Sunday, February 4, 2018

Truck automation

Two bits from Marginal Revolution on truck automation are so good they merit passing on here. Dan Hanson writes this amazing comment
I wonder how many of the people making predictions about the future of truck drivers have ever ridden with one to see what they do?
One of the big failings of high-level analyses of future trends is that in general they either ignore or seriously underestimate the complexity of the job at a detailed level. Lots of jobs look simple or rote from a think tank or government office, but turn out to be quite complex when you dive into the details.
For example, truck drivers don’t just drive trucks. They also secure loads, including determining what to load first and last and how to tie it all down securely. They act as agents for the trunking company. They verify that what they are picking up is what is on the manifest. They are the early warning system for vehicle maintenance. They deal with the government and others at weighing stations. When sleeping in the cab, they act as security for the load. If the vehicle breaks down, they set up road flares and contact authorities. If the vehicle doesn’t handle correctly, the driver has to stop and analyze what’s wrong – blown tire, shifting load, whatever.
In addition, many truckers are sole proprietors who own their own trucks. This means they also do all the bookwork, preventative maintenance, taxes, etc. These people have local knowledge that is not easily transferable. They know the quirks of the routes, they have relationships with customers, they learn how best to navigate through certain areas, they understand how to optimize by splitting loads or arranging for return loads at their destination, etc. They also learn which customers pay promptly, which ones provide their loads in a way that’s easy to get on the truck, which ones generally have their paperwork in order, etc. Loading docks are not all equal. Some are very ad-hoc and require serious judgement to be able to manoever large trucks around them. Never underestimate the importance of local knowledge.
... a fundamentally Hayekian insight: When it comes to large scale activities, nothing about change is easy, and top-down change generally fails...
I would add silicon valley software companies, and media commentators to the think tanks and government offices, on the list of pundits that tend to denigrate the skill, knowledge and intelligence required of what are very wrongly call "low skilled" jobs. I note several times the "paperwork" required in trucking as well.

The comment was on an earlier MR post covering Alexis Madrigal in the Atlantic on self-driving trucks. Automation in any industry reduces costs and increases quality. These mean the industry expands, so labor demand may even grow. Think of the computer you're reading this on. Labor gets to specialize at the things people are good at, which is higher productivity, and wages rise.
Uber does not believe that self-driving trucks will be doing “dock to dock” runs for a very long time. They see a future in which self-driving trucks drive highway miles between what they call transfer hubs, where human drivers will take over for the last miles through complex urban and industrial terrain.
And fill out that paperwork.
if the self-driving trucks are used far more efficiently, it would drive down the cost of freight, which would stimulate demand, leading to more business. And, if more freight is out on the roads, and humans are required to run it around local areas, then there will be a greater, not lesser, need for truck drivers. 
The article misses the second, more important effect. As low-cost trucking expands, other businesses that use trucks expand, and they hire people too.

And I just googled "truck driver shortage" to get the latest media story to rediscover this fact. From NPR "Trucking Industry Struggles With Growing Driver Shortage",
The trucking industry is facing a growing shortage of drivers that is pushing some retailers to delay nonessential shipments or pay high prices to get their goods delivered on time.
A report from the American Trucking Associations says more than 70 percent of goods consumed in the U.S. are moved by truck, but the industry needs to hire almost 900,000 more drivers to meet rising demand.
It's a tough job.  Young people aren't going in to it. There are competing opportunities. My last Uber driver just quit long-haul truck driving. He earns a bit less, but gets to see his family every night.

(Of course we don't use the word "shortage" in economics unless there is a government-imposed price control. This just means wages will go up.)

Update: Thanks to the comment from Unknown below, here is the response from Tom T in MRs comments
This comment reminds me of the guy in Office Space who ends up helplessly screaming at the downsizing consultants, “I have people skills!”
It’s trying way too hard. Sure, the truckers do all of these things. But is there any reason to think that the self-driving AI won’t be just as good, if not far better, at confirming that the load is secure, verifying the manifest, monitoring vehicle maintenance, and interfacing with the weigh station? Does anyone really think a computer can’t split loads and optimize returns? Even accepting the dubious premise that the poor driver is somehow providing effective security by sleeping in the cab in front of the load, the whole need for that sleep-time security disappears when the vehicle is self-driving and doesn’t have to sleep. Truckers who are self-employed and have to do bookwork, taxes, and customer relationships will have a lot more time to devote to that end of the operations if they don’t also have to drive the truck. Loading docks that are ad hoc will find themselves either standardizing or paying extra for failing to do so, like every other aspect of an industrial, computerized world.
I’ve been a litigator for 25 years. At one time, we saw our profession as so paper-based and quirky and fact-dependent that it would always be safe from automation. Sure enough, OCR text recognition, keyword searches, and predictive algorithms completely changed the economics of document review. No one is immune from automation.
I heard somewhere, I forget where, the quip that the ones who really should fear automation are accountants, book keepers, and other fairly routine paperwork office workers. It sounds awfully sensible. Of course our government is pretty good at expanding the demand for paperwork!


