MIT is working on a self-driving car that works on unmapped back roads

Self-driving cars have a ton of potential, but typically you see company’s testing applications on major highways and cities. One reason for this is, obviously, more people will use self-driving vehicles in these areas, but it is also because these areas are so well documented. Road signs, clear lines, and plenty of GPS documentation makes these areas perfect for the technology, but it doesn’t address drivers who are from smaller towns and areas where this documentation isn’t as readily available.

Well, MIT is working on a solution for this. MIT CSAIL with cooperation for Toyota is working on a self-driving car that uses a newly developed framework called MapLite which looks to address some of these issues.

Basically, by using LiDAR and GPS, the vehicles can estimate the sides of the road and any issues that way arise around the vehicle. The GPS is used to determine the final destination and this combination is what makes it possible for the cars to handle back roads. For crossroads and intersections, the system bases its decisions on a number of parameters in an effort to make the best decision.

We’re still some time away from seeing any of this tech in the current stable of autonomous vehicles, as it still has issues with some situations and large elevation changes, these advancements are a great step in the right direction. “Traditional” self-driving cars are great, but struggle when strange scenarios pop up, this could be the answer for those times.

Are you ready for the self-driving revolution or are you still extremely skeptical? Let us know in the comments below. 

Josiah Motley

Former KnowTechie editor.