Yesterday, Google launched Chromecast, a streaming solution for integrating mobile devices with TV, part of another salvo against Apple. Google vs. Apple has been the hot story now in Silicon Valley for a couple of years. Before that, Google vs. Facebook. Before that, Google vs. Microsoft. Technology loves narrative, and setting up a battle of titans always gets the crowd worked up.
Lately, I’ve been thinking about the next fight Google might be inadvertently setting up, and wondering whether they are ready for it.
Self-Driving Cars or Self-Driving Trucks
It turns out I’m not the only one who noticed that Google’s incredible push for self-driving cars actually has more likely applications around trucking. Yesterday, the Wall Street Journal wrote an excellent piece about Catepillar’s experiments using self-driving mining trucks in remote areas of Australia. It had the provocative headline:
Daddy, What Was a Truck Driver?
This is the first piece in the mainstream media that I’ve seen connecting the dots from self-driving cars to trucking, even with a lightweight reference to the Teamsters at the end.
Ubiquitous, autonomous trucks are “close to inevitable,” says Ted Scott, director of engineering and safety policy for the American Trucking Associations. “We are going to have a driverless truck because there will be money in it,” adds James Barrett, president of 105-rig Road Scholar Transport Inc. in Scranton, Pa.
The International Brotherhood of Teamsters haven’t noticed yet, or at least, all searches I performed on their site for keywords like “self driving”, “computer driving”, “automated driving”, or even just “Google” revealed nothing relevant about the topic. But they will.
Massive Economic Value
The statistics are astonishing. A few key insights:
- Approximately 5.7 million Americans are licensed as professional drivers, driving everything from delivery vans to tractor-trailers.
- Roughly speaking, a full-time driver with benefits will cost $65,000 to $100,000 or more a year.
- In 2011, the U.S. trucking industry hauled 67 percent of the total volume of freight transported in the United States. More than 26 million trucks of all classes, including 2.4 million typical Class 8 trucks operated by more than 1.2 million interstate motor carriers. (via American Trucking Association)
- Currently, there is a shortage of qualified drivers. Estimated at 20,000+ now, growing to over 100,000 in the next few years. (via American Trucking Association)
Let’s see. We have a staffing problem around an already fairly expensive role that is the backbone of a majority of freight transport in the United States. That’s just about all the right ingredients for experimentation, development and eventual mass deployment of self-driving trucks.
Rise of the Machines
In 2011, Andy McAfee & Erik Brynjolfsson published the book “Race Against the Machine“, where they describe both the evidence and projection of how computers & artificial intelligence will rapidly displace roles and work previously assumed to be best done by humans. (Andy’s excellent TED 2013 talk is now online.)
The fact is, self-driving long haul trucking addresses a lot of the issues with using human drivers. Computers don’t need to sleep. That alone might double their productivity. They can remotely be audited and controlled in emergency situations. They are predictable, and can execute high efficiency coordination (like road trains). They will no doubt be more fuel efficient, and will likely end up having better safety records than human drivers.
Please don’t get me wrong – I am positive there will be a large number of situations where human drivers will be advantageous. But it will certainly no longer be 100%, and the situations where self-driving trucks make sense will only expand with time.
Google & Unions
Google has made self-driving cars one of the hallmarks of their new brand, thinking about long term problems and futuristic technology. This, unfortunately, is one of the risks that goes with brand association around a technology that may be massively disruptive both socially & politically.
Like most technology companies in Silicon Valley, Google is not a union shop. It has advocated in the past on issues like education reform. It wouldn’t be hard, politically, to paint Google as either ambivalent or even hostile to organized labor.
Challenges of the Next Decade
The next ten years are likely to look very different for technology than the past ten. We’re going to start to see large number of jobs previously thought to be safe from computerization be displaced. It’s at best naive to think that these developments won’t end up politically charged.
