Automated car research around the world has invested a great deal of time and research. The Massachusetts Institute of Technology has announced that it is working on a system that could benefit autonomous cars that drive in rain and snow. The new system developed by the team uses an existing technique called geo-test radar to send electromagnetic pulses underground.
MIT’s system, known as the Local Geodesy Radar, creates a map for cars to use when they return to the same land later. The system can launch waves to the ground and quantify specific elements in the ground to create fingerprints. The data can then be used to identify the future location of the car.
The purpose of this feature is to allow navigation on snow-covered roads or in heavy rain. With this information, the car doesn’t need a camera or laser to run automatically. In the tests, the team found that in the event of snow, the average error range of the navigation system was about 1 inch compared to when used in clear weather.
The self-driving cars used in the tests were able to run for more than six months, and the team never had to step in. The team tested the car at low speed on a closed county road. The researchers say the system can easily be extended to highways and other high-speed areas.
The MIT system represents the first time self-driving developers have deployed geo-radar. The technology was previously used mainly in building planning, mine detection and lunar exploration. The system cannot operate independently because it can’t see anything on the ground. It needs to be paired with lidar and vision systems.