The State of the Connected Autonomous Nation

Connected, autonomous vehicles traveling down their own designated, uncongested lanes and improving mobility on a smart highway.

That is the reality which a joint venture aims to create in a first-of-its-kind connected corridor between downtown Detroit and nearby Ann Arbor, home to the University of Michigan. CAVnue, as the project is called, stands for “connected and automated vehicle corridor,” and states on its website that it plans to “develop technology and infrastructure for projects” that seek to improve infrastructure through the application of technology.

Its partners include some heavy hitters like Sidewalk Infrastructure Partners (a Waymo sibling that’s also part of the Alphabet family), Ford, the University of Michigan and various state and local government entities. The joint venture is still in its early, planning stages of development but, if it pulls off a real-world testing area for vehicles and smart infrastructure, that could be a gamechanger according to some experts. “It’s almost like a living laboratory to find out what works and what doesn’t work,” Bruce Belzowski, managing director, Automotive Futures, told TU-Automotive.

One aspect of the plan that Belzowski and other experts admire is the planned separation of AVs and connected vehicles from legacy models. That would allow for cars, trucks and city buses to move freely and experiment in a safer, more controlled environment than on a typical urban freeway. The dedicated lanes can also help metropolitan areas meet transit goals as buses and shared mobility vehicles may move faster in these lanes. “Separating autonomous vehicles and connected vehicles from regular vehicles will make it so much easier… taking the confusion and challenge of interacting with regular vehicles out of the equation,” Belzowski said.

He also said this is the way to structure a real-world test of AVs, if there’s room for a closed corridor in other locations, otherwise, he said: “You either build new lanes or take out lanes that are already there. HOV [high-occupancy vehicle] lanes may be an example of taking lanes away (from unconnected vehicles and non-AVs).”

The CAVnue project and other similar research is building on the University of Michigan’s Mcity test facility where researchers have been conducting experiments and running scenarios for AVs and connected vehicles. Henry Liu is professor of civil and environmental engineering at the university as well as director at the Center for Connected Automated Transportation. In an interview, Professor Liu said CAVnue could be a blueprint for the next step in testing automated vehicles: “If successful, I think CAVnue will provide a template and a test bed for the technology. We need it to validate the technology. If it works, it can be a model replicated in other places and expanded to other scenarios.”

He added that now is the time for these projects to get started and get funded, given the massive $1Trn infrastructure bill that may soon bring a windfall to this industry and speed the process for making these plans a reality. Liu notes that as automakers continue developing Level 4 and Level 5 vehicles, dedicated corridors such as CAVnue may also help lower costs for automakers, governments, and eventually for consumers. “Infrastructure-assisted automated driving will reduce the number of sensors and costs for vehicles, so fewer and less expensive sensors will be needed.”

He also sees infrastructure service providers (probably tech firms) subsidizing some costs of these test projects to speed up deployment of AVs. Liu says he’s currently working on a smart intersection project with the US Department of Transportation to help pave the way for national AV deployment. His $10M project uses 21 connected intersections in Ann Arbor that will communicate and convey information to AVs. Liu said, “My research group is also working with the American Center for Mobility to deploy an AI-based scenario generation toolbox integrated with augmented reality testing methodology to speed up testing of AVs.”

There are other considerations and learnings from real-world testing that experts are keeping a close eye on. CAVnue should be considered as much a policy experiment as a technological one, according to Jonathan Levine, professor of urban and regional planning at the university. He advocates for public transit and high occupancy connected and autonomous vehicles to receive priority or exclusive access to dedicated corridors or lanes.

Levine said: “New technologies have the potential to do great harm as well as great good. A future of self-driving vehicles could be a boon to environmental and societal goals, or it could exacerbate… the gaps between the transportation haves and have-nots.”