Automate The Freight: Autonomous Buses To Start Operation In UK

The UK will get its first full-size autonomous bus service this summer, if final road testing that begins in the next two weeks goes according to plan.

Known as Project CAVForth for the UK government’s Center for Connected and Autonomous Vehicles (CCAV) and the Forth bridge, over which the buses will travel, it is said to be the most complex test of autonomous on-road mass transit yet undertaken in Europe. The full-size single-deck motorcoaches, five in total, will ply a 22-km (14-mile) route into Edinburgh from Fife, crossing the famous Firth of Forth on the Forth Road suspension bridge. The buses will carry about 36 passengers each and run at SAE Level 4 autonomy, meaning that a safety driver is optional under good driving conditions.

The riding experience will not, however, be entirely human-free. Based on focus group feedback, the bus company will be recruiting “Autonomous Bus Professionals” from their pool of trained drivers. Two of these employees will be on each bus — one in the driver’s seat to monitor the systems, and one to serve as a “Captain,” who will walk the aisle and chat up the passengers. It seems like a good idea to allay any fears the passengers might have about looking up and seeing nobody driving the bus. There’s historical precedence for this; elevator operators were very much a thing even well after automatic controls had been figured out, for precisely the same reason.

The central contention of “Automate the Freight” is that full autonomy will be deployed to commercial trucking long before it becomes mainstream in the consumer space, and when you consider that buses really are nothing more than trucks for human cargo, this project supports that argument. And when you think about it, public transportation is a great test case for full autonomy — the route is well-defined, it’s laid out in an urban setting with a lot of infrastructure, and there’s the bonus of self-loading cargo.

Granted, the stakes are a little higher when you’re carrying freight that has the tendency to sue if anything goes wrong. But somebody has to go first, and it’s encouraging to see that the project’s participants and sponsors have enough confidence in the system to field it. We’ll be watching carefully to see how it turns out.

[via Singularity Hub]

40 thoughts on “Automate The Freight: Autonomous Buses To Start Operation In UK

    1. Not only on the road but also in the buses themselves.
      A part of the job of a bus driver is to act as an authority figure and keep an eye on their passengers.

      It may even be a mayor factor in whether this succeeds or fails.

    1. Well, you got to get paid somehow.

      AI vehicles actually create jobs more than they lose, because there’s the programmer, the manager, the system administrator, technician, procurement people, all the supply chain and logistics for making the electronics, the legal department, the public relations team, the advertising team… but we must also not forget the other company staff like cleaners, clerks, security, the guy who serves food at the cafeteria…

      In the end, there’s bound to be more than one person per bus making the AI drive the bus around.

      1. Most of that supply and support chain already exists for normal busses. It is likely that AI busses WOULD eliminate the job of “bus driver” from that list. Although I think they’ll have to add a “bus guard” back into the equation so that there is someone on hand to keep an eye on things and deal with issues. In the end it’s doubtful it’ll have a major impact on the number of jobs, but it might negatively affect the pay of some levels of those jobs.

        1. @ThisGuy “it might negatively affect the pay of some levels of those jobs”
          I don’t agree as the “driver” must have a greater skill if (s)he is needed to interrupt the AI and take over the driving.

        2. The company that manufactures the bus is different from the company that operates it, is different from the company that provides, installs, and maintains the AI system. This way they can all sell services to each other, which counts as growth of productivity in terms of GDP, rather than just an increase in the cost of operating a bus.

    2. I work in a bus company, and we and all our competitors have a huge staff shortage – even before covid – even though the job does pay well for the base skill level (as in: there’s not much jobs you can get that pay this good with the required school diploma, 40 hours/week on average, with paid breaks and 45 free days/year).

      The main problem is that you can only become a bus driver at 21 years old, while people who love driving big rigs can become a truck driver at 18, and if you’re willing to make the hours, earn more at a young age as a truck driver.

      The cost of the driver is about 40% of the total operating costs of a bus, but a service agent (without a driving license) would be a lot cheaper, and you could omit those on non-revenue trips. This cost reduction could go into offering more transit which means more people have access to jobs.

      As a bus driver your job might be in jeopardy, but for the population as a whole, 250 years of mechanization and automation has shown that there is no risk for a decreasing number of jobs.

      1. > 250 years of mechanization and automation has shown that there is no risk for a decreasing number of jobs.

        The question is rather, are those productive jobs? When a person moves up to a higher economic sector, the work that they do begins to consume resources instead of producing or facilitating the production of resources. A bus driver facilitates work and economic activity, whereas a youtube star merely consumes electricity and industrial goods for trivial amusement.

