By on September 3, 2015

Google Autonomous Vehicle Prototype Circa December 2014

I’ll admit it. I, the millennial managing editor of an automotive blog, would absolutely rely on an autonomous car for my day-to-day errands or long-distance commutes. Why wouldn’t I? I can kick back, relax, talk with people, get some writing done, or anything else I could possibly do on an airplane. As long as all the other vehicles on a roadway are autonomous, it’ll be safer, too!

Why do I think this way? The majority of the driving we do is boring. I can just imagine hailing an autonomous car on my phone, waiting for it to arrive to my home, and setting it to drive me wherever I want in relative comfort. Why should I need to stay alert at a four-way stop if technology can make that a thing of the past?

Except it probably won’t work quite like that.

Instead, we are seeing autonomous tech come out in trickles.

While the Google autonomous car project is the one garnering the most press at the moment and Tesla is also in the running for headlines, Cadillac, Audi and other manufacturers are putting semi-autonomous technology in their cars right now or in the not too distant future.

Cadillac Super Cruise will be here next year on the CTS and CT6. This is probably the best application of our semi-autonomous near future. Long-distance driving is great when you get to enjoy your surroundings. It isn’t so great when you’re staring down the rear bumper of an RV with F-150 taillights.

But, in order for the autonomous future to happen — the one where we have roadways dedicated to fully autonomous cars and us humans aren’t allowed to manually drive — the autonomous cars need to outnumber the ones that don’t.

This happened back in the early 20th century. “Automobilists” as they were called back then were on the constant defensive against livery stable owners and other businessmen who had a vested interest in keeping real horsepower on city streets. Take this exchange between two groups at a meeting in New York City discussing automobiles in Central Park.

Screen Shot 2015-09-02 at 5.08.35 PM

Lawson N. Fuller said that he had been fifty-six years in New-York, and had driven more miles in Manhattan Island than any other man in the city. “And I must say from experience,” he added. “that the vehicle which frightens my horses most is the baby wagon. Next to that is a wet spot on the road.

“Maybe your horses are nearsighted,” interrupted Alderman Bailey, with a sneer. “There are nearsighted horses, you know.”

“Yes there are nearsighted horses and near sighted mules and nearsighted asses,” replied Mr Fuller, at which the Alderman subsided in a burst of laughter from the crowd.

New-York Tribune., November 10, 1899, Page 6, Image 6

It wouldn’t be until a large proportion of the population adopted the automobile that horses became a thing of the past. And just like the stable-owner-friendly alderman in 1899 New York City, it’s the shared nearsightedness today that sees autonomous vehicles as just a blur in the distance; an unformed, not quite solidified idea of the future.

Personally, I think the autonomous future will happen a lot sooner than most of us would like to admit, and it will happen one car at a time just like the Toyota Prius. The idea of a hybrid in 1997 was a fantastical idea. Now they’re ubiquitous and have been for quite some time.

So, my prediction … in 2030 there will be more autonomous cars on the road than those of today. What say you, B&B?

Get the latest TTAC e-Newsletter!

92 Comments on “QOTD: When Will Autonomous Cars Outnumber The Manually Driven?...”


  • avatar
    Scuttle

    What about motorcycles? I cant see motorcycles being made autonomous so do they ban them from the street and make them track only in the future? I will say that autonomous cars would make the streets safer for those of us who ride.

  • avatar
    HerrKaLeun

    0 cars now, and you base your “2030” prediction on what?

    Hey, let me write a click bait article with no content: When will drones transport people and replace cars? Drones are already more advanced than cars, so my guess will be 2037. but that only is so late because the FAA will take till 2035 (around June 3rd, approximately 2:03 p.m.) to approve them.

    At least my numbers are on a more solid foundation than yours.

    • 0 avatar

      Zero cars? Really? There are semi-autonomous cars on the road RIGHT NOW in private hands.

      • 0 avatar
        HerrKaLeun

        DOT approved to drive truly autonomously while the car is empty or driver sleeping?
        Where exactly can I go today and buy this car and have it drive legally autonomously without supervision in all OECD (just to limit it to developed countries) countries?

        Semi-autonomous is BS and not autonomous. If you need a fleet and an actual human supervising a car, it misses the point.

      • 0 avatar

        It’s a long way from semi-autonomous to autonomous.

        I get joy out of most of the driving I do here in the Boston area, as long as the traffic is flowing reasonably well, and I’m not only on highways. A day when I don’t drive somewhere is a day I feel deprived.

        I’m somewhat torn about autonomous cars–they certainly would be wonderful for people who can’t drive for whatever reasons, and if they reduced traffic in cities because you could call for them when you needed them and you wouldn’t have to search for parking places, that would certainly make life easier. But I want to keep driving as long as I live.

      • 0 avatar
        White Shadow

        First of all, semi-autonomous is not the same as autonomous. Secondly, semi-autonomous can be very loosely defined. A car with basic cruise control can fit the definition. Jump down off your high horse….

    • 0 avatar
      Vulpine

      It is not ‘zero cars’, HerrKaLeun; Google has had autonomous cars on the roads for a few years now and a few states have made them legal for use–under human supervision. They’ve already proven their capabilities and the (very) few accidents in which they’ve been involved they have both been at very slow speeds and with one exception caused in one way or another by the human driving the other car (with the only exception being the human driving the Google car manually.)

      Aside from that, nearly everyone driving one of the newer luxury cars and even some of the more mid-level cars has some level of automated driving in the car, either in the form of parking assist, lane-change warnings or even active cruise control.

      No, while I don’t disagree with your slightly longer timeframe, I do disagree with your reasoning. The human factor alone will be why it takes that extra five years or so as they, like you, will continue to resist change. Change is inevitable. The motor vehicle took almost 40 years to go from being a tinkerer’s toy to a common sight on city streets and country roads. It took another 25 years to become the predominant mode of travel for individuals. The train took thirty years to go from horse-power to steam power, if not longer when you consider mining carts rode rails of one sort or another for centuries in some areas. Air travel took almost thirty years to go from little more than a powered kite to a means of public transportation if you ignore balloons and airships, which themselves took nearly 200 years to become more than a means of ground observation with the introduction of the internal combustion engine (steam was too heavy to haul up). To expect less than thirty years ignores how people resisted change and the technologies themselves had to improve enough to become reliable. BUT… your argument is more an example of that resistance rather than a refutation of the thesis.