Large companies, in particular, are vulnerable to political action, as they are large targets. Amazon actually may have been the first consumer tech company to stumble onto this issue, with the outcry around the loss of the independent bookstore. (Interesting, Netflix did not invoke the same reaction to the loss of the video rental store.) Google, however, has touched an issue that affects millions of jobs, and one that historically has been aggressively organized both socially & politically. The Teamsters alone have 1.3 million members (as of 2011).
Silicon Valley was late to lobbying and political influence, but this goes beyond influence. We’re now getting to a level of social impact where companies need to proactively envision and advocate for the future that they are creating. Google may think they are safe by focusing on the most unlikely first implementation of their vision (self-driving cars), but it is very likely they’ll be associated with the concept of self-driving vehicles.
I’m a huge fan of Google, so maybe I’m just worried we may see a future of news broadcasts with people taking bats to self-driving cars in the Google parking lot. And I don’t think anyone is ready for that.
Psychohistory 
“We have a staffing problem around an already fairly expensive role that is the backbone of a majority of freight transport in the United States. That’s just about all the right ingredients for experimentation, development and eventual mass deployment of self-driving trucks.”
Isnt this equally true for the airline industry? We have been having the capability to fly places on autopilot mode. But has that reduced the need for pilots? Has it reduced the pilot salaries, which increases proportionately with their “experience”, even though technology has made the experience gap redundant? Isnt expensive pilot and associated problems from pilot unions been the single most factor for no airline making any money?
So how is truck industry different from airline industry that it will allow driverless trucks to replace drivers? Probably, companies might invest money to buy long range trucks that have driverless capabilities, but I don’t believe drivers can be replaced in the near term.
This is not true about airplanes, for a few obvious reasons. Freight is split across rail, truck and air (for the most part), and the share to air is relatively small. In some ways, you can look at this as proportionality: # of drivers required per ton of freight / mile / time. Trucking is more than 2/3 of all freight in the US, and is relatively labor intensive compared to the other options. The capital costs and flexibility of trucking, however, continues to dominate.
This post, BTW, is about the long term (10+ years) not the short term (3-5 years).
It’s already happening in a big way with aircraft. This is now! http://www.mcclatchydc.com/2013/05/06/190332/drone-cargo-helicopters-prove.html#.UfQpExb9lgI
Awesome article. I never thought about the value of self driving cars in respect to trucking before but it is brilliant.
In the last 30 years China replaced many industrial jobs in America with their high skilled low wage workers and provoked an acceleration to the knowledge economy. Technology will dis-intermediate workers again maybe pushing us all to more fragmented, entrepreneurial activities.
Hey Adam. Finally read the post. Definitely interesting. A few things, as I used to work at the Port of Boston and helped manage some of the container operations to the railhead and trucking infrastructure. Assuming truck drivers get displaced, there are many local jobs at ports that need to be filled, including security, that could actually make one’s life easier — less travel. Generally, I see a lot of other infrastructure needs (and therefore jobs) sprouting up as these kind of technologies advance. (Unrelated, here’s the post I wrote generally about government policies: http://blog.semilshah.com/2013/07/26/man-vs-the-government-machine/)
I’ve been thinking about this too. I imagine a lot of interstate/long-distance driving being automated (indeed, I can easily imagine a future where most inter-city driving provides preferable treatment to automated vehicles – perhaps initially in dedicated lanes like HOV, but later, entire roadway systems) – but the so-called “last mile” might face a lot more social & political resistance (the legal liability of a city-street crash involving an automated vehicle would be a very interesting case).
So – yeah, I imagine that a lot of the long-distance driving will not have a driver. But there might still need to be a driver on-board to take over within city limits (or when not on highways), or perhaps suburban truck-stops will take on new roles where local drivers come out to meet the trucks and manually drive them into the cities/towns/etc. It may well shake out, for socio-political, if not technological reasons, that much like with airline pilots, we need a human presence, even if they’re there mostly as a backup/supervisor, not directly controlling.