  1. Theres a typo in the name at the start - Its Dan Hanson.

    Also see this comment on MR as a good response:

  2. The comment cites Hayek's observation that top-down change is frequently a fatal conceit. But there is another Hayekian observation that is also appropriate. Which is that entrepreneurial knowledge -- which is certainly what the trucking owner-operators have -- is highly situational:

    "These people have local knowledge that is not easily transferable. They know the quirks of the routes, they have relationships with customers, they learn how best to navigate through certain areas, they understand how to optimize by splitting loads or arranging for return loads at their destination, etc. They also learn which customers pay promptly, which ones provide their loads in a way that’s easy to get on the truck, which ones generally have their paperwork in order, etc. Loading docks are not all equal."

  3. "Of course we don't use the word "shortage" in economics unless there is a government-imposed price control. This just means wages will go up."

    That's the theory, but it doesn't always work in practice since rising wages imply rising prices, which the market won't always bear. Back around 2000 I know of a couple restaurants which closed because they couldn't get staff and couldn't raise prices (and thus wages) without losing business.

    There's also the question I remember from an old economics textbook of what happens if the supply of workers won't expand to meet demand regardless of how high wages go up.

    Any change toward self-driving trucks will have to deal with all of the non-driving issues involved and you are right that many non-driving tasks must be taken into account along with practical measures which require mechanical actions. Will a self driving truck have the ability to put chains on when going into the mountains in snow (many Silicon Valley types probably don't know what car/truck chains are let alone when and how they are used).

    Still, self driving trucks will probably become reality. It may mean a move toward larger trucking firms with a "load specialist" who manages all the non-driving activities at the source and another at the destination. This could actually mean better jobs -- the "specialist" would not have to be away from home most of the time as truckers do today.

  4. Not an economist here, but love learning the discipline.
    I have a brother and sister both truck drivers and will add a few human interest notes.
    My brother hauls building materials, custom glass and tile, from central Wisconsin to the center of large city building projects in NYC, San Francisco, Portland, etc. He can maneuver an 18 wheeled flatbed into Times Square. He can angle by inches in and out of narrow city alleys. He has guided a tractor with a literal blown engine block safely to the edge of an interstate, without impact.
    My sister states her company has a designated driver to retrieve abandoned tractors. Enough drivers give up and walk away from their tasks that her company assigns a driver to retrieve them. Three reasons for the high turnover. 1) social isolation, esp. difficult for the natural extrovert, 2) lack of fitness opportunities. The lifestyle leads to weight gain and stiffness. 3) Not being able to sleep in ones own bed.
    The hub model with skilled drivers on the last 50 miles would be a huge lifestyle improvement. This would draw in more drivers.

  5. I used to drive an inner-city route for an express air carrier from the hub to the airport. Some thoughts: Any automated trucking will only work on an interstate or similar route. The variables of traffic, accidents, pulling over for first-respondents; not to mention the expense of dedicated lanes or expanded right-of-ways and land acquisition cost necessary to accommodate it.

    Paperwork can be automated and truck brokers already use computer software to consolidate and re-route loads. The proof will be if automation can take on the nuance and multi-tasking that we all use to drive a car and add to that 21 gears and a box 53' long.

    So I'm sure there will be some experimenting, and the insurance industry and lawyers will be waiting at the loading dock; so the reality is very...very, far off.

    By the way there exist a system that's partially automated, has two drivers and can carry maybe up to 50 trailers (maybe more) and it runs on rails- that'll do for now.

    1. On your last comment. It is interesting that we don't have self-driving trains, particularly at Amtrak given recent crashes. This seems orders of magnitude easier than self-driving cars. Maybe just the market is too small to make it worthwhile

    2. It's probably an insurance or union issue.

    3. Copenhagen metro has no drivers. Works perfectly.

    4. The Copenhagen metro is indeed very nice. While riding it recently, I wondered if any of our US cities, in building a subway, could bring themselves to build a driverless system, or whether they would feel the pull to "provide jobs." Many in the US left admire Denmark from afar as a model, but I have also noticed a certain ruthless efficiency there that doesn't quite match the US image.

    5. I would not ride on unmanned mass transit in a major American city. Not without a concealed carry permit.

    6. You don't need to cross the Atlantic to see a driverless metro. You can find one 15 miles north of the US border - in Vancouver. And it works great.

  6. We have had driverless movement of semi trailers for 30 years. Put them on a flatbed railroad car. It has taken a certain percentage of long distance trips. You can see long trains piled with containers at most railroad crossings. But, It hasn't stopped there from being lots of tractor pulled containers too.

