
Another possible use for autonomously driven cars would be as taxis.
While “around town” driving is pretty demanding the DARPA Urban Challange proved it’s doable and that was a couple of years ago.
Where there’s some really interesting possibilities for synergy is in the combining of autonomous driving capability with an electric car.
Just getting rid of the driver dramatically alters the economics of the taxi cab. But combining those advantages with the *economic* advantages of electric cars starts to undercut the value of privately-owned cars for people who don’t often drive long distances.
The convenience of the private automobile with the economic advantage of only “owning” it for those time when youu actually have a use for a private car, would be a very attractive package of advantages. Add to that not having to actually control the car, simply sitting back and relaxing, and I believe you have the makings of a society-altering development.
The advantages of autonomous cars go much farther then what is being discussed I think once you start looking at what could happen when they are adopted at the mass level:
1) Car sharing could become much easier with automated pick-ups requested from your home tablet to ensure you are picked up by an automated taxi. This will increase expendible income for people who choose this route instead of buying/leasing cars.
2) The cars themselves can start talking to each other, and traffic lights, conditions, creating a network of information sharing to expedite traffic flows. This will both allow people to arrive to work on time with less hassle, but actually relax, talk, read the news, etc on the way to work.
3) Imagine the add-ons for automated cars? Sensors and cameras that can report and track whatever is deemed necessary for the common good (track license plates, do facial recognition, report accidents automatically, and a myriad of police duties). Even drive you to the ER if needed, communicating with all the vehicles on the way to get you there ASAP (imagine ambulances that can have 2-3 EMTs working on you instead of having to drive?). Yes you get into some scary “big brother is watching” scenarios, but there will alwas be checks and balances.
4) Automated or not, would love to see all cars be moving WifI hotspots. Metro WiFi made possible. Since electric cars have more room typically you can start putting interesting things such as this.
5) Back to car sharing, you can dial up the type of car that you need for certain situations too. So if you need a utility vehicle like an SUV or pick-up for hauling home your new couch, or a van for a family trip, you’re all set!
I think the evolution in automated cars will come with the possibilities of mass adoption over time, and once people get over the NEED of having to own a car that is theirs, their precious… I’ve always treated cars as tools to get from point a to b, yes sometimes they are fun to drive, but I can have MORE fun in an authomated convertible than one I have to drive myself, unless I want to be a racecar driver
I agree that there is so much more to discuss, beyond all of the economic befits that you have mentioned which I fully agree with, nice post by the way, the biggest impact this tech will have is automated thinking machines will be interacting in the world on large scale. These machines will be making decisions and navigating with goals to accomplish. I can think of a great number of biological species that do this. As the 2nd and 3rd, 4th, 5th generations of these deceives evolve the we will quickly realize we are sharing our world with early version of a non-biological species.
Chew on that, I’d love to hear your thoughts.
Which future headline will we embrace?:
Autonomous vehicles reduce traffic fatalities by 80% in the U.S.
or
Robot cars killed 4,500 last year in the U.S.
In the US last year, traffic fatalities were down to roughly 30,000 thanks to safer cars. Industry experts think that autonomous cars could reduce that number by 80% or just 4,500 fatalities. Which number will the public focus on?
On the right side of this page right now is the Martin Ford book asking the question… “Will Automation Lead to Economic Collapse?” I can’t wait for automated cars but I deliver food out of my car, the first thing I think of is these vehicles taking my job. Now let’s skip me. Taxi’s would be first to go to automated, but what about all those taxi driver jobs(sad). The argument has always been, that with new technologies come new jobs. However I definitely believe our technology can/will get to the point that automated things can do almost every job. Service jobs should be one of the last to go, I can easily see robot waiters coming. Managemnt might be last to go, but really…. how many decision makers can be employed. Not a whole world full of them. I know there’s nothing we can do to change the path we’re on, I just hope when those Taxi Drivers lose their jobs to automated cars, google will continue to pay their salaries.
Actually, the argument’s always been that with new technologies, new wealth. The jobs that are typically displaced are low-skilled and repetitive jobs which means that cheap machines, as opposed to expensive human beings, are doing the jobs. By looking fixedly at one side of the equation you can completely ignore the fact that, as the end result of centuries of those sorts of displacement, the condition of humankind is massively better then if those displacements hadn’t occurred.
