Let's face it -- common sense rarely applies to many drivers. Look at the car in the lane next to you. There's a good chance the driver is texting, checking his email or updating his Facebook page via smartphone. Throw in all the road-raging tailgaters, cutthroat lane-changers and speed demons blowing red lights, and you have roadways that resemble scenes from "Mad Max 2: Road Warrior."
On a serious note, since 2000, there have been 110 million car accidents in the United States, more than 443,000 of which have been fatal. That averages to 110 fatalities per day. Statistics like these make traffic accidents one of the leading causes of death in this country, as well as worldwide.
Despite many advances in recent years -- blind spot warning systems, traction control, cars that park themselves -- engineers continue to forge ahead in their quest to achieve the long-term goal of intelligent transportation: cars that can "see" and communicate with other vehicles to prevent virtually 100 percent of crashes.
As intelligent transportation systems (ITS) become more standard, they will still have to contend with human-operated vehicles as long as older models remain on the road.
That's why a team of engineers from the Massachusetts Institute of Technology are working on a new ITS algorithm that incorporates models of human driving patterns to warn drivers of potential collisions and that ultimately takes control of the vehicle if needed to prevent traffic accidents.
The theory behind the algorithm and experimental results will be published in the journal IEEE Robotics and Automation Magazine.
Co-authors Rajeev Verma, a visiting Ph.D. student, and Domitilla Del Vecchio, assistant professor of mechanical engineering, determined that driving actions fall into two main categories: braking and accelerating. Depending on which the driver is doing, there is a finite set of possible places the car could be in the future. Combined with predictive models of human behavior (when and where drivers slow down or speed up), this set of possible positions all figured into the new algorithm.
CURIOSITY.COM: 10 Real-World Applications for Quantum Mechanics
The result is a program capable of computing, for any two vehicles on the road, a defined area where two cars are in danger of colliding. Using information from onboard sensors, along with roadside and traffic light sensors, the ITS-equipped car reacts accordingly, making situation-based decisions to prevent a crash.
When both vehicles are equipped with ITS, they're able to cooperate with each other by communicating their postions to avoid a collision.
The team tested their algorithm with two miniature vehicles (one autonomous and one controlled by a human) on overlapping circular tracks. Eight volunteers participated to account for varied driving styles. Of 100 trials, there were 97 instances of collision avoidance.
[Via PhysOrg]
Credit: Melanie Gonick
Tags: Auto, Cars, Driving and Safety, Transportation, Transportation Infrastructure
comments ( )