The News Wheel
No Comments

MIT Tech Helps Driverless Cars Navigate in Dusty and Foggy Conditions

Decrease Font Size Increase Font Size Text Size Print This Page

Driverless cars face many challenges as the auto industry moves closer toward fully-autonomous models. For one thing, they’re still learning how to predict and respond to human driver behaviors. Additionally, their LIDAR systems have trouble detecting dark-colored cars.

Autonomous vehicles have another huge obstacle to overcome: “seeing” through fog and dust. Since these models rely on light-based image sensors to detect surrounding obstacles and other vehicles, light-obscuring conditions like fog and dust can temporarily blind AV sensors.

But good news — researchers at MIT have developed a new technology that uses sub-terahertz wavelengths to help driverless autos navigate in low-visibility conditions.


Low on Tread? Simplify your search for new tires


How the technology works

The new technology can detect sub-terahertz wavelengths, a type of wavelength that ranks between microwave and infrared radiation on the electromagnetic spectrum. And sub-terahertz wavelengths can be sensed through fog and dust clouds. Conventional LIDAR systems rely on infrared wavelengths, which AV sensors just can’t detect in foul weather.

The new system works by sending a signal through a transmitter. Next, a receiver measures the absorption and reflection of the rebounding sub-terahertz wavelengths. It then transmits the signal to a processor to generate an image of the object.

Future goals

Ruonan Han, director of the Terahertz Integrated Electronics Group in the MIT Microsystems Technology Laboratories, says that the new technology is a cost-effective solution meant to work with conventional LIDAR systems, instead of replacing them.

Only time will tell if this innovative new tech will reach the market-ready product stage, so automakers can incorporate it on driverless vehicles.


Smart Technology: Discover the advantages of OnStar


News Sources: The Drive, MIT News