        The risk is that with more and more automation forcing people out of productive jobs, they’re no longer able or even required to return their own value to the society. We’re adding the cost of the automation on top of the cost of the people and the actual economic efficiency of the system goes down instead of up.

        1. What that last part means: we need more natural resources to sustain the same population. More energy, more materials, plus there’s the trouble of distribution when your wealth becomes totally decoupled from the concrete value you put back into the system.

          Income disparities rise (already has) as the economy is no longer based on basic production and supporting services, because income becomes a matter of playing the market and making other people consume more rather than creating new consumable wealth.

        2. I would suggest people like your youtube star can work out very profitable to society – LockPickingLawyer for instance has content full of mechanical details, security and social engineering that can perhaps spark good ideas in others, ThisOldTony despite all their humourous antics has still managed to introduce me to new concepts of workholding/tools/methods to make a part that again can help spread good ideas that work out directly useful to the production of goods. (To pick just two of who knows how many good options – including more story telling stuff like D&D games – dealing with tricky issues and alternative points of view has real world benefits)

          Just because you are one step or more further removed from something doesn’t mean you didn’t have an important part to play, even if its entirely unknowingly.

          1. For every old Tony there are a hundred people who just copy other people’s videos to steal advertising money, which in itself is a horrible way to make money because the actual payers cannot negotiate over the price – it’s just transparently added to the price of everything.

          2. >Just because you are one step or more further removed from something doesn’t mean you didn’t have an important part to play

            Indeed. It just gets more difficult the further up the abstraction ladder you climb, because it becomes less obvious how you are (not) helping the society. It’s exceedingly easy to wind up with an economy based on well disguised pyramid schemes.

          3. Oh wow Scam artist on the internet now, what a shocker…

            Considering that sort of cheap ripping other peoples hard work off happens in every possible area of life – really decent paint/soap/chocolate/etc brand puts in lots of work to make a good product, in a super ethical environmentally sound way and looses out to scam artists and cloner making heaps more money doing a good job of pretending to be them just long enough to get money of somebody…

  1. @Dan Maloney said: “Based on focus group feedback, the bus company will be recruiting “Autonomous Bus Professionals” from their pool of trained drivers. Two of these employees will be on each bus — one in the driver’s seat to monitor the systems, and one to serve as a “Captain,” who will walk the aisle and chat up the passengers. It seems like a good idea to allay any fears the passengers might have about looking up and seeing nobody driving the bus. There’s historical precedence for this; elevator operators were very much a thing even well after automatic controls had been figured out, for precisely the same reason.”

    This is good – autonomous vehicles mixed with unpredictable wetware drivers is asking for trouble – baby steps please. Before cutting autonomous vehicles loose on their own in a mixed driver (human + robot) environment, collect lots of operational safety statistics over a significant period of human-supervised autonomous operating time. An automated driver must be orders of magnitude more reliable than a human driver before it is allowed to operate without supervision, something I feel we are a long way from demonstrating in a mixed-driver environment.

    Dedicated elevator operators are in my opinion a bad comparison. Early elevators had mechanical gates and doors without safety limit switches and in most cases a person was needed to manually stop/start the device at each floor. For practical and safety reasons a dedicated human had to run a device like that. As soon as full automation was introduced, the elevator operators were rapidly retired. There was no significant overlap period to build user confidence in automated elevators. Anyway that’s how I saw it when I lived through that period in America.

    There were and still are exceptions. Take Japan for example, even today in high-end retailers and glitzy high-rise buildings it is possible to find impeccably dressed attractive young elevator operators greeting passengers, holding doors, and pushing buttons, even though the elevator is capable of safe fully-automatic operation. In Japan, elevator operators are a sign of culture and respect – and yes, nostalgia.

  2. There is ZERO comparison between a redundant-relay-based elevator and a poorly-trained (as in EVERY) artificial neural network.
    This is a REALLY bad idea. They should be losing their jobs BEFORE anybody gets hurt, NOT after.

      1. Yikes! That looks like exactly the sort of sci-fi horror I would’ve loved back in the day, when computers could only do what was explicitly-programmed.

        [Hey, bold-deniers… believe it or not, we can’t do italics in here. And, believe it or not, there are MANY folk in the position to make decisions who don’t understand that finite-state machines differ VASTLY from Neural Networks. If my short-bold-statement gets even ONE of them to question decisions like these, it’ll be worth the ostracism you gave me.]

    1. Those who cannot do specialized jobs become entertainers and trivial servants, stock market gamblers, “artists”, etc.

      It’s basically the same as putting them on permanent welfare, with the difference that they have to do some make-work which consumes further resources to earn their keep.