You think when Google makes automated taxis it should pay the salaries of former taxi drivers? Did Henry Ford pay the salaries of former buggy whip makers?
Don’t make any assumptions about what would be the first to go not that I’m suggesting you just sit there and hope your head isn’t on the chopping block. It’s just that subtle dynamics make some obvious advancements non-starters.
For instance, if you look back at the history of motorized flight vertical take-off and landing were simply assumed. But when real aircraft started showing up, and it was realized through hard experience how tough vertical flight really was, what we now refer to as “conventional” aircraft became the dominant form.
That doesn’t relieve you of the personal responsibility of seeing to the value of the skills you have to offer, just that assuming the worst is simply another conceit.
In fact, a case can be made for autonomous vehicles coming to dominate in the agricultural sector before they show up on the roads. First, that’s the direction agriculture’s headed anyway and has been for a long time. Second, it’s a much more controlled environment with fewer in the way of legal and regulatory hurdles, a calculabe return on investment and more then a little in the way of engineering resources to push the technology to the next step.
Or automated, aerial package delivery might burst on the scene completely altering the economics of the industry. First, out in the countryside were population density is very low and package delivery costs are commensurately high but as the technology improves, moving into more densely-populated areas.
Or something completely surprising will occur and have us old fogeys scratching our heads mumbling “I didn’t see that coming”.
So start looking to the skills you have to offer that other people are willing to pay for. It really is the only defense you have against obsolensence. The sooner you come to terms with that hard reality the better off you, and those who depend upon you, will be.
Automation is here so to think that autonomous vehicles (AV’s) aren’t on the horizon, the very near horizon, is setting yourself up for failure. The benefits far exceed the negatives:
o Elimination of driver restriction – Nearly anyone can use an AV
o Elimination of driver – An AV can drive itself without a circadian cycle
o Elimination of parking scarcity – more space with less sprawl
o Fewer crashes – and less lives lost
o More roadway
The list goes on for the positives of what this technology is bringing to the table. I think Arpad summed it up rather well in the final sentences of his post below.
“By looking fixedly at one side of the equation you can completely ignore the fact that, as the end result of centuries of those sorts of displacement, the condition of humankind is massively better then if those displacements hadn’t occurred.” – Arpad
However, don’t think it will take centuries to see the net benefit. It is hard to believe someone in a history book will look back and sum it all up as Michael Schermer has done with many other aspects of life. Instead we will see the benefits immediately. With our ability to acquire and assimilate data through various methods (cloud compilations are only the tip here) we will be constantly reminded of how many fewer accidents there are, how many more people share mobility, and we will laugh at the tedium of actual driving.
We’re looking at emerging technologies right now that overcome the hurdles set in place to make autonomous driving a reality.
o A negligible (if any) increase in infrastructure and likely a decrease as we can do away with a LOT of the road signs.
o Working at or above human standards – over all safer (this one bullet sums up many)
o Cheaper costs – as implementation increases the cost will decrease
We’ve already accepted with little debate autonomous assistance in driving with assisted parking, back-up assistance, blind spot coverage… we’re proving everyday that computers are enhancing our lives on the road. Isn’t this the next logical step?
From a personal point of view as a business owner–and of course I’m bias–I would rather see an autonomous drone vehicle schlepping cargo around on the road than a human, not only for that fact that they don’t need to sleep and only rarely stop, but also to take that person out of that unhealthy and dangerous job. At first that seems harsh but we’ve seen the impact long time driving has on truckers.
I’m going on entirely too long here. Most of us here could write a book on the advantages vs. the disadvantages. Instead perhaps the next process of thought should be what is the next hurdle of implementation, legal, moral, media, or something else?
No, it certainly won’t take centuries but I believe the hurdles are going to be legal/regulatory and not technical. The DARPA Urban Challenge had autonomous cars driving in relatively simple, simulated urban environments and that was in 2007. Of the eleven teams involved six completed the course. That was five years ago.
I wouldn’t be at all surprised to see autonomous vehicles debuting this year in a commercial, but tightly-controlled environment, like an open pit mine, farm or in a large factory. On the open road and for sale to private parties? Don’t know and won’t take a guess.
Since you quoted me let me correct your misperception.