    I too am inclined to think that AI/robotics proponents have overclaimed. But they have been doing that for years. I grew up reading Issac Asimov's novels about robots. We are nowhere near realizing his vision of a true multi-purpose androform robot.

    BTW Asimov's robots had brains made out of "platinum-iridium sponge" At $1,000/oz. There would not have been enough of them to make a difference. Our real robots use silicon, which is wildly abundant.

    1. A friend in the railroad business advises, don't put anything on a rail car you don't want shaken to bits.

  7. It maybe, but that is just a question of how well you wrap the goods. I knew a man who shipped eggs from Ohio to Hong Kong in containers that were sent by rail and ocean.

    The point is that there are many ways of moving freight containers other than having a diesel powered tractor operated by a human driver pull them over the highway. The driverless tractor, if such a thing is ever commercialized, is not that big a change in the world.

  8. Fun post. Yes, we have labor "shortages" in the US. In all 12 districts according to the latest Beige Book. The price signal is ineffective in the US...or so one would believe.

    1. Talk to employers who need what you might call non-professional labor. Their biggest problem is finding people who pass their drug screens.

  9. I think the point is not whether "the self-driving AI" will be "just as good, if not far better, at confirming that the load is secure, verifying the manifest, monitoring vehicle maintenance, and interfacing with the weigh station". Sure, maybe in 100 years humanoid robots will be able to do all that, which implies that they will also be able to do almost all else and labor will no longer be a scarce resource. Who knows? But that's exactly the point: the impact of any innovation is unpredictable, so one should tread lightly when making forecasts.

    As well, while reading this interesting post, I couldn't help smiling bitterly, thinking that most economists who are well-known for their work on technological change and whom I greatly respect, people like P. Romer, P. Howitt, P. Aghion, Sam Kortum and others, fall exactly in the category of scholars who largely engaged in deductive reasoning from their offices. Economic historians like N. Rosenberg, J. Mokyr, Chris Freeman, K. Pavit, and others who offer detailed analyses of specific technological innovations throughout history are largely ignored. Maybe it is because the conclusion drawn, that the impact of each major innovation has been different and not easily predictable, cannot be captured as easily by neoclassical models relative to evolutionary models (e.g. by Nelson and Winters) or complexity models (e.g. by Brian Arthur). In any case, though I consider myself a neoclassical economist, I think our perspective, at least on the issue of technological change, would improve from greater methodological pluralism.

  10. OT but way fun:

    "Arizona Senate Votes to Accept Tax Payments in Bitcoin" | Fortune

    Egads. Theoretically. this huge. Money has value if governments accept it for payment of taxes. One could even see some sort of exchange emerge where people sell bitcoins to Arizona taxpayers.

    I cannot imagine sovereign nations and central banks will let this stand.

    But, if other states and nations start accepting bit coins….is this an expansion of the money supply? I think so….

    1. Actually not such a big deal, as long as it's not a fixed number of bitcoins. If you owe $100, but can pay in bitcoin, euros, or pesos, at the current exchange rate, that makes little difference.

    2. John--

      Thanks for your reply--but while the supply of bitcoins near to fixed (I think bitcoin supply is planned to operate like an asymptote) the value of bitcoins is obviously not fixed.

      So, if bitcoins become worth $1 million each, then I think the money supply has expanded, as long as government accept bitcoins as payment for taxes.

      As in: I buy a bitcoin for $100k. Later it becomes worth $1 mil. I pay off my long overdue taxes with it, and still have plenty of money for consumption. No one has given up consumption so I could do this---no one in the private sector paid me $1 million in US cash for my bitcoin.

      The government now has $1 million in US cash spending power at its disposal.

      Though the value of bitcoins is set by a few marginal transactions, the value of bitcoins in total can rise dramatically.

      In this case, it appears that money is being created out of thin air.

      I think.

  11. Add on: I think a key is whether, say, Arizona, then buys services, spending bitcoins. So I have my $1 million bitcoin, and I send it in to AZ to my off long overdue taxes. I bought the bit coin at $100k.

    AZ uses my bitcoin to buy $1 mln worth of highway paving. The highway paver pays vendors in bitcoin and on down the line. Everyone is willing to accept bitcoin, as they have value in paying taxes.

    So, you get a second bitcoin economy operating on top of the US dollar economy, but both currencies making claims on real output.

    I think….

  12. I think you could make the case that tailors provide many services that mass produced clothes cannot match. I am sure they can likely produce a better suit than anything I can pick up at a local department store. They are still a luxury product.

    Once the automation technology exists, there will be a relentless pressure to standardize. We have seen already the effect of containers for ships. Why wouldn't we see the same effect for trucks? Most of the tasks described by the comment seem to be related to poor standardization. Designing a good standardization system is hard and there is little point to do so if there is a human standing around anyway.

  13. I think all jobs except economic professor seem very complex and indispensable :-)

  14. The topic was really an interesting one. Definitely a good read.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.