  • avatar
    thelaine

    When hope dies.

  • avatar
    bubbajet

    2030 near side, 2040 is more likely IMHO. That’s just to cross the 50% +1 number though. I, too, would welcome it. There was a great editorial post from The Oatmeal(r) pointing out the benefits, in particular, for older people. He used his mother as an example but think of it in terms of your family. I’ll soon have parents that shouldn’t drive, but will still have the faculties to want to socialize. What better way than an automated car, allowing all of us to maintain the dignity of independence a little longer. This is on top of the safety benefits for all. Of course,there are security concerns, and driving can be fun for sure.

    • 0 avatar
      Speed3

      A lot can happen in 15 years. That’s 3 model redesigns into the future if you are Toyota/Honda. 2 if you are GM (and maybe one and a half if you are FCA, lol).

      Think back to 2000 though. The iPod didn’t even exist, now think about all the functions of the iPhone 6. Maybe it will be a long time before all vehicles are fully autonomous, but technological change can happen faster than imagined. Its also hard to predict 15 years out.

  • avatar
    RideHeight

    When bleary, exhausted illegals get them provided for free because by the time widespread AVs are feasible that’s who’s going to be the biggest demo on the roads.

    The rest of us will be dead or jacked-in and not driving much.

  • avatar
    ydnas7

    unlike horses and cars, autonomous and human and combinations thereof are quite able to coexist on the road. After all, there is not much reason to think a windows car will consider a human powered car to be any more dangerous than a mac car, particularly if the human car has a inter-vehicle communications occurring.

    • 0 avatar
      mmreeses

      Fully automated airline flights will happen long before fully automated cars as it’s much easier and cheaper to retool aviation infrastructure when compared to retooling the roads…

      And you still aren’t hearing a plea from the public for robo-flights, despite all the lives lost over the years due to some sort of human error.

      • 0 avatar
        Vulpine

        “And you still aren’t hearing a plea from the public for robo-flights, despite all the lives lost over the years due to some sort of human error.”

        Nearly every commercial flight today is a ‘robo-flight’ to a good extent. From the time the plane leaves the ground to the time it touches down again the auto-piliot has at least partial, if not total control. The accidents come when a situation arises for which the auto-pilot cannot correct, at which point it hands over control to the human pilots. When you hear of a mid-air collision, it’s not a commercial plane with another commercial plane, it’s a commercial plane with a privately-owned plane that doesn’t have an autopilot on board. Or it’s two private planes with neither carrying an autopilot. The few ‘near misses’ you occasionally read about between commercial planes are almost always in the vicinity of an airport where in some cases they can come within a couple thousand feet of each other just under normal operating conditions and there will be fifty to one-hundred large planes flying in a relatively small piece of sky. The only times you hear of a collision between commercial planes is when at least one of those planes is still on the ground, under manual control.

        • 0 avatar
          carguy67

          I think your understanding of autopilots is lacking; autopilots are perfectly happy to fly you into other aircraft. To date, there are no autopilots linked to TCAS systems, or capable of performing evasive maneuvers; see (among others):

          http://www.airliners.net/aviation-forums/tech_ops/read.main/286700

          Autopilots will fly a heading, an altitude and (in some cases), a flight plan (including departures and approaches). That’s it. Aircraft with autothrottles will maintain a specified airspeed, but they are relatively few in number (but increasing).

          Most midair collisions happen at or near uncontrolled airports; where the ‘see and avoid’ concept is most critical. Now, if you’re equating an autopilot with a ground controller with radar coverage of traffic, you have something (and all the controller can do is call out traffic and/or vector one or both of the aircraft at risk).

          • 0 avatar
            Vulpine

            Try again, carguy. They include the capability to maneuver to avoid another similarly-equipped aircraft… usually by having one climb and the other descend to avoid contact. Yes, the planes do already have the ability to communicate between themselves as a safety factor.

  • avatar
    ydnas7

    autonomous drones
    brings a whole new meaning to

    drop-off

  • avatar

    I have no problem with autonomous driving ont the highway.
    My cars have adaptive cruise control that sets distance and speed above 30 mph, but requires you to drive under 30.

    Thing is, my cars don’t have electric steering which is a necessity for “self driving” and “self parking”. GPS, lane keeping cameras…

    Computerized, electronic steering/throttle makes you susceptible to hackers.

    Insurance companies will NEVER go for this. Nor will states generating revenue from tickets. Can’t fine a robot.

  • avatar
    mmreeses

    I can kick back, relax, talk with people, get some writing done, or anything else I could possibly do on an airplane.

    this is why precisely, it’s not unreasonable to think that a world of AVs will worsen traffic in the future.

    Why take the train if you currently take mass transit and have the $$$? You AV to work. You send your AV home as a deadhead for your kids/spouse to use. Your AV comes back to pick you up. So now 4 one-way car trips, when there used to be zero.

    You don’t use an “Uber AV” cuz the same reason why you don’t use a cab today. you want to use your car, not one that’s been shared by thousands of people.

    • 0 avatar
      dal20402

      This is why congestion pricing (in central cities) and tolls (at highway bottlenecks) are coming whether we like it or not. You’ll take the train because it’s too expensive to pay for those four AV trips, just like you take the train today because it’s too expensive to park.

  • avatar
    PeriSoft

    The real problem with *fully* autonomous cars is that right now there’s no way to handle a lot of tricky things, and nobody seems to really be thinking about those problems.

    For example, a few weeks ago I drove up to Watkins Glen for a race. I drove out into the paddock and parked around near where my Chump team was setting up. There isn’t anyone even *working* on an autonomous car that would even have had a method for me to pick where to park! It wasn’t roads; it was just a flat expanse of tarmac with various little connectors, cones, and random objects that are different all the time. How would you even specify where to go to your car? “Oh, go over by the 173 team with the Miata.” Nuh-uh. Not yet.

