Researchers at the University of Arizona have taken on the issue of intelligent multimodal traffic monitoring with their development of a high-resolution radar sensor that combines the best results of cameras and radar. Cameras work well in settings with good visibility but are limited in rainy, dark or foggy conditions. Conventional radar provides movement and position data but does not effectively distinguish between the various approaching entities (i.e., cars, bikes, pedestrians).
The sensor distinguishes between cars and pedestrians and supply counts, speed and direction of each moving target in all lighting and weather situations. The high-resolution millimeter-wave (mmWave) radar sensor used in the prototype provides a richer picture than conventional radar and does better than cameras in low-visibility conditions. Unlike lidar and other light-based systems, mmWave radar can reliably resolve the speed of a moving target.
"The mmWave radar is also different from other sensors in that it can provide relatively stable radial velocity, which is very helpful for us to identify the speed of vehicles," said Dr. Siyang Cao. "The key problem in multimodal traffic monitoring is finding the speed and volume of each mode. A sensor must be able to detect, track, classify, and measure the speed of an object, while also being low-cost and low power consumption. With real-time traffic statistics we hope to improve traffic efficiency and also reduce the incidence of crashes," Cao said.
The UA team’s prototype is low-cost, lightweight and compact so it is easy to install. The researchers installed their sensor at an intersection in Tucson and monitored it with a laptop from a nearby parking lot. A signalized intersection with mixed traffic was chosen because conflicts and crashes tend to involve multiple transportation modes. With the multivariate Gaussian Mixture model they developed to interpret the information gathered from the sensor, they were able to achieve promising results in terms of detecting and correctly identifying objects.
"We realize that sensor technology is moving to a stage that's going to have a lot of new applications. On one hand, the cost of sensors is dropping and their performance is improving significantly. Meanwhile the surrounding technology — for example battery technology, communications, and computation enabled by artificial intelligence — is also improving. For multimodal traffic monitoring, a sensor can collect information that can be shared with drivers via a next-generation communication network to improve mobility and safety at intersections," Cao said.
The researchers plan to further develop their model so it can interpret more complex data and identify motorcycles, bicycles, trucks and buses.