March 26, 2014

The Driving Force behind Autonomous Vehicles

While developers like Google believe autonomous vehicles (AV) – or self-driving cars – will be market-ready by 2018, others are more sceptical.  Whether the reasons are legal, technological or human, doubters suggest there may be lots of reasons why this particular utopian vision will not be ready for prime time.

AVIndeed, it may be that a fully autonomous vehicle has a ways yet to travel – but I expect that we will see aspects of the technology introduced ASAP, if not actually mandated.

The reason?  Safety.

If ten of thousands of people can be saved from being maimed or killed by the introduction of sensors that can respond more quickly than any human is capable of – and then override an inattentive driver – why wouldn’t we want that technology in every new vehicle?  Indeed, why wouldn’t government require it?

Or if not government, then how about the insurance industry?

And that’s when things get interesting.  Presumably there will be a substantial discount for any car outfitted with accident-preventing features – and a commensurate increase for any car that doesn’t.  Arguably, the differential may be so great that many people simply won’t be able to afford the liability of not having the latest technology, and also won’t be able to afford a new vehicle that does.

They will then be faced with choosing the obvious alternative: car-sharing.

Already car-sharing offers a practical alternative for those who don’t need, don’t want or can’t afford a full-time vehicle that sits idle most of the time.  And even if it won’t be possible for some time to call up an AV that can drive independently to a customer’s location, the cost advantage of subscribing to a service that offers the latest options, without the burden of individual liability insurance, will make car-sharing substantially more competitive than it already is.

Further, the fleet managers will maintain incredibly complex vehicles impossible for the average person to service on their own or provide a performance guarantee for resale.

A further advantage to government will be the ability to introduce road pricing almost invisibly – as a tax on each trip or contract, integrated into the overall cost, calculated by the internal software, paid through the service provider, without the consumer being necessarily aware of the tax imposed.

Oh, there will be lots of unexpected and unpredictable consequences, as there always is with revolutionary technologies and the behaviours that result.  But I believe the reason this advance will occur far faster than many expect is the disproportionate effects of insurance costs once  the technology is mandated.  And given the number of lives it can save, how can it not be?


Credit to Michael Ohnemus, a student of Urban Studies professor Anthony Perl, whose research paper on AVs stimulated these ideas.


Posted in


If you love this region and have a view to its future please subscribe, donate, or become a Patron.

Share on


  1. This assumes that insurance companies will actually increase base rates rather than just apply discounts to the newer technology. I don’t see how the insurance industry on it’s own will do that unless drivers become liable for larger damages in the case of a collision.

    1. That’s right. Presumably cars driven by humans will crash no more than before, and perhaps less, if the robo-cars around them are better at compensating for their mistakes. It would be lower insurance for robocars, and the same for human drivers.

  2. Love to see that ..


    imagine an accident between two automatic cars.

    Who is at fault:
    a) the first or second car’s electronic system
    b) the first or second car’s environmental sensors
    c) the first or second car’s brakes
    d) the GPS system
    e) the wrong human override given by the driver of the first or second car
    f) the speed sensor signals, owned by the City of Vancouver, sending he wrong signals
    g) the first or second car’s tires
    h) …

    Many things will have to be tested first in a lab, then on a very small scale, then in pilots in the real world, then on a mass scale .. and that will take DECADES !

    The idea of cell phones existed in the 1970’s .. but really caught on in the 2000’s only on a large scale, over 30 years later, with far less accident potential and far less $ per device .. in a $15,000+ per car environment it will be 40-60 years !

    Fascinating though .. and probably available as features in luxury cars first, on certain inner city or highway segments in the 2020’s to 2030’s

    1. Just the opposite. The first will probably low speed share vehicles in city centres. A town in the UK is planning on having a fleet of 100 up
      and running in 2017. They will seat two people and only go up to 30kph. If a crash does happen, it won’t be that serious.

      At highway speeds, errors will be fatal so the equipment and vehicles will need more expensive components that won’t fail and require much more testing.

    2. Remember, driverless cars are already on the road. Though everybody worries about the legal, I don’t think that it’s as big a problem as people think it will be.
      After millions of km, there has not been a single crash, let alone injury.
      Driverless cars will be coming as soon as 2015. Get ready, and brace yourselves.

      1. 2015 ? I very much doubt that. Planes would be far easier, and most planes still have pilots. Only drones in remote regions have no drivers. And a road is far more complex than an empty sky with no pedestrians, bikers or other vehicles close by.

        We will see gradual implementation, say GPS or parking or automatic braking when too close, or side sensors, or boring highways .. and yes we will see pilot projects here and there by cash rich firms like Google, Microsoft, Apple, BMW or VW to test certain features and get marketing mileage .. but for regular folks sitting in the back seat reading a newspaper ? Say 2050 or 2060 .. 30+ years from now at best !