    Other examples: Before fully autonomous vehicles are practical, the way road work is done will have to be completely changed and standardized. I drove my kid to his day camp this week, and there was a crew on the road trimming trees. They had a big cherry picker thing that swung into the opposite lane occasionally, so they had one dude standing there directing me to slowly drive off the edge of the road onto the shoulder to get around. He was working with hand signals and eye contact.

    Is anyone even *trying* to make an autonomous vehicle that can handle that situation? Are they even proposing anything which might be able to handle that?

    And if you take a road trip of any length, you’ll encounter those situations regularly, things where you rely on eye contact, ad-hoc barriers, context cues.

    Meanwhile we have people seriously saying stuff like, “Oh, we’ll laser scan everything and keep it up to date enough so the maps will be updated on the fly.” Seriously? You’re telling me you’re going to laser scan every ass-end Upstate NY road when a pothole gets worked on before the next car even shows up?

    Be serious.

    This doesn’t even get into the issue of what happens when you try to run an autonomous car in an area that doesn’t have perfect southern-California weather. “Oh, the cars can read the lines on the road! They have detailed maps!” Yeah, that’s great. What happens when the road is covered in snow for six months? When two lane roads become defacto one-lane roads due to build-up from plows? When snow covers the road signs? When there’s six inches of packed snow that totally changes the way the road surface appears and makes curb edges undetectable?

    What about terrible weather? Low visibility? Are you telling me that people will only control their cars manually in the worst of conditions? That sounds like a GREAT idea! Or that my car will refuse to go to the hospital with my pregnant wife because it’s snowing too hard? That sounds like a great idea too!

    Autonomous cars can be practical fairly soon in limited, controlled circumstances. Interstates, major arteries. But suggesting that fully autonomous, point-to-point vehicles are going to happen in 15 years is, in my opinion, utter folly, not because technology *can’t accomplish what’s necessary* but because infrastructure surrounding the way we drive needs to be changed from the bottom up to enable it to occur. It’s easy to change technology. It’s not so easy to do a 180 with 100 years of accumulated expectations and traditions of traffic management, and that’s what’s required to create a situation where fully autonomous cars can even *begin* to function.

    And that’s after the Silicon Valley tech types wake up out of their utopia and realize that not everyone has the same lives they do.

    Will it happen? Yeah. Will it happen the way the pundits and tech wunderkind would have us believe right now? Hell to the no.

    • 0 avatar
      Vulpine

      Carrying things to the ridiculous can be a good way to buffalo your opponent in a debate, but you have to be careful not to take it too far. Much of what you argue can be accommodated with the addition of a bit more technology.

      For instance, your discussion of the California roads and heavy rain; have you never heard of an electric fence? A very tiny bit of research could find you a practical test of such technology on a stretch of freeway that was being re-built on the south side of Washington DC. There is even video of it, where four or five sedans are running down the freeway nose-to-tail no more than a couple feet apart, their drivers holding newspapers up in front of their faces very visibly as they cruise at about 60-70mph up the road. Those cars weren’t using lasers or cameras, they were using low-powered radar to maintain their distance between vehicles and reading a wire in the pavement to maintain steering. Essentially they were following an invisible track. For expressway use and even most Federal and state highways this would be ideal for keeping the cars driving smoothly. But by itself is insufficient for navigation. What it can do is provide that pin-point accuracy to supplement GPS and cellular mapping where those others can have anywhere from a 10-foot to 100-foot margin of error.

      The camera and laser technologies are another way to help the vehicle monitor its location and conditions. The stripes on the roads could easily be supplemented with a metallic component as well, allowing for the on-board computer to make efficient and accurate lane changes even when visibility is reduced through, for instance, snow and ice on the pavement. We humans tend to migrate to the middle of the road in such cases–or the middle of the freeway in order to stay as far away from the edges as we can. This is also one reason why accidents on icy roads are far more common–that and the fact that some people seem to have no concept of how slippery the road really is under their tires. An autonomous car could easily test its grip on the road and adjust its speed accordingly anywhere from hundreds of times per second to perhaps every few seconds depending on conditions.

      BUT… All of this takes not only the technology but also the programming to cover these situations. Additionally, in a manner not too different from the railroads there would need to be a traffic management hub and likely satellite locations to control average speed, the traffic lights and other aspects of a metropolitan area that would affect traffic flow. The advantage is that when such a central hub can control the rate of acceleration, braking and other maneuvering of all the vehicles in the city without interference from impatient drivers running red lights or unobservant drivers sitting at a light for several seconds before realizing the light has changed, the traffic would be able to travel much more smoothly and as a result much more quickly through the city. In the event of a blackout, the vehicles would still have enough autonomosity through communications with the cars around it that traffic need not be stopped, though efficiency would fall as they would have to run in roughly half-block sized ‘packs’ between intersections with members dropping out and new members joining as each vehicle approaches its selected route.

      Yes, autonomy is a very complex issue and some parts of that autonomy may not be realized for a long long time–maybe not until most of us have passed to the next world, whatever that may be for you. Our freeways are the easiest highways to automate as the restricted access and relatively mild curves will be easy for the vehicles to follow using even current technology. Autonomous cars like Google’s can work in a city environment, but it’s not as efficient as it could be and it still has to contend with manually-controlled vehicles. Eventually, that laser dome may be unnecessary as the vehicle’s other sensors pick up the load and more vehicles learn to communicate with each other.

      But even there comes the added risk of security. People have become complacent about their security with their, ‘I gotta have it now’ mentality. Instant gratification has made it too easy for thieves to steal personal identity. And as we’ve already seen, it is also possible now for malicious individuals to kidnap or even murder a victim through their car’s systems. Making these systems more amenable to outside control would also make them easier to intercept. Conversely, it would also make it easier for Authorities to locate and intercept criminals. The balance of personal security vs public security will be a very fine line and it will probably take at least 20 years to find that point where people will ignore the potential disadvantages to accept the very real advantages.

