Interstate highway lanes for autonomous trucks could move freight more efficiently, UI study says

This story first appeared in The Long Haul 2025, an annual special section that celebrates National Truck Driver Appreciation Week by looking at Eastern Iowa’s trucking industry.

Under current federal law, truck drivers may drive for no more than 11 hours before taking at least 10 hours’ rest. But what if the truck itself did a good share of the driving?

“We were thinking, ‘Should that be different for autonomous vehicles?’” said Qi Luo, assistant professor of business analytics at the University of Iowa’s Tippie School of Business. “(Drivers) are free 95 percent of the time, so their fatigue level won’t be comparable.”

Luo tested that idea against data on present-day truck traffic between Chicago and the West Coast. Preliminary findings from his study suggest designating interstate highway lanes for semi-automated long-haul trucks during overnight hours could move freight more efficiently.

“That’s the safest and easiest scenario that came to mind,” he said. “It’s a controlled environment. It does require certain changes in infrastructure.”

It’s likely years before you encounter a semi-truck along Interstate 80 with no one behind the wheel during the wee hours. But in May, Pennsylvania-based trucking firm Aurora began operating driverless “Aurora Driver” trucks between Dallas and Houston.

“Aurora is a really well-known company in the industry,” said Omar Ahmad, deputy director of the University of Iowa’s Driving Safety Research Institute, which conducts research into autonomous vehicles. “(Luo) has an idea that’s worth consideration.”

The company put human drivers behind the wheel after a few weeks at the request of the trucks’ manufacturer, but “this observer will not operate the vehicle — the Aurora Driver will continue to be fully responsible for all driving tasks,” according to company CEO Chris Urmson.

“The technology has a lot of potential, both for improving safety and enhancing the business of trucking,” said Ahmad. “Many companies have been working on this, and there’s been a lot of predictions about how soon we’ll have fully automated vehicles and trucks. Up to now, pretty much all of those predictions have not come to fruition in its full implementation.”

With the industry short around 80,000 drivers — a number the American Trucking Association expects to double in coming years from retirements — some form of driver assist, even short of full-self-driving, could improve efficiency. Luo’s study suggests designating rural interstate lanes for semi-autonomous trucks during the overnight hours, much as high-occupancy vehicle (HOV) lanes are designated during morning and evening rush hours in metro areas.

Designated lanes would allow trucks to go faster and form convoys, improving fuel mileage.

“It really depends on how the government wants to operate,” said Luo. “Separate, dedicated lanes would be an easy infrastructure upgrade, and they’re temporary, only opened at night, so it would be easy to implement without much cost.”

“Interstates are our most predictable roadways,” said Ahmad. “Everyone is traveling in the same direction, and you can only get on and off at well-marked locations. Everyone is more or less traveling the same speed. With any automation, the question is how predictable is the environment? The more unpredictable, the more challenges it’s going to have.”

Qi Luo, assistant professor of business analytics at the University of Iowa’s Tippie School of Business (University of Iowa)

  Buy Photo

Omar Ahmad, deputy director of the University of Iowa’s Driving Safety Research Institute (University of Iowa)

  Buy Photo

prev

next

Self-driving systems are rated on a six-point scale, from Level 0 (traditional manual control) to Level 5 (fully automated; vehicles won’t even have steering wheels, accelerators or brake pedals). Aurora’s semis are Level 3, conditional driving automation, with “environmental detection” capabilities that allow the vehicle to make decisions such as accelerating past a slower vehicle.

“You still need an operator, but the trucks have braking manually,” Luo said. “The driver still needs to be alert, like the Tesla autopilot.”

Aurora’s semis are equipped with more than two dozen laser, radar and camera sensors “that can see over four football fields away,” according to the company.

“Some technologies are very good at looking forward a long way,” said Ahmad, who’s supervised the construction of a semi-autonomous bus at the UI center. “You want to know about how far away it is, but you may want to know more about it: how fast is it moving, what direction is it moving in? Some technologies are really good at looking far away but not as good as determining what it is.”

Aurora’s semis may drive short distances on secondary roads, but operation off multi-lane, limited-access highways would be restricted.

“A sensor package that’s designed for interstates may not work well in rural areas,” Ahmad said. “You also need to look off to the sides. A sensor package that can pick up traffic has to be set up to look really far ahead in the left or the right direction. That kind of package might fall apart in the rural areas.”

Cost may be an issue for manufacturers and the truckers who buy their rigs.

“Some of those sensors have been pretty expensive,” Ahmad said. “You never want just one kind of sensor, you want redundancy. A $100 part is a big deal to manufacturers, it’s such a competitive space.”

The motoring public’s comfort at sharing the highway with a driverless semi is likely to be a factor.

“Public acceptance needs a lot of validation of these types of technologies,” Luo said. “With trucks, it’s a little scary to be driving next to them.”

“More than 30 states have laws regarding (autonomous) cars,” Luo said. “On trucks it’s unclear. Trucking conditions vary, and it’s more difficult to accumulate training data. Certain fleets owned by large companies driving between their warehouses will keep collecting data. That’s how I envision this field will take off.”