        1. Google has over 500,000 km driven in uncontrolled settings (ie LA and San Francisco). Autonomous cars are already legal in a number of states, and a few manufacturers have plans for commercially available cars within 4 years. It’s coming MUCH sooner than you think. The technology is ready now.

        2. Of course, it might not just be Google or any North American City that gets driverless cars first. Remember, China’s light-years ahead of the US when it comes to technology. They have already 10,000 km of high speed rail lines, the US & canada have one combined (and it’s not a true HSR either).

          It’ll probably be a chinese city that first implements mass use of driverless cars. The chinese are watching closely on google’s developments, just like how they launch their version of facebook (called weChat).

  3. I watched autonomous heavy equipment driving on a mine site years ago. Self parking is now available on the Ford Focus, hardly a luxury vehicle. It is wrong to see the move to autonomous vehicles as all or nothing, we will see incremental implementation and that has already started.

    In terms of sorting out accident liability, imagine scenario one: Two cars crash, he says one thing/she says another. Scenario two: There is a video and computer record of the vehicle’s actions right up to impact. I think scenario two would be easier for the insurance adjuster to deal with.

  4. Autonomous cars are basically PRT. The word “car” will be inaccurate, because there would be no need for privately owned cars. All “cars” would in a sense become transit vehicles. It’s exactly like how elevators work, except the ‘elevators’ come to you, and can go anywhere, everywhere.

    The societal implications of this would be just like the people on the axiom in Wall-E. The main barrier is legal, but I don’t think that’s where the main troubles will be. Like how the car invaded the 20th century, the driverless PRT machines will change the way we move in the 21st.

  5. 1) No one even notices that our Skytrains are driverless.

    2) This change will happen quickly because Google is behind it and they can make a lot of money knowing where you have been.

    3) Everything will change. Our cities may be wonderfully improved.

    1. Yes, in time. Skytrain has no pedestrians walking on track nor cross traffic nor bikes. Let’s not confuse a pilot project with a few dozen cars in a controlled setting to a mass rollout. Big difference.

  6. “And given the number of lives it can save, how can it not be?”

    Ever heard of photo radar?

    I agree with Jeff that this will happen incrementally over time. But pricing new cars out of the reach of people via government mandate – doesn’t seem likely to me…

  7. I for one welcome the idea of the Car2Go that comes to me when I need it, and goes away when I can’t find anywhere to park it. And if I can read or watch a video instead of driving it, so much the better

    1. That would be lovely .. I think unions will block that as long as they can .. namely the cab driver and the bus driver unions .. but yes, it makes far more sense especially on less traveled evening routes that having an empty bus circle around every 20 or 30 minutes like we see all over Vancouver !

      There are smart phone apps today where you can catch a ride with people from A to B for a small fee .. and even that is challenged in court in many US states !

      See here for car share apps:

      or here for mere cab apps:

  8. The insurance industry will raise premiums for cars that lack the proper sensors because the cost of personal injury claims continues to climb faster than almost everything else. Those with “improved” cars will initially get discounts, then rate freezes and eventually they’ll join the inflationary ladder. By then almost all cars on the road will be semi or fully autonomous.

    The only disagreement I see is in how long the transition will take.

    One of the areas where I think drivers will be angry is autonomous cars adhering to speed limits. People have become accustomed to doing 60-70km/h between red lights and get angry when heavy traffic takes that option away from them. Some may take out their frustration on highways where going 150 in a 100 zone is already common. Fortunately we already have a generation accustomed to watching movies while getting from point A to point B. They would much rather have a robo-chauffer than be forced to stay in one position paying attention to the road for hours on end.

  9. I share the author’s optimism and the natural link to carsharing, especially the one mentioned by Stephen Rees, allowing cars to drive themselves to the next user, rarely sitting idle and unproductive.

    The car cost will climb, but the insurance cost go down; so we will be bribed in a sense to let the car be connected to the grid and to drive itself. Shared cars will be the best way to deal with the higher costs. It also allows the province to feel less resistance to telling a driver (e.g., aggressive, distracted, repeatedly-DWI, or a older, dotty one) to never sit behind the steering wheel again. Shared cars connected to the grid allows the authorities to know who is behind the wheel of every car and in control.

    And cities, which are getting ready to charging car users (not necessarily drivers) for their use (distance, time of day, local congestion factors) will also like the plugged in cars sending out data continuously (allowing collisions to be dissected to the nth degree).

    When we know that an AV can detect pedestrians and cyclists and avoid them 100% of the time, kids and elderly grandparents will once again be allowed to walk autonomously on our streets.

  10. Maybe adding coexisting technologies such as our BiModal Glideway Dual Mode Transportation System, which offers a separate glideway for autonomous use only. This will help with congestion and the safety concerns between autonomous and manual controlled vehicles.