      • 0 avatar
        carguy67

        So, you’re saying all we need is a massive infrastructure expenditure to facilitate autonomous vehicles? Because we’ve–the US, that is–have been so willing to commit massive resources to infrastructure the last few years?

        • 0 avatar
          Vulpine

          Some of that is already in place, carguy. Not fully capable, but the simple fact is that most large cities do have a traffic management office where the traffic flow is monitored and the light cycles modified to at least attempt to improve flow at given times. And yes, these systems are already at least somewhat automated while including the ability for the operators to adjust sequences anywhere from a macro to a micro-scale down to triggering individual lights if necessary. What they can’t handle yet is a power failure in any one location, though they can attempt to adjust the zones around it to minimize the disruption while sending police to try and manage individual intersections until the system can come back online.

          Yes, there will be some expenditure; it’s inevitable. But if a fully-automated system that can go so far as to control the individual vehicles goes online, traffic flows can become far more efficient through simply eliminating the accordion effect at each traffic light and letting the vehicle travel more closely together than is currently safe with human drivers.

    • 0 avatar
      tedward

      Perisoft +1

      We already have, in the newer autonomous cruise cars with electric racks, all of the hardware needed for useful semi auto driving. Software will allow this to expand to new development suburban driving in the near future. Everything after is facing the diminishing returns issue, as well as the truly nasty liability and safety problems that you so perfectly outlined.

      I liked your point about drivers only really piloting in bad weather. That is a nightmare scenario for me personally.

    • 0 avatar
      trackratmk1

      @ Perisoft – To add to your excellent comment, what people seem to chronically fail to grasp is that programmers can’t program judgement calls into these systems. They can only program scenarios they think could happen, and then the proper response from the car. What will the fail safes be like? Handing control back to a driver, or attempting to stop/shut down the car? The trolley problem debate has barely just begun.

      In reality, there are limitless situations that humans, though often incompetent behind the wheel, make judgement calls on everyday.

      It’s been tested already that the most dangerous thing to do is cancel autonomous control and hand it over to the inattentive (and ever more inexperienced) driver in the middle of a “moment”.

      No doubt we’ll have relative autonomy in ideal conditions (highways, good weather, etc), but there is no way for these cars to conquer all conditions, anywhere.

      And that’s before touching the legal ramifications about responsibility; “it’s not my fault, the car did it!” And the insurance issues; “it’s STILL not my fault!”

      • 0 avatar
        Charliej

        This will happen in other countries before the US. Plenty of other countries are not run by the lawyaers. In Mexico, there are no personal injury lawyers. Amounts to be paid in an accident are determined by the government. There is no lottery effect after an accident. A death calls for a payment of approximately $60,000. Self driving cars will be here before the US because no lawyers will be trying to make a big score on the technology.

  • avatar
    Goatshadow

    Even if autonomous cars were the only things on the roads, there would be endless hilarious ways for humans (and nature) to disrupt and f**k with them. Not even taking into account those who would hack their own cars to take advantage of the others.

    Relying on software and programmers to make this dreary, rubbish utopia happen safely is too ridiculously optimistic for words.

    • 0 avatar
      mmreeses

      An all-AV world would be a Manhattan jaywalkers drive come true.

      Got a red? no problem. you don’t have to flee like a rabbit anymore when crossing against the red light.

  • avatar
    thelaine

    The final victory of the mass-transit folks.

  • avatar
    ixim

    Horses to cars was a quantum leap in terms of function, convenience and cost. AVs are a great way to commute. Do/will they serve as well for freedom of movement? What about the privacy issues? I know, there’s no privacy now, anyway. Will an AV tow my boat to the lake, stay with me for a few days and take me and my friends home at a reasonable cost? I know, it’s just a software problem like so many other things I can do just fine by myself but here’s this cool gadget…..but this sounds dystopian to me.

    • 0 avatar
      thelaine

      Agreed. The use of the word “autonomous” is Orwellian. The car better not be and the occupant? Not likely for long. Someone else will eventually be the “expert” in determining your best route and destination, in your own best interest. Just sit back and relax.

  • avatar
    Vulpine

    I’d say you’re close, Mark, but just a bit too optimistic. A full twenty years if you ask me, if not slightly longer. Too many people today grew up with having to drive their cars themselves and few of them will be willing to trust automatics until they’ve proven themselves. That trust will probably be developing over the next fifteen years as these older models slowly fade away and their drivers are pretty much forced to buy newer cars, but with the average age of a car between 10 and 15 years before ultimate retirement and today’s cars only semi-automated, I think we really need to add another five years to your estimate.

    2035, I believe, will be the breakover point.

    • 0 avatar

      The 30 year old Millennials of today (that’s me) are open to this future. We will be 45 in 15 years. Kids born today, who have no recollection of a life without internet, rotary telephones, carburetors, or cassette players will be almost old enough to legally drive … if we use today’s standards of licensed driving.

      I think you might be right, but I also know technology evolves much faster than we think it does.

      • 0 avatar
        Vulpine

        “The 30 year old Millennials of today (that’s me) are open to this future.”

        That’s my point, Mark. The 40-year-olds and older are more likely to resist that change as demonstrated both by some of the arguments above and discussions on other TTAC threads about which vehicle type is better, etc. I’m sure you’ve noted many conversations about pickup trucks where there is unreasonable resistance to even the concept of a compact pickup as a viable type. Most of those resisting are in the 40-plus group even though many of those supporting compact trucks are also in that group. You aren’t likely to see another stage of acceptance until you reach the Baby Boomers who were the ones that first accepted desktop computers as a viable home appliance for communications and entertainment. I’m one of these people and I look forward to the automated car–but also realize its potential limitations and agree with retaining some level of manual control for emergencies.

        The technology will be there in a usable form; the human factor will take longer to accept that technology. My own mother, while taking advantage of SOME technology, is computer-phobic. Even when I try to get her to use a tablet so she can keep in touch with me, she prefers to use a land-line telephone and typically keeps her cell phone turned off. Worse, even at the times she needs her cell phone, it needs charging first. She’s aiming to reach the century mark and really needs the technology to retain her independence, but refuses to use it. I intend to surpass her age and an autonomous car would make that easier.

