On 30 August, a startup plans to add its “smart pavement” to an intersection in an industrial corner of Denver. The company has encased assorted electronics within four slabs of concrete and will wedge those slabs into the road between a Pepsi Co. bottling plant and two parking lots.
Integrated Roadways says its product, which can deduce the speed, weight, and direction of a vehicle from the basket of sensors buried in the pavement, will face its first real-world test at that discreet Denver junction.
The company can then use that data to alert authorities to accidents or prompt officials to reconfigure lanes to relieve congestion. It’s one approach to so-called “smart roads,” which aim to combine sensing and intelligence in ways that reduce the hazards and hassles of vehicular travel.
If all goes well in that first test, Integrated Roadways will replace 500 meters of pavement along a dangerous curve in Highway 285, just south of Denver, with its product in early 2019. The goal is for the pavement to detect when a driver careens off the road’s edge—the kind of accident that happens dozens of times every year on Colorado’s mountainous highways.
Amy Ford, director of communications for the Colorado Department of Transportation (CDOT), says that section of highway cannot easily be widened and is too narrow to support the addition of a guardrail. When accidents do occur there, it’s crucial to alert emergency responders as quickly as possible.
“I think that’s kind of the Holy Grail [for smart roads]: to figure out how you can reduce the time to discovery or recognition of an incident, whether that incident is a run-off-road [collision], or an accident, or a sofa that’s fallen into the street,” says Brendan Morris, an associate professor in computer and electrical engineering at the University of Nevada, Las Vegas (UNLV) who has worked on other smart road initiatives.
CDOT has given Integrated Roadways a total of US $2.75 million for both projects. Elsewhere, smart roads monitor traffic, direct drivers to vacant parking spaces, automatically issue speeding tickets, or warn of icy roads ahead. Still others measure air quality or listen for gunshots. Almost all of these systems feature some mix of sensors, processing, and fiber optic cables or wireless networks.
To build its version, Integrated Roadways starts with a piece of concrete that’s manufactured as a slab, rather than poured on-site. To each slab, it adds a three-axis accelerometer that measures vibrations to predict a vehicle’s path of arrival. A sensing fiber optic cable detects strain in the pavement by measuring subtle changes in the way light travels through the cable. And a magnetometer gauges the width of a car’s axle to help the system figure out what type of car is driving overhead.
Together with two central processing units, these sensors determine the position, size, speed, and trajectory of vehicles in real time. Separately, a gyroscope records the position of the concrete slab itself to determine whether the road has shifted out of place.
These sensors send their data through a power-over-Ethernet connection (in which both electricity and data travel through the same cable) to control centers every 800 meters (0.5 miles) along the road. Each control center serves as a mini–data center stuffed with server racks and a wireless base station. The whole system is powered by the grid.
Leif Sjögren, research director for infrastructure management at VTI, Sweden’s National Road and Transport Research Institute, says the product sounds similar to an intelligent roads concept he developed with funding from the European Union. He’s been surprised by the slow pace of progress since his project ended because sensors have gotten smaller and more efficient, and cloud technology has rapidly advanced.
Many U.S. cities have installed cameras to watch intersections and switch lights from green to red, or to help officials decide when to reconfigure lanes to relieve congestion. But Morris of UNLV says one reason smart-road technologies haven’t been more widely adopted is that transportation departments often lack the funding to make a large investment, and don’t want to bet on a technology that hasn’t already been adopted by others.
“We’ve improved tremendously but turning those [advances] into products and getting the huge contracts required to deal with those things on a large scale is a little bit difficult,” Morris says. “And I think it’s hard for cities to understand how to integrate them into the whole traffic system.”
Integrated Roadways, which patented its technology in 2016, believes that in addition to making roads safer, smart pavement will ultimately help roads pay for themselves. This will reverse the funding “death spiral” that CEO Tim Sylvester says public agencies routinely face.
His plan is to install the pavement and then charge other companies fees to add a flurry of sensors and gear to the control centers along the side of the road and to three extension ports buried in each slab.
Sylvester envisions four primary customers: telecommunications companies eager to put up 5G antennas along roadways, insurance companies that wish to install instruments to collect collision data, Internet service providers that want to set up Wi-Fi access points for customers, and real estate developers who want to put up cameras to watch traffic patterns near a new project.
Within 15 years, Sylvester projects Integrated Roadways, which is based in Kansas City, Mo., will generate $1 million per lane per mile per year from these installations. At that point, a new installation would pay for itself within 5 years.
John Harvey, who directs the pavement research center at University of California, Davis, isn’t convinced that this business model will pan out.
He says the information that could be gathered at control centers isn’t particularly useful to companies or agencies, except on occasion, when they need to update their design assumptions. Furthermore, he says, there are cheaper ways to gather that data.
For now, the product that Integrated Roadways will install in Colorado is twice as expensive as the concrete used to build roads today—but Sylvester expects the product’s price to drop as the company scales up production. Ford of CDOT says the current price is comparable to other early-stage technologies the agency is testing in its RoadX program.
To further improve his company’s product, Sylvester says his team plans to embed more strain-sensing fiber optic cables into each slab. He also wants to add a separate fiber optic cable to transmit data down the road itself, as an alternative to burying cables underneath or alongside roads. Eventually, he says he’d like the entire road to power itself by somehow “harvesting the energy that vehicles are constantly shedding.”
Still, as UNLV’s Morris points out, the product that Integrated Roadways will install in Colorado is meant to alert authorities after an accident happens. A more powerful tool would prevent accidents in the first place—the aim of many companies developing self-driving cars.
Morris thinks there may be an easier approach to smart roads in the short term by simply taking advantage of the many traffic and security cameras that are already mounted on light posts and buildings today. “We have a lot of coverage but there’s also a lot of room for growth in that most of those aren’t smart cameras. They’re just passively observing,” he says. “You can imagine how adding intelligence can add a lot of value to those cameras.”