Terry Bennett argues that in order to unlock the full potential of autonomous cars, transportation planners and policymakers need to think seriously rethinking infrastructure to better accommodate them. Some of his recommendations are familiar, e.g., that we should encourage more public-private joint ventures, but the core idea is the following:
Create ways for cars to collect, coordinate, and upload roadway info, so the physical environment can “listen” to roadway sensors and optimize traffic performance in real-time, as well as “learn” and adjust longer-term patterns. By leveraging and installing wireless transponders called Roadside Units or similar smart embedded sensors, cars can feed safety information into our highways and rural roads. Such information would include static road hazards like curvy roads or low bridges; changing risks such as construction; and information about traffic density, flow, volume and speed. It’s not unlike transferring data back and forth to high-speed tolling lanes — but with much richer data.
Autonomous cars have more potential to reduce traffic congestion than high-speed rail, which is designed to serve a relatively small number of intercity passengers. Investments in intelligent road infrastructure might thus be more cost-effective than investments in high-speed rail, particularly if these investments allow for better bus and paratransit service, as seems likely.