      1. And of course: politicians. One of the most pointless “jobs” invented in the 21st century is a “social media influencer”, which is basically a modern version of a petty demagogue who picks on trivial issues and invents problems for money. Almost like an updated version of a TV-evangelist.

    2. I agree with you to some extent.

      I live in a place were we have 1 bus that comes to this side of the woods and takes you to the nearest train station. It does not operate on weekends or very early/late hours during the week. Because it costs too much to operate the bus during those times due to the lack of passengers who pay for it. Well, if I go to work and have to leave late for some reason, now I’m left without a ride.

      This would allow that bus service to be available during those times. Not to mention that it would allow for service to neighboring towns too, since right now that does not exist at all.

      1. Well, take the other guy who said the driver is about 40% of the expense of running a route. Of course the machines aren’t free either, so suppose the automated bus would be 30% cheaper. How much more service would that get you? Would that solve your issue?

  3. Andrew already pointed out the DLR, but there’s a few others too… :-)
    * Glasgow Subway (Since 1980)
    * London Underground: Victoria Line (1968), Central Line (1999), Jubilee Line (2011) and Northern Line (2014)
    * Numerous APMs at Birmingham, Gatwick, Luton and Stansted airports, and other locations.
    * Crossrail (The Elizabeth Line) and the Cwmbrian Coast lines use ETCS which means automated operation is already supported, but not necessarily in use.
    # TVM 400 signalling (HS1; Eurostar, Southeastern) supports supervised autonomous operation on controlled high-speed lines, but this is very rarely used.
    (All sources: Wikipedia)

    The British mainline network will probably move over to ATO as time, technology and the erosion of workers rights progresses (ASLEF will put up a big fight against this!) but Britain’s present desire to be a “World beating” country means that there will probably be a few nasty accidents along the way. Tories value profit over human life as we all know, so the technology will be rushed in even if it’s only UKCA compliant… ;-)

    …And that’s why Britain testing autonomous buses scares the Brexit out of me. In France, Germany or Japan such tests would go as well as they could – They might be a bit iffy, but they’d be no riskier than human-driven buses – But without European safety standards to protect us who knows what could happen. The saving grace is that with the tests taking place in Scotland human safety will still take precedence, which is not something I could be assured of for tests taking place in England.

    1. If they really valued profit over life there would have been no pandemic lockdowns at all, very bad for business, and as bad as Covid was it was never bad enough to pose much threat to society as a whole, pretty much just the folks that are at greater risk by nature of being already ill or old really would have on average come off worse without.

      For most folks in the working productive societally very useful age range maybe they’d end up taking some extra sick days, for a few maybe they end up with half pay etc for longer term health absence, and almost nobody in that range comes off worse than that or dies – it would still be a big number as when the total population is in the millions even half a percent is a very scary sounding large number…

      Not sure how Scotland works out ‘safer’ than England either, or what Brexit has to do with anything – UK standards for most things are often tougher than the EU at least when introduced, and no EU standards got thrown out with Brexit (yet anyway).

      Also a few nasty accidents is 100% certain with anything given time, including long established ‘safe’ practices – there always has been and always will be safety problems nobody saw coming and winning the really bad shit happens lottery despite mythical perfect safety procedures…

      Really drivers on trains hasn’t made much sense for pretty much all of my life, except for on those few routes that are supposed to be someday or have been altered to make them safer anyway. The only reason they are present on most trains is because they were already there before the technology caught up, so nobody initially wanted to spend the huge sums it would have cost upfront in the 90’s to implement the technology and make them all redundant. Might well pay for itself in a year or two even back then with the wage bill reduction, but its a very large sum of money upfront in a very underfunded and since privatization convoluted cost wise rail network.

      1. >If they really valued profit over life there would have been no pandemic lockdowns at all,

        Whose life, yours or theirs? You forget that the government, the top politicians, as well as the CEOs and owners of large corporations tend to be older people right in the risk zone. Of course they would want to limit the spread of the disease onto themselves.

        1. But all they have to do for that Dude is continue to live in the isolated little cloud land and not meet their constituent/employee for a while… Protecting themselves while ignoring potential peril to the minion classes wouldn’t have been hard, and undoubtedly works out better for the economy. Just less well for some significant fraction of the at risk groups.

          Afterall you can even sell it as patting yourselves on the back for following the guidance that says remote working is suggested where possible to protect ‘everyone’, when all they really care about in this hypothetical is themselves.

  4. Yeah, because I trust auto companies and the government to make and program software that has lives in its hands.
    Not to mention there’ll be no one to pull the bus over and call the cops when the junkies, muggers, and rapists get out of hand.

Leave a Reply to NicoCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.