Arkansas has a bad reputation when it comes to its highways. In 1999, it was voted the state with the worst overall highways in the nation in a survey of truckers for Overdrive magazine.

But identifying and maintaining road surfaces is not as easy as it may seem. Typically, a highway department will invest thousands of person-hours each year into surveying road surfaces visually and then cataloguing which areas show cracks and damage. Add to that the need to revisit stretches of highway to gather more data or check human error, and the process of assessing the health of a highway system can be cumbersome.

So the Arkansas Highway and Transportation Department collaborated with the department of civil engineering at the University of Arkansas to develop a system that allows highway department employees to work more efficiently than ever before.

Let's Go to Video

One of the results of the collaboration with the University of Arkansas was the multimedia-based highway information system (MMHIS). Development of this system began in 1996.

The MMHIS takes video images of a highway, currently gathered with the help of a vehicle called an Automatic Road Analyzer (ARAN), and digitizes the images into MPEG-2 format. Back in the office, the images are synchronized with basic road data (county, road mile, width, recent work completed) and data from the scan (rutting, roughness).

"We're trying to cut down on field trips; [that's] the biggest time and money saver," said Bobby Bradshaw, MMHIS coordinator for the Arkansas Highway and Transportation Department.

Although an initial drive over all the road surfaces in the system is required, the data and video can then be used again and again.

"We can make initial visits and answer the vast majority of questions without going back out," said Bradshaw.

For example, a crew was going to remove a sign from a bridge and was unsure whether they would need to close one or both lanes of traffic. Without the system, someone would have driven to the bridge to visually inspect the sign placement; with the system, the image was available on the desktop for immediate decision-making.

Ready for the Future

The University of Arkansas is not content to rest on its laurels with the MMHIS. Kelvin Wang, a professor of civil engineering and a principal investigator on the MMHIS project, is an advocate for using technology to reduce this major budget item for states. "Highways are the biggest expenditure of public funds," he said. "To monitor the health of highways is very important."

Wang has helped developed a data collection van that is even more precise than the commercially available ARAN. The new van incorporates a digital camera capable of shooting 12 frames per second. The computer then links the images together depending on the speed of the vehicle - more images are retained at higher speeds, fewer at lower speeds. The computer system, using a series of proprietary algorithms, then processes the images and identifies the length and width of cracks. All of this is completed with greater efficiency and accuracy than a team surveying the highway manually is capable of. Wang estimates that a manual survey is, at most, 2 percent to 10 percent accurate. The video survey, on the other hand, completely covers 14 feet of highway width, enough to survey an entire 12-foot lane.

The potential cost savings are clear. Wang estimates a purchase price of $250,000 for the digital van and computer system. A manual survey would require a highway department to hire 50 workers to survey 10,000 miles.

The digital system is not currently available commercially, but Wang expects it to be available soon. He has formed a company, WayLink Systems Corp., with the University of Arkansas to pursue bringing the technology to market. At this writing, they are in negotiations with a larger company that would be the entity for commercialization.

Where does Arkansas plan to take this technology? "We hope to add more data to make a more robust system," said Bradshaw.

The department also plans to save money - not through reductions in staff, but through increases in efficiency. They are looking at incorporating a GPS element into the system to better identify portions of highway, and they hope to deploy the data gathered to all personnel with a need to use it. "It's a logical extension of the things we have here," Bradshaw said.