Purdue News
Purdue News
If you're tired, you can take a nap. If you get hungry, your vehicle will find the best and nearest place to eat. If weather makes a highway dangerous or impassable, your vehicle will find the best alternate route. If you use a toll road, you won't have to stop at each booth -- fares will be electronically deducted from your debit card as you proceed.
"You may be able to go as fast as 200 miles per hour without touching your steering wheel, accelerator or brake pedals," says Kumares Sinha, professor of transportation engineering at Purdue University.
This transportation paradise may be several years away, but researchers of intelligent transportation systems (ITS) see it just over the horizon. ITS includes a wide array of technologies including information processing, communications, control and electronics. Several aspects of this research take place at Purdue, scattered across diverse disciplines: civil, industrial, mechanical, electrical and computer, and agricultural engineering.
"Everything falls under the general category of making transportation more efficient and safer," says Jason Papastavrou, associate professor of industrial engineering.
Eventually, intelligent highway systems will reduce congestion, inefficiency, danger and pollution -- the problems that beset modern transportation infrastructure.
Every year, according to government statistics:
"Technically speaking," says Sinha, "a highway has infinite capacity -- you could have vehicles bumper to bumper going 200 miles an hour. But realistically, right now the capacity is about 2,000 vehicles per hour per lane. In many places it is not feasible to just add more lanes, so we have to find some other way to increase highway capacity."
That's where the "intelligent" in ITS comes in.
According to Sinha: "Automobiles are becoming very smart, with features such as cruise control, electronics in the engine, temperature sensors, and so on. However, roads are still dumb. The question is, how can we make the infrastructure smart in conjunction with the vehicles?
Sinha and Papastavrou represent two of the many thrusts of this research: traffic planning and driver safety.
Sinha has observed and participated in such research for nearly 25 years, and he says the U.S. research effort accounts for several hundred million dollars a year, with an emphasis on private-public-university consortia.
He heads the transportation and urban engineering area in Purdue's School of Civil Engineering and also directs the Joint Highway Research Project, a 60-year-old state-funded program through which Purdue conducts transportation research for the state of Indiana.
Much of the work done by Sinha and his transportation engineering colleagues has focused on the Borman Expressway, a 15-mile stretch of Interstate 80/94 in northwestern Indiana. Each day it handles around 125,000 vehicles, of which one third may be trucks.
"It's one of the most heavily congested traffic corridors in the country," Sinha says. It's the site of 15 to 20 daily "incidents" -- accidents, mechanical breakdowns, emergency spills and the like -- that produce nonrecurring congestion.
"Recurring congestion is what you find during peak periods of highway use," he explains. "You can predict they are going to happen. You cannot predict nonrecurring congestion. They come from incidents that occupy part of the highway and reduce its capacity, leading to a backup."
The research on the Borman Expressway aims to develop a network of sensors and decision-making centers that monitor traffic, detect incidents, and respond by (1) deploying emergency crews to accidents, breakdowns, or other obstacles to smooth the flow of traffic, and (2) alerting motorists "upstream" and suggesting alternate routes.
Whereas Sinha's research focuses mostly on the relationship between vehicles and highways, Papastavrou is more interested in the relationship between drivers and their environment -- including their vehicles.
Papastavrou's research will have an impact on driver efficiency and safety, he says, as two examples illustrate:
"My mother, when she sees the light come on, immediately goes to the first gas station to fill up," Papastavrou says. "So she fills up practically eight times when only seven would be required. If the car knew that my mother was driving, it would give the warning at one-sixteenth of the tank remaining, to make her more efficient."
"Those warning signs are fixed," Papastavrou notes, "but if you think about it, they should change with the time of day, with the weather conditions, with the traffic patterns. Why is it always 65? On days when it's snowing, it might be dangerous for anybody to go above 45. But when the weather is good and traffic is light, it might be safe to go 85. An intelligent system would have signs that adapt."
Papastavrou and his industrial engineering colleagues, particularly Mark Lehto, associate professor of industrial engineering at Purdue, also study safety and human factors. Recently they completed simulation testing of a mechanism that will help drivers determine when it is safe to pass a vehicle.
"The warning mechanism would look at the car in front of you, and at the car coming toward you. It would measure the speed of those vehicles and know the speed of your vehicle. It would also know your driving tendencies and would give you a green or red light on whether or not it is safe to pass," Papastavrou says.
Most important, the system would learn how a driver performs and adjust its recommendations accordingly.
For those who fear that the high-tech future seems too intrusive, there's good news. At this stage of its evolution, the research is trying to assist rather than replace drivers. As Papastavrou says, "So far we have not designed a system that will reach out and make a driver pay attention to what the system is trying to say."
aas/sinha.its/9703f18
Sources: Kumares Sinha, (765) 494-2211; e-mail, sinha@ecn.purdue.edu
Jason Papastavrou, (765) 494-5414; e-mail, papastav@ecn.purdue.edu
Writers: Sig Kriebel, Office of the Dean of Engineering, (765) 494-9988; e-mail,
kriebel@ecn.purdue.edu
Amanda Siegfried, Purdue News Service, (765) 494-4709; e-mail, amanda_siegfried@purdue.edu
Purdue News Service: (765) 494-2096; e-mail, purduenews@purdue.edu