      • 0 avatar

        15 years is way too optimistic.
        It’s going to be a few yrs before anyone can jump thru regulatory hoops and actually bring a car to market.To expect half the human driven cars to be replaced by autonomous cars in a decade is unrealistic.Do you really expect a GM or Toyota to be able to make and sell autonomous versions of their line-ups w/in a decade?

  • avatar
    Joss

    USPS couriers etc would benefit from Google-type small vehicle for residential deliveries. You get a phone call on it’s arrival go out and get your package. Mind you don’t get hit by a passing vehicle…

    • 0 avatar
      Vulpine

      Wouldn’t work, Joss. Not yet, anyway. That autonomous USPS vehicle could maybe feed your mailbox, but if you’re not home you certainly wouldn’t be able to accept any package it may be carrying.

  • avatar
    sportyaccordy

    Carlos Ghosn had a really astute take on the idea of “autonomous” cars. We are still way off from completely self-driving cars, which to a large degree is a shame. But there are already tons of “assists”. It started with extreme stuff like ABS and traction control, but then there was cruise control. Now we have cruise control that can maintain a certain distance to cars ahead and keep the car in a lane, cars that can park themselves, etc. etc. THAT’S how it’s ALREADY happening.

    It’s no different than a plane pilot… they have hella assists and there are planes out that can’t be flown without assists or computer control. Yea there are drones but they are all unmanned and still controlled by humans. Nobody is getting on a plane without a pilot. So I think there will always be a human element.

    I for one am OK with a larger proliferation of self-driving cars and a bigger car sharing culture. I would love to catch a ride to work, at my own convenience and at a decent speed. Carpooling and PT suck in those regards… but a self driving car that could come and pick me and some coworkers up would be awesome. That would be much cheaper than owning a car all the time, which would let me spend money either on a truly fun car or motorcycle, or more track time. I’d gladly give up driving on the street to commute if it meant I could go to the kart track or do track days whenever I wanted.

    • 0 avatar

      I share this view. As a motorcyclist, I don’t mind giving up road riding if there is a place dedicated to riding where I don’t need to worry about someone texting behind the wheel. People who “commute” are a bigger risk than those who are enjoying the drive in the moment.

      If I could do my errands while I do my errands (maybe making plans for a party on my way to get supplies) then this makes so much sense. On the weekend, I can strap myself to my American iron and go for a rip.

    • 0 avatar
      Vulpine

      @Sporty: Slightly backwards on your technologies. Cruise control has been an option for cars since the late 60s and is now almost standard on every car–though still considered an option by some. ABS became mandatory and to some extent is still not totally trustworthy which is why traction control had to be added. These technologies on average only showed up about 20 and 10 years ago respectively. But you are right; as the technology continues to improve, the cars become more capable of full automation.

      Am I a fan of a car-sharing culture? Not in the sense you describe. Well, not exactly, anyway. Humans are too self-centric to consider a car-sharing system viable except where owning a private vehicle is impractical–meaning in densely urban environments. I’m much more a fan of the suburban/commuter rail system where the personal vehicle takes the commuter to a train station, where they take the train into the city and then use shared transportation to reach their final destination. It would be more efficient in the long run and reduce vehicle congestion within the city, making travel itself faster and more convenient. I use a similar method myself when I visit cities like Washington, DC where driving in the city itself is almost impossible during certain parts of the day. Much easier to drive to the ‘outer loop’ and catch the Metro train/subway to reach the inner city. It’s even possible to reach the airports by train as I did at BWI a few weeks ago.

      • 0 avatar
        sportyaccordy

        The problem with the suburban commuter rail system is it combines the worst of both worlds. You still have to own, insure, fuel and park a car, only to have to pay again get on a crowded commuter train, often only to have to pay a third time to transfer to some other public transportation mode.

        Plus aside from the general suckiness of PT, there are general day-to-day issues like reliability. I grew up and spent ~28 years in NYC so I’m speaking from experience here. If you work in the city and live in LI, and your rail line goes down, you’re screwed. No way around it. If you live and work within subway territory and your line is down, you’re a little less screwed, but still screwed. And it’s not an uncommon occurrence. I lived in the Upper East Side, where there is pretty much only 1 line (the green line). The trains would be backed up for a sick passenger or signal trouble probably 3-4 times a month. It was so bad I rode my bike or motorcycle to work every day the weather permitted.

        Now in NYC there are just too many people and not enough roads for vehicle sharing. They just need to run one long train in a continuous loop on every line during rush hour lol. But I’m in NC now and I drive to work. Car sharing would address all the issues of private car ownership and public transportation. With a floating fleet of autonomous cars I could pretty much leave for work whenever I wanted, and come back whenever I wanted. I could get picked up from my house and dropped off at my job and vice versa on the way home. I could run out to do an errand or go to the gym. If these little pods are fuel efficient, their cost per mile would be a lot cheaper than a privately owned car. And the benefits of a public transportation vehicle unencumbered by fixed routes and schedules are obvious.

        So I think suburbia is exactly where car sharing would thrive. If it were wide spread enough it could put a legitimate dent in private ownership and environmental impact. It seems like a no brainer to me.

  • avatar
    ajla

    I’m a tick younger than you and I don’t expect the steering wheel to disappear in my lifetime.

  • avatar
    Fred

    Another point about converting from horses to cars, was that it was suppose to reduce pollution. Imagine the smell of all those horses pooping in a city like Chicago. Unfortunately cars introduced their own pollution. So I have no idea when automatic cars will be hear. I do predict they will introduce a whole new set of unseen problems.

    • 0 avatar
      Vulpine

      Fred, the odds are high that the robo-cars we’re discussing will be either fuel cell or battery-electric. No pollution in that sense. The hopes would be that they’ll have sense enough to know when their charge is low and where to receive a fresh charge without inconveniencing their passengers. I don’t think that pollution itself will be a problem; with luck, such a shift would see the city become brighter as the old layers of soot and smoke gradually wash away. On the other hand, graffiti will probably become even more evident as these self-driving cars get ‘marked’ by street gangs. We’ll just have to wait and see.