The “centuries” of precedent to which I was referring was in response to someone up the thread who was worried about the taxi drivers autonomous cabs would replace. I was pointing out that for every person displaced by technological progress many more people have benefited. Looking at only those who are injured by technological progress means, largely, precluding all technological progress.
To get back to the subject at hand, I think you’re being entirely too optimistic about the time line for the acceptance of autonomous vehicle. The litigous, and risk-averse, nature of our culture means there will be non-technical impediments to the adoption of autonomous vehicles.
Proponents of the “precautionary principle” will have a field day and the first time anyone’s injured by an autonomous cab, or can convince a lawyer who stands to benefit handsomely from any out-of-court settlement let alone a favorable decision that they’ve got even a borderline valid case for a law suit, the prospects for the technology are damaged.
There are other considerations as well.
For instance, will there be an emergency overide of vehicle safety programming? Under what circumstance will a vehicle refuse to convey a passanger to their destination? Will icy roads mean all autonomous vehicle traffic comes to a standstill? How comprehensive will autonomous vehicle “awareness” of unusual conditions, like an ambulance flying up from behind or traffic barriers, have to be before they’re safe on public roads? If you live in the California hills will an autonomous cab taking you home blithely carry you into a fire? And if it does, who’s on the hook legally?
The non-technical problems aren’t minor and, as the recent enactment of law in California and Nevada on the subject of autonomous vehicles on public roads shows, only just being engaged and thus far from solved.
We’re already starting to see driver-less features slowly worked into new cars. Lane monitoring sensors, cameras on the back and sides, and even breaks that automatically kick in and stop you if the car detects you’re going too fast and about to hit something. Oh and don’t forget automatic parallel parking, and some companies are working on “advanced cruise control” for highways, that will keep the car in the lane and at the right speed, you just have to monitor it and take over when you need to change lanes, slow down, or get off the highway.
Personally, I think we’ll see the first commercially available self-driving cars as a high-end option, for those who can afford it, before the end of THIS decade. Of course it will take a little longer to reach the general market, but it will do that rapidly enough.. Don’t forget, this isn’t just a fancy feature for cars, this is a SAFETY feature! It’ll be expensive at first, but because of the enormous benefit, far greater than any other safety feature, it WILL become common place rapidly. People will demand it, politicians will use safety to advance their careers and make themselves look good.. There’s a lot that will push this forward.
I think by the mid 2020s we’ll realize it’s common place, and here to stay.
I think you’re exactly right. We are already seeing some of the semi-autonomous features you mention in high-end cars. They’re known in the industry as ADAS: advanced driver assistance systems. And they’re trickling down from the Cadillacs et al.
USDOT studies have already shown they reduce accidents and save lives, and a cultural shift will make ADAS more desirable to drivers. For baby boomers and gen x, driving and owning a car was a huge privilege. (Imagine, children, once upon a time, if you wanted to buy something, you had to go out and look for it!)Younger folks would much rather play a game or watch a video while they get somewhere. They’ll think it’s great when the car takes over.
And don’t forget v2v and v2x communications. USDOT is thinking about mandating that all new cars carry a dedicated short-range communications modem to broadcast heading and speed over spectrum already set aside by the FCC. Cars so equipped will be able to automatically slow down when they approach smart traffic signals or warn drivers — or take over — if another car is too close. Just for example. If it’s mandated, we’ll see v2x-equipped cars on the road in 2016.
Um… the next decade? I drive too and from work in Silicon valley, and once in a while I\’ll see one of the google cars in traffic. While there is someone behind the wheel, they don\’t have their hands on it and they look way more relaxed than anyone else in traffic. I think that it will take one year for Nevada to legalize self driving taxis.
From there … were off to the races. Other states will have a model. Aging boomer populations won\’t want to let their freedom go and they\’ll have a new option (beyond being dangerous drivers). Busy mothers will be able to put their kids in a self driving taxi for as cheap as it would be to drive them. 12-16 year old will gain freedom and never *want* to drive since the freedom of the car and the responsibilities will be separated.
Next steps… HOV lanes become \”robot\” car lanes packed with 3-5 times as many cars, maintaining constant speed and not rubbernecking for accidents/traffic infractions/etc.. going from point A to point B outside of a robot driven car won\’t make sense (with traffic it\’ll be 10-70% slower).
Robot taxis will take over for car ownership, trucks that drive themselves will take over driving jobs we\’ll loose 10% of our existing jobs.
