Why automating trucking is harder than you think
Photo by Emil Lippe for The Washington Post via Getty ImagesAn interview with Karen Levy, an assistant professor of sociology at Cornell, about her new book, Data Driven: Truckers, Technology, and the New Workplace Surveillance Continue reading…
As a bevy of classic Hollywood movies has shown, truck driving is an occupation intertwined with American ideals of freedom and machismo.
But technology is threatening the trucker’s traditional independence. Seeking to reduce crashes, the federal government issued a mandate in 2017 that truckers use Electronic Logging Devices (ELDs) to record their driving hours, replacing pencil-and-paper logs that were easily fudged.
Are ELDs unwarranted surveillance, or are they a vital safety technology? The answer depends on whom you ask, but pretty much everyone agrees that their adoption constrains the role of human drivers in an industry employing over 3 million people in the US.
Karen Levy, an assistant professor of sociology at Cornell, explores the impacts of trucking technologies like ELDs in her new book, Data Driven: Truckers, Technology, and the New Workplace Surveillance. As Levy describes, implementation of the federal ELD mandate led truckers, employers, and regulators to alternate between collaboration and conflict as they scrambled to comply (or at least make others think that they were).
Levy emerges skeptical not just of ELDs but of autonomous driving as well. She writes:
“It is true that humans have faults — but automating them away isn’t a solution. There will always be a role for humans in (so-called) autonomous systems — in their design, operation, maintenance, use, and oversight. Those rules may look different than they did before, but the notion that humans can be eliminated from systems is fundamentally false.”
Image: Princeton University Press
I spoke with her about her book and the lessons it holds for an iconic blue-collar profession that is a linchpin of the US transportation system. Our conversation has been edited for clarity.
How did you land on truckers and technology as a research topic?
In 2011, I was a graduate student studying sociology, and I had a law degree. I’ve always been obsessed with rules, and I wanted to find a space that was moving from enforcing rules manually to enforcing them through digital power.
I heard an NPR story that explained how the federal government was considering mandating that truckers keep their time digitally after 70 years of doing it manually. I got excited.
That day, I went to a truck stop in Portland, Oregon. I said to myself, “Let me see what it feels like to talk to the truck drivers.” I was immediately hooked because people were so happy to share their life stories with me and so willing to explain things to me even though I had zero knowledge. I got hooked, and I kept doing it for what’s now a quarter of my life.
In your book, you vividly describe the lives of today’s truckers. Their mental and physical tolls are considerable, and wages have fallen significantly since deregulation 40 years ago. Why do truckers still want to do this?
When you talk to truckers about why they do the work, almost all of them say something about how they don’t want someone looking over their shoulder and that they want a job where they have some control over their time.
There is a romance about driving across the country and being in control of what you’re doing. It’s hard to find that in other blue-collar jobs. So a lot of truckers select into the industry. There’s a strong culture associated with masculinity and pride in the central role truckers play in the global economy.
When you’re driving on the highway, do you now worry about the truck driver in front of you causing a crash?
In some ways, I’m unworried because I’ve met so many hyper-competent truckers who’ve driven millions of miles without an accident. They have almost a fatherly orientation toward other drivers on the road, and they know those drivers make a lot of mistakes.
At the same time, I know that those experienced drivers are the ones most likely to be driven out of the industry due to the surveillance technology I write about in the book. As a result, there’s more of a push now to bring younger drivers into big rigs, including 18-year-olds. Those are not the drivers you want next to you on the highway.
“Those are not the drivers you want next to you on the highway.”
You used the phrase “Knights of the Highway” to describe how truckers used to be seen. How does that sense of pride and individualism affect their attitude toward technology?
Truckers aren’t Luddites. They actually know a lot about technology because they often have to fix these complicated 80–100,000-lb vehicles with very little help. Technology also relates to the nature of their work; they were big adopters of CB radio back in the day.
But they are understandably resistant to technologies that impinge on their sense of self, especially if they have accumulated knowledge about how to deal with different driving conditions or their own biophysical state, which is how they know when to stop driving. When confronted by a technological regime that suggests they’re not trustworthy — that they’re liars and cheats — that hits them pretty hard.
Which brings us to Electronic Logging Devices. Why were they introduced?
Since the 1930s, truckers have been subject to these timekeeping regulations that limited the number of hours they drive daily and weekly. But it’s a very open secret in the industry that no one has taken this requirement all that seriously and that truckers frequently falsify their paper logs. In part, that’s because the logs are disconnected to how they’re paid, which is by the mile. Sometimes truckers use chemical aids to go way beyond what their bodies are capable of, which is obviously dangerous.
The US Department of Transportation has had legitimate concerns about unsafe driving, and one of the levers regulators had at their disposal was mandating Electronic Logging Devices. That led to a 25-year discussion in Washington about whether truckers should be required to log their hours electronically, and it finally became mandatory in 2017. The idea was to make it harder for truckers to flout the rules.
In the book, I described how the requirement that truckers have ELDs acts as a scaffold that also enables supervision and surveillance by their employer, the trucking company. Because now that truckers have to have this black box in the truck, the companies say you might as well use a black box that also tracks your fuel usage, how hard you’re braking, and other such things.
It’s almost as though the federal government said everyone has to buy a phone that makes calls. But of course, you can’t easily buy a phone that only makes calls; you have to buy one that also texts, accesses the internet, takes pictures, and so on. That’s basically what happened with truckers and ELD adoption.
You write that truckers resent ELDs and that a common refrain among them is that ELDs treat them like criminals. But you also note that 5,000 people die in crashes in the US related to trucking (about 12 percent of all crash deaths). How much should we care about truckers’ feelings when so many lives are at stake?
