KEY POINTS

  • Ground-penetrating radar can map roads even in snowy or foggy conditions
  • It is more accurate than existing LIDAR technology
  • It can be used in combination with other technologies to make self-driving safer

One of the biggest challenges for self-driving is unpredictable variables such as weather, which can affect the journey. A team of researchers from the Massachusetts Institute of Technology’s Computer Science and Artificial Intelligence Lab (MIT-CSAIL) seems to have found a solution to this vexing problem.

The researchers have developed technology that will let self-driving cars “see” through snow and fog, using a Ground Penetrating Radar (GPR). Using the technology, the car systems will be able to see beneath the surface of the snow and map routes accordingly.

“Self-driving cars are likely to be safer, on average, than human-driven cars. But they may fail in new and catastrophic ways that a human driver could prevent. This project is designing a new architecture for a highly dependable self-driving car,” CSAIL stated on its website.

Currently, most self-driving cars use LIDAR (Light Detection and Ranging)-based sensors, but these can get affected by weather such as lightning or fog. LIDAR works like human drivers do —  it navigates the road in front of the vehicle, except it can do this for much longer distances, and thereby makes self-driving cars safer. But, extreme weather will affect the accuracy of these sensors as they can only map what is in front. If the visibility is low due to fog, or the route is not properly visible due to snow, it can affect this mapping.

GPR is, therefore, a much safer method of mapping a route since it sends electromagnetic pulses into the ground to map out the route. Since it uses feedback from beneath the ground, it doesn't face the issue that LIDAR does and therefore, can improve the quality of navigation.  The CSAIL team has used a refined version of GPR, the localizing GPR or LGPR, developed at the MIT Lincoln Laboratory, which can map the route with even more accuracy.

"If you or I grabbed a shovel and dug it into the ground, all we're going to see is a bunch of dirt. But LGPR can quantify the specific elements there and compare that to the map it's already created, so that it knows exactly where it is, without needing cameras or lasers," CSAIL Ph.D. student Teddy Ort told Engadget.

The radar needs to be refined for rainy conditions when water seeps into the ground, but it can be used in combination with other systems, such as LIDAR to make the navigation much safer. It refines mapping for self-driving vehicles to make the technology even more precise, which can help navigation, especially over long and varied routes, with different terrains.