Green dots travel up and down highway veins on a computer map of the United States. The dots represent the cell phones, PalmPilots and laptops of users of CoPilot Live, a satellite navigation system developed by Operations Research Professor Alain Kornhauser GS '71. The forecasted travel times for each user are displayed on a panel.
Soon, CoPilot users, who are voluntarily tracked, may be able to use this database's information to avoid traffic and seek the quickest route in real time as they are traveling.
"This can help legions of individuals make slightly better decisions each day," Kornhauser said. "By providing the appropriate information, people can save a lot of time."
The United States currently has only 1,000 CoPilot users, considerably short of the million needed to provide forecasts based on live information about all U.S. roads. But Kornhauser predicted that Europe may boast 100,000 users by year's end, because Motorola's new A780 cell phones will be able to run CoPilot.
"Since this technology is from private companies, not the government, if people want this, it will happen quickly," Kornhauser said, referring to an alternative and time-intensive plan that would have the government install sensors on roads nationwide.
"This is almost the antithesis of central planning," he added. "What I like about all these things is that there are off buttons. I'd rather let it be my call than [the government's] call."
Kornhauser received his bachelors, masters and doctoral degrees in aerospace engineering, but said he found the transition into operations research to be fairly smooth. "[It was] still going from point A to point B, except now point B wasn't Mars anymore," he said. "I decided to use what I learned to maybe improve urban society."
The technology originated in the pioneering computational transportation work that Kornhauser and others at the University have done since the 1970's, when they started using graph models to analyze railroad routes.
He tests CoPilot by frequently taking wrong turns in his 30 year-old black Mercedes. The handheld computer mounted on the dashboard instantly gives him a new set of directions — useful in a traffic bottleneck.
The direct predecessor to CoPilot Live emerged in 1995, when ALK Technologies — a company founded by Kornhauser in 1979 — released Road Trips, the first consumer software program that gave door-to-door driving directions.
Road Trips was similar to today's Yahoo Maps or Mapquest online driving directions in that it used an algorithm to determine the most efficient way to get from one point to another.
"Road Trips planned routes based on 'costs' associated with left and right turns, depending on what kind of street you were on, and toll roads, and one way roads and so on. There was the computational side of it, and the art of making routes people themselves would think is the fastest. It certainly wasn't the shortest route," Kornhauser said.

Then Microsoft entered the market with its own route planning software, and Road Trips was left behind commercially.
Now that CoPilot can use live information to find the best route every few seconds, Kornhauser's ALK, which has continued serving commercial clients, may move back into the consumer market in a big way.
Kornhauser said that the soon-to-be-released eighth version of CoPilot Live is close to his original ideal. "What we always wanted to do was to get [navigation software] onto cell phones, so that you could generate routes every second, based on real-time information, where users both contribute and consume information."
When a device running CoPilot Live is turned on, the device's GPS receiver scans the sky for satellites. If there are no obstacles like tall buildings, otherwise known as "urban canyons," which interfere with satellite signals, then CoPilot uses a GPS receiver to receive data from four satellites to compute the user's position, direction and travel time.
To give this information, every CoPilot tells a database about how long it took to travel between two waypoints, or "monuments." The database in turn tells other users how long it would take them to travel between the same monuments. Much of Kornhauser's work has been in determining the location of monuments, of which there are currently about 125,000 in the United States and 125,000 in Europe, and in forecasting estimated travel times.
Kornhauser heads the University's program in transportation and is also a co-director of the Center for New Jersey Transportation Information & Decision Engineering (TIDE). He said that both programs have been based around comparatively small-scale decisions within operations research.
"Much of the field focuses on big questions like military deployment or corporate asset management. We focus on decisions made by millions of people at the individual level each day," Kornhauser said.
Even though he may be working on a small scale, Kornhauser said he still thinks of transportation as having the potential to be revolutionary.
"We don't have to have systems that look like this," he said, pointing to the complicated lighted dials and controls in the inside of his almost 30 year-old black Mercedes. "If we were in 1905 designing an automobile and came up with something like this, people would have thought we were crazy," Kornhauser said. "To think that we're still going to have this in 2105 is unthinkable."
Kornhauser is also affiliated with the Wilson school, and said that his scientific work cannot be separated from his "dabbling" in policy.
This fall, Kornhauser is teaching ORF 467: Transportation, which will focus on the planning and systems analysis behind transportation — including the problems arising from hurricane Katrina. He is also teaching WWS 527a: Transportation Policy Analysis & Systems Planning, which will cover the same topic from a more policy-oriented perspective.