Smart cars could prevent collisions

Large-scale U.S. test program paves the way for eventual collision-free driving

Published: September 19, 2012, 9:00 AM

GM Vehicle-to-Vehicle (V2V) technology diagram

The idea that cars, rather than drivers, could prevent themselves from colliding with each other is not a new one. Several manufacturers already offer such features as active cruise control and automatic emergency braking.

But there's more that they could do. A lot more if it were possible for them to talk to each other.

That's the premise behind a large-scale "smart car" test program, involving nearly 3000 vehicles, begun recently by the University of Michigan Transportation Research Institute (UMTRI).

It's a year-long study called the Safety Pilot Model Deployment (SPMD) – part of a $22-million (US) partnership between UMTRI and the U.S. Department of Transportation, supported by the auto industry.

Eight automakers – Ford, General Motors, Honda, Hyundai, Mercedes-Benz, Nissan, Toyota and Volkswagen – have all provided vehicles for the test program.

The project's purpose is simple: to see how well automatic wireless communication among vehicles works in the real world.

The participating vehicles will be driven by local residents of Ann Arbor, Michigan, (where UMTRI is located) over the course of a year, while researchers collect and analyze data to assess the systems’ effectiveness.

The vehicles include passenger cars, commercial trucks and transit buses and they are fitted with wireless devices that enable both V2V (Vehicle-to-Vehicle) and V2I (Vehicle-to-Infrastructure) communication.

In simpler terms, the vehicles can "talk" to each other, as well as to traffic lights and other road signals located at intersections, curves and highway sites throughout a defined test area in and around Ann Arbor.

The data transmitted and received include such things as vehicle position, speed and direction. But that communication will be invisible to the vehicles' drivers – they'll never know it's taking place – except in the event of a potential crash situation.

In the case of a vehicle unexpectedly braking ahead, a sudden lane change or merging traffic, for example, the drivers in some of the vehicles will be given a visual or audible warning.

The V2I communication shares information about traffic signals, road attributes and surface conditions from 29 roadside transponders installed over approximately 120 lane-kilometres of area roads.

The wireless communications take place on a secure channel (the 5.9 GHz Dedicated Short Range Communications (DSRC) band) allocated by the U.S. Federal Communications Commission specifically for the purpose.

While all the vehicles can talk to each other and those transponders to establish the overall scenario, only some are equipped with the warning systems.

While it is technically possible to have at least some of the vehicles take over partial control from the driver in the event of an emergency, given the availability of features such as emergncy braking, this project doesn't extend that far.

It's limited, at this time, to identifying potential problem scenarios and warning the drivers accordingly. If it works as well as hoped, the next step could go further.

Connected vehicle technology has the ability to address as much as 80 percent of all crashes involving unimpaired drivers and to greatly reduce carbon emissions, according to UMTRI director, Peter Sweatman.

"This is a game-changer for transportation," enthused program manager Jim Sayer, an associate research scientist at UMTRI.

And U.S. Transportation Secretary, Ray LaHood is just as enthusiastic. "This is a big moment for automotive safety," he said "This cutting-edge technology offers real promise for improving both the safety and efficiency of our roads."

At the conclusion of the Michigan test, the U.S. DOT will consider rulemaking related to these technologies.

While SPMD is the largest program of its type, it's not the only one. A 120-vehicle fleet is being used to test 20 experimental driver assistance technologies as part of a four-year research project called "Safe Intelligent Mobility" (SIM) in Germany.

And another European project called SARTRE (Safe Road Trains for the Environment), which researched the potential for implementing electronically-connected road trains on conventional highways, has just concluded.

Governments, standards organizations and global automakers are all work together on these projects, not just to test the technology but to help establish common standards and security frameworks so vehicles from different manufacturers can communicate and interact with each other.

According to Mike Shulman, technical leader of Ford Research and Advanced Engineering, the earliest the systems could be implemented would be six to eight years.

But, says Hariharan Krishnan, a GM R&D technical fellow, "It will take approximately another five years of market penetration for customers to truly benefit from the technology."

All of which means that, within ten years, cars really may be able to keep from colliding with each other.