I agree that these robotic beasts will be here sooner than most folks expect.
As I said before, completely autonomous vehicles will become available around 2017 and mainstream by 2019.
This is actually a huge issue. It could make an enormous impact before the next presidential election.
What I think people don’t “get” is that this could drive car transport from an ownership model to a service model. Right now I pay $2 a day to own a car (insurance) even if I don’t drive it at all. I use the car 2-4 hours a day (my commute sucks) and it sits idle for 20-22 hours a day, depreciating and collecting dust (which costs me ).
If a robot drove, the car could be driven 16-20 hours a day 800-1000+ miles a day and 300,000+ miles a year. If you run a robotic car service with tens of thousands of cars, fuel and maintenance and engine/motor life become MUCH more important than initial vehicle cost. The difference between 30mpg and 50mpg at $5/gallon is $20k a year, over a 5 year life it’s $100k.
At that price a carbon fiber car body, hybrid drivetrain, and a natural gas engine all make sense.
Upsides: As a side effect of this the US will end it’s addiction to foreign oil. We will also save most of the 40k deaths per year in auto deaths. Savings of hundreds of millions of hours a year stuck in traffic.
Downsides: loss of transport and auto manufacturing jobs.
You’re right, and you’re just scratching the surface of the changes that will be possible. Currently, the design of cities is based on an old transportation model. Parking meters, lots, garages, ramps…all of them exist to allow people to store their cars reasonably close to their final destinations. And all of them occupy valuable real estate.
Now imagine this: Your car drives you to work, drops you off at the front door, then heads off to the grocery store. You’ve already emailed the store with your wish list, and they have the goodies ready for a (robotic?) employee to load into your trunk.
Then you call your car to pick you up, and voila! You’re greeted by your vehicle, loaded with your merchandise and ready to take you home. Not only does this system save everyone time and energy; it also frees up vast tracts of urban space. The layout of every city is about to change.
First, by “next” presidential election, I assume you’re talking about 2016, not 2012. Either case, I think you’re being a bit optimistic. Most people keep their car for 4+ years (probably longer now than a few years back), and the average car is driven for over 10 years before it is retired. So it will take some time for the fleet of cars to turn over, even if they all magically were 100% robotically driven starting next year.
Second, you’re assuming people would quickly move to a rental system. There’s compelling reasons right now for people in urban areas to rent a car, and it’s a gaining traction, so to speak, but still a very small part of the car ownership picture. I’m not sure adding robotics would immediately change that aspect of it. People in very rural areas and those with special transportation needs (building contractors, farmers, etc.) would still have their own vehicles as well for the imaginable future.
But in the long term, I agree with you, those are some interesting points, particularly that parking lots would probably be centralized and not extensive at every single store. This also gives mobility to the elderly, drunks, the disabled and the young.
If we could merge robotic cars with mass transit, that would be another big step. Take a train for trips over 100 miles or in congested areas, use electric power or natural gas for short trips to your final destination.
Yes 2016, not this years election.
Let me clarify. I don’t think we’ll see all the change that I describe by that time. But I do think that professional driving jobs will be one of the first targets for this technology. Commercial trucks that drive themselves would pay for a robot system in a year or two, if that starts in 2015, it will have an enormous economic impact for the 2016 election.
If you get rid of the people driving the trucks you also get rid of DOT hours of service regulations. These automated trucks will have very little down time and wont need to stop to sleep. Once automated trucks are available the trucking industry will convert as quickly as they can because it will more efficient and and probably safer (lowering insurance costs). They could also cut back on office staff since they wont be dealing with human drivers any more.
Has any one considered the implications for pedestrians? How will they be affected by autonomous cars? Will roadways be safer to cross because human error and inattention has been eliminated? Will they be more dangerous because vehicles are traveling so much faster and without any requirement to stop at intersections? Will there still be dedicated crossings, where cars are signaled to stop to permit crossing? Will those be the only places safe to cross? Will intersections change all that much if cars still need to stop frequently?
I look forward to cars driving themselves, thought the train will likely be my main mode of transport. Google has shown the tech does work, many of my friends LOVE speeding and doing burnouts etc.. (on public streets) however I really don’t care for it, if I want a cheap thrill I will go to a theme park roller-coaster ride. Nobody at real risk then!