If you think talking on a cell phone while driving is distracting, consider a police officer flying down the highway, reeling off a license plate number to a dispatcher at the same time he or she is turning on the cruiser’s siren and blue lights.
That’s what the brains behind Project54 did more than 10 years ago, before cell phone use in automobiles became as common as CD players: they thought about the risks of performing those kinds of tasks at any speed and figured out a way to make it safer.
Project54 is a system that was developed in UNH’s Consolidated Advanced Technologies Laboratory (CATlab) to operate a patrol car’s equipment—radio, siren, lights, radar, etc.—with voice commands, allowing police officers to keep their hands and eyes on the road. To date, more than 1,000 emergency vehicles in 180 agencies in New Hampshire have the hands-free software that is used by first responders as well as members of law enforcement agencies.
Named for the 1960s television show “Car 54 Where Are You?”, the program was launched in 1999 with funding from the U.S. Department of Justice and support from then-U.S. Sen. Judd Gregg.
In addition to voice command, Project54 allows individual cruisers to be integrated into agency-wide data networks. That means police can run background checks on drivers they may be pursuing, or share information they are gathering in the field with individuals back at headquarters.
Queries can be made using voice commands via the car radio or the information can be keyed in using the CD touchscreen and keyboard installed in the patrol car as part of the Project54 program.
“The Project54 technology has made a very positive impact throughout New Hampshire,” says Andrew Kun, associate professor electrical and computer engineering and Project54’s principal investigator. “You can ban ordinary drivers from using cell phones but police officers have to communicate. So you work on minimizing the distractions.”
One of the next pieces of the work is figuring out what kinds of answers computers should give in response to voice commands.
“Should they be long, short, spoken fast, slow?” Kun says. “These are aspects that haven’t been explored. If we’re going to build computers that you use like cell phones, that’s not much gain. We’re interested in how to learn from human actions.”
That interest includes finding ways to analyze cognitive load, which relates to the amount of information and interactions one can process simultaneously, and how that impacts human -computer interactions.
“That’s an exciting area: human -computer interactions,” Kun says. “There’s a lot of exciting science that needs to be done in the area.”
For more information on Project54 visit http://project54.unh.edu/.