U. Michigan Launches World’s Biggest Sandbox for Wi-Fi Connected Cars

A not-so-futuristic world in which drivers, cars and roads operate in a highly connected network of instantaneous data got one step closer to reality today.

The scale of the project, managed by the University of Michigan’s Transportation Research Institute (UMTRI), is unprecedented. For the next 12 months, nearly 3,000 cars — operated by drivers specifically recruited because they frequently drive in the same quadrant of Ann Arbor, Mich. — will be integrated via Dedicated Short Range Communications (DSRC), a technology similar to the wi-fi network you use at home or the local cafe.

But instead of your laptop or smart phone connecting to the web—so you can check your latest Facebook messages—these thousands of cars will beam safety messages and warnings to their drivers, each other and to a dedicated cloud of computers. Each vehicle will transmit about 10 messages per second.

“The participants are parents driving kids to school, picking them up after school, or driving them to ballet or football,” said Peter Sweatman, UMTRI’s director. “We have a platform, with five or six applications on that platform intended to avoid certain major classes of crash, whether at an intersection, a lane departure, a rear-end or whatever it might be.” When dangerous conditions occur, drivers will be warned via some combination of visual signals, sounds and vibrations.

Big Data on Wheels

All of the data will be recorded, so researchers can determine the accuracy of the warnings, and which types of alerts are most effective at helping drivers avert danger. At this point, there is no automated vehicle control, but given the number of sensors on today’s vehicles, that’s a logical subsequent step, according to Sweatman.

Most of the vehicles are owned by the participants and fitted with after-market safety equipment and one-way communications devices. In addition, 64 cars supplied by participating automobile companies will be loaned for a year to drivers. These vehicles have been outfitted with embedded communications equipment—connected to the car’s onboard computer network—and fitted with the carmaker’s customized warning interfaces and multiple video cameras.

This project is the culmination of about 10 years of work by the U.S. Department of Transportation, as well as a long list of partner organizations, automotive companies, and educational institutions such as UMTRI.

This latest phase, which launched today, represents the deployment of technology into the real world with everyday drivers. As part of this phase, the connected vehicles can also “talk” to dozens of traffic signals, and sections of roads deemed to be particularly dangerous.

Incredible Platform

The combination of data and video gathered in Ann Arbor will make researchers practically omniscient. “We not only know what messages and warning are being sent to drivers. We know exactly what the driver is doing in great detail,” said Sweatman.

“We know how many occupants there are, what they are doing, the driver’s facial expression, and where they are looking. We can see if they’re texting, what they are doing with their hands, and with video outside the vehicle, we know what the traffic scene is.”

The model deployment is a $25-million pilot with approximately 80 percent of the funding provided by the U.S. Department of Transportation. When you consider that there are 34,000 fatalities, costing about $240 billion annually, according to the National Highway Traffic Safety Administration, the cost seems more than justified.

Eight major automotive manufacturers — Ford, General Motors, Honda, Hyundai-Kia, Mercedes-Benz, Nissan, Toyota and Volkswagen — are providing support for the research through partnering agreements.

“In the longer term, I’m really interested to see what else gets built on this incredible platform,” said Sweatman. “It’s a sandbox, a starting point for entrepreneurs for all kinds of applications, not only to avoid crashes, but to make the traffic flow better and to save energy.” He said that all architectures, standards and specifications are publicly available.