Street smart: is the highway of the future here?
Posted: January 17, 2025
On a three-mile stretch of the I-94 between Ann Arbor and Detroit, Michigan, you can now take a test drive on the highway of the future.
Mind you, this is a far cry from I, Robot, the sci-fi film that has Will Smith stylishly cruising futuristic tunnels in a sleek, self-driving sports car. (Luckily the I-94 doesn’t feature attacks from hostile robots, either.)
For now, all that’s visible on the interstate are slim poles lining the median every 200 meters. Their job is collecting and transmitting the essence of a smart road: reams and reams of data.
Our Industrial Life
Get your bi-weekly newsletter sharing fresh perspectives on complicated issues, new technology, and open questions shaping our industrial world.
How do smart roads work?
The 55-foot poles went up this summer as part of a trial by Michigan's Department of Transportation and Cavnue, a subsidiary of Sidewalk Infrastructure Partners, the infrastructure innovator spun off from Alphabet.
They contain cameras, radar units and small computers to gather real-time information on traffic, weather, potholes, obstacles and, eventually, connected and autonomous vehicles, so drivers or their cars can directly react to an accident or stalled vehicle, for example.[1]
The project is one of several in the US and across the world meant to advance smart roads, which envision a new era of connected infrastructure that features cutting-edge technology automating everything from traffic control to road maintenance.
But fully entering this new era will, first and foremost, require collecting lots of data from a variety of sources including:
- Traffic and environmental sensors to measure vehicle volumes and speeds, road surface conditions such as temperature, moisture and ice build-up, or even vibration and strain to keep tabs on the long-term structural health of roads and bridges.
- Cameras and computer vision systems to detect congestion, accidents, and unusual vehicle behavior, or scan license plates for automated toll collection and law enforcement.
- Connected cars themselves, which can share their speed, location, fuel levels and even internal diagnostics to complete the picture.
National Highways, the UK government agency that manages the country’s major road network, funded trials of six new smart road technologies this year, including a tool that automatically detects hazardous objects and a real-time road data analytics platform meant to provide road traffic insights and model accident risk.[2]
And in Australia, toll road operator Transurban is testing traffic management and communication with connected vehicles using the more than 5,000 cameras installed across its roads, which already detect anomalous traffic patterns.
Smart road features: from adaptive traffic signals to predictive maintenance
What does all this mean for actual road management? In theory, all that data should directly translate into insights that can help transport managers, or autonomous systems, better direct traffic, maintain infrastructure and prevent accidents.
Data about the cars on the road can feed into adaptive traffic signals, for instance, which adjust in real time based on current conditions. A sensor flagging heavy traffic at an intersection could extend a green light to push more vehicles through, cutting congestion and even emissions from idling cars. Outdated traffic signal timing currently accounts for more than 10% of all delays and congestion on major routes, according to the Federal Highway Administration.
Similarly, a smart road system could open additional lanes during rush hour and adjust speed limits in response to road damage or bad weather. During an accident, it could reduce incident response times and even reroute traffic altogether. Over the long-term, sensors keeping track of road health could also flag areas that need repairs before they fully deteriorate.
“There’s just not enough technology out there on the roadway to provide observations around risks and then the insights to deal with them,” says Tyler Duvall, the CEO and Co-Founder of Cavnue. “It’s not enough to see what’s going on, you have to provide judgment and insights to government and operators and roadways to make interventions.”
Much of this is still theoretical, however. Although the technology exists, it hasn’t always panned out as planned.
In the UK, radar-based stopped-vehicle detection systems were found to have suffered hundreds of power outages during a roughly two-year period, with some outages lasting several days.
Concerns over data privacy could cause yet more growing pains. A typical connected vehicle, which forms an integral part of future smart road systems, can generate nearly 25 gigabytes of data per hour and collect information from more than 100 different datapoints, according to S&P Global. That means huge potential for privacy violations, and automakers have already faced claims for alleged breaches in recent years.
The future of smart roads
Even Cavnue’s pilot project in Michigan, which will run through the end of 2025, has an uncertain future.
The company wants to expand the smart road to a 40-mile connected corridor and eventually feed information and advisories directly to connected and autonomous vehicles. But Michigan’s Department of Transportation (MDOT) hasn’t committed to it yet, partly hedging against the potential costs. (For now, Cavnue is funding the entire trial.)
"The technology is evolving, and MDOT wants to benefit from that technology,” says Ryan Mitchell, the director of the department’s Office of Major Projects. “But we don't want to put the public at risk for the cost of that technology, and whether it may or may not move in one way, or in one direction or another.”
In the meantime, more trials are set to launch to test smart road technology. Cavnue is deploying a similar system for a smart freight corridor around Austin, Texas, which aims to inform truck drivers of road conditions and eventually cater to self-driving trucks.[3] In Salt Lake City, Utah, commuter buses already communicate with traffic signals to gain a few extra seconds of green light when approaching intersections.
Last year, the Federal Highway Administration also granted $60 million to advance connected and interoperable vehicle technologies through three programs in Arizona, Texas and Utah. Just a few months later, the US Department of Transportation issued a federal blueprint to promote communication between roads and vehicles, with a target to deploy the technology on 20% of the national highway system by 2028.
Announcing the plan, Federal Highway Administrator Shailen Bhatt painted smart roads as all but a certainty. “This is not about pilot programs; this is not about trying to test out technology; this is proven technology that works,” he said.