    • 0 avatar
      Syke

      Vulpine, I have to go with Fred on this one. Somewhere in this brave new technology, there will be an unintended consequence. No, not pollution. That one, as you’ve pointed out, has already been factored in. No, it’s going to be something along the “cars are clean unlike horses” because nobody back in 1900 could visualize smog problem.

      And we cannot see the problem now, because it’ll be a problem that currently doesn’t exist in our world because we don’t have the set of factors to make it a problem. Hardly a reason to pull back on the development. But anybody who’s visualizing nothing but wondrous unicorns in his new technology is a fool.

      • 0 avatar
        Vulpine

        Syke, I won’t argue such a possibility; but I also believe that it will have a direct human cause such as the spread of disease or, as one pointed out, an increased risk of terrorist attacks through using the vehicles as an automated delivery method for weapons of one sort or another. Such vehicles used for mass transit will most certainly need additional sensors to detect health and explosive hazards and it may become necessary even for privately-owned vehicles, though survivalists and other gun owners specifically would see that as an invasion of their privacy. So yes, there is the possibility of a different kind of ‘pollution’.

        That said, the strongest opposition will still be that of the human desire for self control mixed with their “instant gratification” mindset. Today’s city-based society is impatient. They don’t want to wait as traffic sorts itself out, they want to push through despite the traffic in front of them. They don’t want their car to decide their route when they know perfectly well that some ‘short cut’ they know will get them through or past the problem faster than any robot would; even when the alternate route is guaranteed to add as much as twenty to thirty minutes to the travel time were that original route not congested. I’ve done this myself, and been chagrinned to discover that the cause of the congestion would have been passed in five to ten minutes had I just been patient and I had therefore lost an additional ten to twenty minutes for no reason other than my impatience. Avoiding the problem entirely may have saved time, but sometimes an incident occurs at a point where even with today’s rapid reporting tools, you’ve already passed the exit before you learn of the backup and the next exit is two miles down the highway with heavy traffic between you and that incident that’s guaranteed to back up before the next exit. As the character in a certain MMORPG says, “Slow downnnnn.” Enjoy the world.

  • avatar

    The key word is… other traffic not autonomously driven. Experts think that it will decades to phase out conventional cars. Nothing is gonna protect you in your robo-car from a reckless driver hitting you in the rear for instance. You may even run a higher risk of getting hurt. The Google pods are not meant to use in manual override mode.

    • 0 avatar
      sportyaccordy

      I mean you don’t even need to be an expert to see why wholesale change won’t happen anytime soon. Car notes are longer and longer so at the minimum we are looking at like 10 years away. And then there is the small issue of the tech not being ready yet to even begin a full phase out, as well as the legal framework for liability and all that. So we are a long ways away. I’m thinking 30-40 years until we even begin to approach half… plus there are plenty of people who merely enjoy driving and will continue to do so themselves. So all the chicken littleing is silly. People aren’t thinking this stuff through.

      • 0 avatar
        dal20402

        “plus there are plenty of people who merely enjoy driving and will continue to do so themselves.”

        They’ll have to enjoy it enough to pay the insurance rates, which will be much higher on human-driven cars once robot cars are widespread. I don’t think many people are that committed.

        • 0 avatar
          sportyaccordy

          If robot cars are safer, why would insurance rates be higher? Insurance rates are (at least tangentially) tied to what insurers feel is your risk for a claim, which is an accident. If robot cars = less accidents, if anything they should drive insurance premiums down.

          • 0 avatar
            dal20402

            Payouts will get bigger for accidents resulting from human-driven cars when human driving is an avoidable choice. It won’t be “your driving was a bit negligent.” It will be “your choice to drive yourself instead of letting the car do it was a monstrously dangerous act.” Expect the payout for a lost life stemming from bad human driving to go from the $1M range to the $10M range, and required coverage amounts on human-driven cars to go up accordingly.

          • 0 avatar
            sportyaccordy

            Why would a payout be any different for a human vs a robot? Damage is damage, injuries are injuries. Robot cars = less accidents = less claims = cheaper insurance. There are studies on this.

            http://www.iii.org/issue-update/self-driving-cars-and-insurance

          • 0 avatar
            dal20402

            “Why would a payout be any different for a human vs a robot? Damage is damage, injuries are injuries.”

            You’d think so, but that’s not the way it works in litigation. First, there’s punitive damages. Second, juries will award greater compensatory damages when they are more outraged by the facts. Eventually as the differing accident rates become more and more clear choosing to manually operate a car will start to seem like something worthy of punishment.

          • 0 avatar
            sportyaccordy

            Even assuming what you say is true (and it’s not- most claims don’t get anywhere near a court or have punitive payouts), this doesn’t change the fact that robot drivers = less accidents = less claims = cheaper insurance. The only way your theory would make sense is if you could demonstrate how robot drivers would cause more accidents, because accidents = claims. But history shows tech >>>> humans for driving. Accident rates were dropping fast when stability control became mainstream, and started to rise again once people got smartphones. So we are the problem here.

            It seems to me you don’t like the prospect of robot cars but haven’t really thought through why you don’t like them. No, insurance for human drivers won’t increase…. all actual data and logic indicates they will decrease. No, there will never be a point where all cars on the road are fully autonomous with no human override. There are too many legacy cars still on the road with plenty of life in them and there will be for years to come, and there is too much money and profit in cars that are fun to drive.

            Let’s stop with these paranoia stuffed strawmen of how awful automated cars will be. These discussions have really brought out some of the worst critical thinking I have seen in the auto enthusiast realm.

          • 0 avatar
            Pch101

            You are completely misunderstanding Dal’s point.