In theory, that logic would make sense. But the reason I’m not compelled by it is that there is no evidence that ELDs actually make anyone safer. In fact, the evidence suggests that crash rates have gone up after the ELD mandate and that truckers drive less safely because they’re so tightly supervised.
“This is not a good approach if we’re most concerned with safety.”
Truckers have long depended on the flexibility of being able to keep track of their time using paper and pencil. When you remove that flexibility, that means the trucker is less likely to avoid speeding or stop for a cup of coffee if he’s tired. Rather, what it suggests is that the most important thing is to get from A to B on the road as quickly as he can. Studies show that speeding has increased since the ELD mandate and that truck-related fatalities have, too. That indicates that this is not a good approach if we’re most concerned with safety.
How widespread is the use of Driver Monitoring Systems across trucking?
Quite widespread. Once you’ve built the capacity into the truck to transmit data back to the home office, it’s not that much extra work to add another data stream. It’s common to track fine-grained driver behavior like real-time speed, as well as more invasive camera-based or biometric ways to track aspects of the trucker’s behavior or body.
I was struck by your example of a “smart cap” that vibrates. Can you tell me about that?
Yes, it does a constant EEG on the trucker’s brainwaves through tech integrated into the hat and then transmits that information back to the supervisor. It’s looking for signs of disruption or fatigue or irregularity. The goal is to protect these workers, at least according to the company’s marketing.
There’s also technology that can flash lights in the trucker’s eyes, as well as a seat cushion that vibrates the trucker back into alertness if they look like they’re falling asleep.
I can understand why truckers would resent those devices, and I wonder what implications it carries for driver monitoring systems installed on automobiles. So let’s fast forward to a world where DMS is widespread across cars. If you’re tired or tipsy, your car can tell — and it will first send warnings and then refuse to let you operate the vehicle. Do you think the general public will accept the safety benefits of DMS? Or are they going to be livid?
If that comes to pass, I expect we will see moments of resistance, as we do with any technology that is oriented toward controlling people. That’s especially true if it’s controlling people in ways that they weren’t controlled before. We would expect to see more resistance, as well as more companies capitalizing on the pushback to sell aftermarket devices that thwart those DMS capabilities.
Of course, such devices could greatly degrade the efficacy of DMS systems. In the 1970s, there was a powerful popular backlash to safety belt interlocks, which forced Congress to quickly retreat. Do you see parallels with DMS?
“If people see the legitimacy of the rules mandating the technology, then maybe they accept it.”
I think a lot of it comes down to legitimacy. If people see the legitimacy of the rules mandating the technology, then maybe they accept it. But if it’s poorly implemented or if the technology doesn’t always work, it’s very easy for the mandate to lose public legitimacy. Then all bets are off.
Let’s get back to trucking. I’ve heard autonomous technology boosters claim that trucking is a better use case for AV tech than robotaxis because the environment is a lot simpler. Outside the first and last few miles of trips, highway driving seems like a rather predictable kind of activity (especially compared to driving in a city). Is that impression fair, or is it a misrepresentation of the trucker’s role?
I think it’s a gross misunderstanding of what truckers do. They’re doing all kinds of stuff. They’re required to visually inspect their trucks for safety several times per day, looking for a frayed tie that holds something onto a flatbed or examining a tire that kind of seems funky. If they have freight that is refrigerated, they’re also making sure the refrigeration system is working. And they’re ensuring stuff doesn’t get stolen.
These are things that you can’t easily automate, but they are crucial when you’re piloting an 80,000-lb vehicle down the road all day.
Let me challenge that. You’re talking about safety checks, refrigeration, and security. Is it really so hard to automate those activities?
I think you’re right that some of those functions could be remotely automated, eventually. But the timescale is much longer than what autonomous vehicle advocates are shooting for. We don’t yet have a way to conduct visual inspections remotely. We don’t yet have a remote security solution that’s comparable to having a human being protecting very valuable freight.
I think it’s not impossible to think that within 40 years, we might have autonomous trucking, but it will take a while.
Has the pandemic affected attitudes toward trucking and technology?
The pandemic has been this interesting case study regarding the efficacy of timekeeping regulations. A bunch of regulations got lifted for obvious reasons since people didn’t have things like toilet paper. So truckers didn’t have to comply with rules as they normally would.
“I think it’s not impossible to think that within 40 years, we might have autonomous trucking, but it will take a while.”
Accident rates didn’t go up; everything was actually pretty much fine. So now some in the industry say to regulators, “You claim we have to have all these rigid rules, and we have to be under your eye in a very precise way, but in an emergency, we can actually do just fine. So what makes you think we can’t do without the regulations all the time?”
It’ll be interesting to see how that bubbles up in Department of Transportation policymaking in the next couple years.
If you could wave a magic wand and change any federal trucking policy, what would it be?
I would remove the exemption of truckers from the Fair Labor Standards Act so they can get overtime pay.
It would be a game-changer in the industry since compensation is now typically paid per mile driven. It’s not going to solve all the problems, but it would recognize that a lot of work in trucking is wholly uncompensated, like when truckers are caught in traffic or bad weather or when they are delayed unloading their cargo. Truckers would be shown that their work counts for something — not nothing, as much of it does right now.
It sounds like you’re pretty pessimistic that technology in the trucking industry can improve safety.
In isolation, yes. I think technology can be part of a suite of solutions that will actually protect people. But beyond economic reform in this case, I just don’t think that tech on its own will accomplish anything.