So would truck-driving change? yeah – a lot. But I do suspect it’ll take years, if not decades longer for urban routes to be automated compare to when inter-urban routes are automated.
Hi Adam. While it is true that there will be longterm economic disruption of driving as a profession, this project is different from most other projects where people talk about “robots taking jobs.” For decades now, people have talked about this, because the companies building and buying industrial robots have in a majority of cases been expressly building them for the purpose of replacing paid human labour.
Not so here. If you were to go up to any of the folks building self-driving cars and bring this up they would find it strange. Replacing paid labour isn’t even on their minds as a goal, not for most of them. Saving lives is usually the first goal. Then replacing not the 7 billion hours of paid driver labour but the 50 billion hours of unpaid, non-productive time that “amateur” drivers spend driving themselves around. Then the consequences for energy and disruption of the automotive industry, both industries that dwarf the driving “industry.”
So while not denying that there is an issue and consequences for those who make their living driving, the reality is that this is comparatively tiny compared to all the other big effects. Nobody wants to think their job is “in the noise” so it must be planned for, but the reality is rather striking.
While I appreciate the sentiment, the WSJ article clearly cites a number of commercial efforts directly in this direction. The lead time for replacing the non-commercial driver is so much longer, and has a much less obvious commercial benefit, it’s hard to imagine that long haul trucking, particularly long routes with non-hazardous cargo, wouldn’t be the cheapest and fastest high ROI application.
It’s also worth noting that having technologists tell out of work truck drivers that their intention was to focus on non-productive amateur drivers is exactly the type of social/political naivete that I argue is dangerous.
Well it is, for the most part, the truth. However, there is another reason there is more focus on passenger cars than on trucks. The vehicles are all prototypes, and work is all very safety-oriented. There will be accidents, especially in the early days, and it is highly preferred that these accidents be with small cars and not big rigs.
But seriously, while there are people looking at trucks its mostly in the constrained environment space like the Rio Tinto mine, not the open highway. There have been convoy projects for the open highway (though they ran into problems) but these don’t eliminate drivers, in fact the Volvo/Sartre one always had a lead truck driver.
I’m not pretending that there are not big changes coming in truck driving, nor would I advise people to get into it as a long-term career if they are young, but when it comes to the open road, the cars are coming first, the trucks later. And we still have a few years to wait for the cars.
But what’s the better ROI? Replacing a truck driver who is paid $30-$40 per hour, or letting a $500/hour lawyer be productive on the way to work?
Welcome to the present. Of course driverless cars/trucks will displace cab/livery and freight drivers. Their job requires a warm body and ~20/20 vision.
It is the field hand of the 19th/20th century. Once traffic accidents are reduced 90% then the major cause of accidents and fatalities (human error) will be eliminated.
Teamsters can kick and scream, but they are aging out. Older drivers will not be replaced. Some freight will require human supervision, but driving will be a thing of the past.
This is a VERY GOOD THING.
You project that Google will be painted as “anti-union” since they aren’t a union shop? Ridiculous.
The unions have no power. Truck drivers are already experiencing reduced wages, higher fuel prices and lower reimbursements. Go talk to ANY truck driver. It’s a $hit job. Unless you like toothless hookers at truck stops.
Pingback: Winners and losers from self-driving cars | Den grønne Have
Cool article, the thought randomly occurred to me too and this was the first thing I clicked on in google. If I had money to invest, I’d bet long on google, they really stand to make a bajillion dollars off of this tech if they can dominate the market for automated trucking.
And the social impact is pretty scary too… I feel like it would be the equivalent to the cotton gin, no one technology since then would so quickly and devastatingly take careers out from underneath people, assuming that there would be no political interference. But… our government is based on the same principals as american idol, so I guess thatd be unlikely. Cool scary/interesting article tho.
Pingback: Pay Scale & Driver Shortage ?
Pingback: 2017: New Year, New Template, More Posts | Psychohistory