          • 0 avatar
            sportyaccordy

            No, I get his point. He thinks manual driving will be punished because it will be deemed too risky. Again, reality doesn’t bear this out. I ride a motorcycle. Liability/personal injury coverage on it maxed out in NC is something like $200 for the year, compared to 7-8x that for full coverage on my car. I used to drive a 350Z and now I drive a Civic EX. Full coverage on both costs the same, despite the Z having been rated as the deadliest car on US roads in its respective model years. So by dal’s logic motorcycling should have been essentially outlawed by insurance; and yet, I can insure myself against injury for a year on a vehicle wth zero safety devices and performance rivaling a Ferarri for less than 2 months of coverage for a car with airbags, ABS, crumple zones etc.

            Or maybe dal’s point is that we should toss logic and real world evidence out the window and just be opposed to autonomous cars out of fear? Either way I can’t really get with what he’s saying. He’s not speaking rationally or empirically.

          • 0 avatar
            Pch101

            Dal is a fan of automated cars. You really aren’t getting his point.

          • 0 avatar
            Vulpine

            “Damage is damage, injuries are injuries. Robot cars = less accidents = less claims = cheaper insurance”

            True… except that a robot car tends to be far more expensive than a manually-driven car and so damage is 2x-3x more damage to repair. Until robot cars outnumber manually driven vehicles significantly–to the point where it becomes obvious that the manually driven vehicle is the predominant cause of motor vehicle accidents, the insurance rates will at best remain level and at least for a while will probably be higher, the more automated the vehicle is simply due to the cost of repairs on that automated vehicle. In fact, the risk is high that they will be declared ‘totaled’ more often as the high cost of repair would probably exceed 90% of its replacement value. Which is why so many cars in even moderate-speed crashes are declared totaled. About the only “wrecked” cars that return to the road with an insurance repair are the ones surviving low-speed crashes that were drivable from the scene or which didn’t suffer some form of “frame” damage.

            Robots don’t have to cause the accident to be involved in an accident. But that robotic car will be more expensive to repair–if at all. Either way, the insurance company will have to pay out on the damage and until it can be proven comprehensively that a human driver caused the accident, even if through negligence and not even involved in the accident, those high premiums will remain. It can only be hoped that the sensors on the robotic car can capture the data more clearly than human memory. Dash-cams are already proving effective towards determining the real fault in a crash.

  • avatar
    trackratmk1

    Look, the average car age TODAY is 11.4 years.

    The majority of the car market will continue to purchase traditional cars certainly for the next ten years, given OEM future product planning and product lifecycles. I George Foreman guarantee it.

    If that’s the case, and that age holds relatively steady, in 2030 you’re still going to have most cars on the road dating back to 2020.

    Also, let’s work backwards from your 2030 guess. You’ll need manufacturers to be selling well over 50% of all new cars as fully autonomous cars for many years prior to that, in order to displace the existing used cars at a fast enough rate. And again, product planning and real world prototyping would have to be pretty much done in the next couple of years to scale up AV’s that fast.

    So no, not a chance by 2035 – even assuming all the practical obstacles of AV’s are already overcome.

    • 0 avatar
      tedward

      Thetruthaboutinertia.com

      It’s not just the product planning timeline either. It’s where the full set of sensors falls on your average vehicle’s trim level ladder that also determines adoption rates. There’s customer preference as well. I knew I don’t want a sensor studded bumper the next time a family member unavoidably contacts a deer, I’d consider that a huge deal breaker in fact.

  • avatar
    dal20402

    As soon as the technology is proven and available at a relatively wide variety of price points, insurance companies will handle the rest, and the transition will happen within a decade.

    Insurance companies will experience losses from robot cars that *begin* at half the rate of human-driven vehicles (when robot cars are still rare) and get much better than that as robot cars get more common. Rates on human-driven cars will slowly rise and eventually become unaffordable to normal people.

    The technology is closer than a lot of you think, also. Rain or fog? No problem these days. Dense city driving? Getting there. Snow coating the road is still a challenge, but that’s also a pretty rare scenario in most heavily populated places. Odd situations? Either a mode that gives the human driver control after coming to a full stop in the safest possible place, or asking the human driver questions (“point on the touch screen to where you want me to park.”

    I’d guess we have “good enough” working technology in 2020, a low-cost implementation by 2027-8, and a majority of cars on the road autonomous by 2035. The replacement cycle will be faster than it is today because 1) this is the first really compelling feature in decades for the average buyer and 2) lower insurance rates (or, more precisely, higher insurance rates on human-driven vehicles) will make up part of the cost.

    • 0 avatar
      tedward

      Unless the insurance companies experience a higher rate of rear end collisions, face higher average repair costs or face an increase in legal costs with these hypothetical cars. In that case the opposite happens. Right now actuarial data is four cars that already have obscene parts cost, so we don’t know these things yet.

      • 0 avatar
        dal20402

        I would bet my house against that. For insurance companies, robot cars will for all practical purposes wipe out the collisions where the insured is at fault. At first they won’t change the collisions where the other party is at fault, but as more of them are on the road those will go down too. So many of you are focused on bizarre edge cases while the money is in the reduction in routine accidents.

    • 0 avatar
      sportyaccordy

      Like I already explained, human rates won’t rise, they will fall. Less accidents = less claims, less claims = less payouts = lower premiums. You see this in pretty much every form of insurance… the higher the collective risk of a group the higher the premium. Human drivers will be sharing the road with robot drivers, which will make the roads safer for everybody and reduce costs.

      I know you are scrambling to find an angle to damn robot cars but this is not it. Insurers don’t “punish”, they price risk, and as risk goes down so do prices. This is a very simple concept that reality demonstrates everywhere.

  • avatar
    Pch101

    The autonomous car ain’t gonna happen, for reasons that I’ve explained elsewhere. Cars will have more autonomous-style features, but full autonomy is a pipe dream.

    • 0 avatar
      trackratmk1

      Pch, can you link me to your other comments? I have heard virtually no rational counterarguments to the AV fantasy.

      • 0 avatar
        mmreeses

        We (well say 50%-80% of Americans) already already live in a world where you can push a button 24/7/365, wait 3-15 minutes, get delivered door to door without the hassles of car ownership.

        It’s called a Uber or a cab or a black car. The main thing (though it’s a big only) that Google robo-cars provide is the safety of robo-drivers.

      • 0 avatar
        dal20402

        His argument boils down to “no one will take on the liability because robot car crashes will scare people more than regular car crashes do today.” Even though robot car crashes will be far fewer in number.

      • 0 avatar
        Pch101

        The short answer is that it will take only a few spectacular automated crashes for the public to demand human backup systems. And such crashes are inevitable, since it is impossible to have heavy objects moving on roadways at high speed without something eventually going wrong.

        (The human backup systems will actually make things worse in the big picture. But we tend to focus on anecdotes rather than data, so that is irrelevant to how opinions are formed.)

        • 0 avatar
          dal20402

          You think the anecdotes will force behavior that is contrary to what the data would suggest, forever. I think there is too much money involved for that to happen, and the insurance companies’ green eyeshades will easily triumph over the bad publicity, particularly as people get more accustomed to robot cars.

          • 0 avatar
            Pch101

            A prime example of how Americans can’t figure out how statistical data works: American gun control vs. the rest of the civilized world. (Guns statistically make people less safe, but try convincing gun owners of that.)

            Another: American healthcare vs. the rest of the civilized world. (Similar problem.)

            Let’s be serious here: When Christine the angry car kills grandma, you really think that big picture arguments are going to be tolerated?

          • 0 avatar
            VoGo

            I’m usually with PCH, but not on this one. We Americans tolerate the loss of 40K people annually to drunk, distracted or just plain lousy human drivers a year. Why would we bat an eye if the driver’s name was Google, rather than Jimmy?

          • 0 avatar
            dal20402

            We’ve had this argument before, but within a few years of Christine the angry car killing Grandma, what won’t be tolerated anymore is Chris the angry driver killing Grandma, which happens all the damn time.

            If road fatalities were a new and different thing I might be more inclined to agree with you. But the reality will be far clearer, because the numbers will be easier to compile than they are with guns, and it will be easily understandable to the public: “Autonomous cars reduce the risk of at-fault accidents by 98 percent.”

            Even if that is hard to understand, the public will easily understand “insurance for your autonomous car is one-third the cost.”

          • 0 avatar
            Pch101

            When an individual kills another, we blame that one guy and move on.

            When the robot does it, we’re going to blame the underlying technology and the failure to have a human who could have stopped it from doing us harm.

            The argument that we are better off in the long run by letting machines do all of the work is not going to be accepted. We’re not going to let Christine out without a leash.

          • 0 avatar
            dal20402

            “When an individual kills another, we blame that one guy and move on.”

            Maybe you do. I don’t. 35,000 road fatalities a year isn’t because there are 35,000 bad people a year deliberately killing others. It’s because of systemic problems, some of which would be solved by robot cars. I think the public knows that on some level.

        • 0 avatar
          Vulpine

          Historical anecdote. The American F-111 fighter-bomber aircraft suffered more crashes when flying at low altitude under TFR simply because when the pilots grabbed manual control away from the TFR system, they couldn’t react quickly enough. In every instance where a TFR plane struck a hill or mountainside, it was due to the pilot snatching control in panic.

          The fix? When the pilot grabbed control, the TFR automatically kicked the plane up into an emergency climb BEFORE releasing the controls.

          That was 60 years ago.Any spectacular crash of an autonomous car will almost certainly be due to the front-seat passenger overriding the automation.

        • 0 avatar
          sportyaccordy

          Such crashes are hardly inevitable. And you are doing worse than anecdotes… you are basing a theory on pure speculation lmao.

          I agree that wholesale autonomous cars will never happen, for a myriad of reasons, but the biggest two being logistically it’s just not possible to do any time soon, and more importantly there are plenty of people who just don’t WANT fully autonomous cars. There’s too much money in manual cars as well.

          If anything, enthusiasts should be looking forward to the availability of autonomous cars. Manufacturers will have to focus more on drivers/performance cars because that will be who is the main market for non-autonomous cars. Insurance will be cheaper because roads will be safer. So on and so forth.

  • avatar
    05lgt

    These dreamy eyed descriptions of the bliss of being a passenger… I prefer to drive. There are many times when there are more than one qualified drivers in the car and given the choice it’s me at the wheel. Call a friend, ask a spouse, taxi/Uber/black car… NO! I like to DRIVE. Am I the only one?

  • avatar
    ixim

    I’m plussing 051 there. Why are we conflating safety with another form of mass transit? The economic argument is with the latter. People will always find ways to get hurt traveling. AVs safer? Different, more technology driven risks but unintended consequences are guaranteed.

  • avatar
    Superdessucke

    The way people drive around me, one might argue that it’s already happened.

  • avatar

    Outnumber not before 2050 minimum.

    • 0 avatar
      RideHeight

      I agree, at least that long.

      One big suck about being a Boomer is that we’ll all be watching some awesome technology that’s just on the cusp of widespread adoption when we go into our terminal nap.

      Autonomous cars, organ regrowth and cognition recovery come to mind for me.

  • avatar
    wumpus

    When will it happen in Japan? I’ve often claimed that self-driving cars appeal to three types of people. Drunks, soccer moms, and the elderly. The point being that the elderly always vote, the soccer moms usually vote, and the drunks often lose politically even with the entire liquor industry behind them.

    Then there is Japan with its famously aging population and its equally famous high-tech ways (personally I can’t take Japanese software seriously, but maybe they will import it). My guess is that whatever google and co. wants to do with autonomous cars, they will be able to do in Japan (assuming they are willing to sufficiently humor MITI).

    I suspect the US could easily make plans for autonomous cars by watching Japan, but will simply blunder blindly into a brave new world in the name of “American exceptionalism”.

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • The_Imperialist: I’ve had my Venue since July and this review sums things up quite accurately. If I had to...
  • Diewaldo: That has to be the most ugly car of the year so far. It looks … just wrong from every angle. I have...
  • wolfwagen: While Going to HS in eastern Pennsylvania, we had a new girl that came from Bayonne, NJ. Pretty attractive...
  • jagerninja: This is kind of unrelated, but what’s with the Lexus ad video that appears to be embedded on every...
  • wolfwagen: Agreed

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Mark Baruth
  • Ronnie Schreiber