Most of today's autonomous vehicles are being tested in specific areas - typically major cities - where details of the environment have all been carefully mapped out by hand. These self-driving vehicles, say the researchers, rely heavily on these 3D maps and only use sensors and vision algorithms to avoid dynamic objects.
"The cars use these maps to know where they are and what to do in the presence of new obstacles like pedestrians and other cars," says Daniela Rus, director of MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL). "The need for dense 3D maps limits the places where self-driving cars can operate."
Given the difficulties in mapping roads that are unpaved, unlit, or unreliably marked, and the lack of incentives to map those that are less trafficked, say the researchers, means that "there are huge swaths of America that self-driving cars simply aren't ready for." The answer, they say, lies in creating autonomous systems that are advanced enough to navigate without such maps.
To address this, the researchers developed a framework - called MapLite - that allows self-driving cars to drive on roads they've never been on before without the benefit of 3D maps. The system combines simple GPS data of the type found on Google Maps with a series of sensors that observe the road conditions.
Using a Toyota Prius outfitted with a range of LiDAR and IMU sensors, the researchers were able to use MapLite to autonomously drive on multiple unpaved country roads in Devens, MA, and reliably detect the road more than 100 feet in advance.
"The reason this kind of 'map-less' approach hasn't really been done before is because it is generally much harder to reach the same accuracy and reliability as with detailed maps," says CSAIL graduate student Teddy Ort, a lead author on a related paper about the system. "A system like this that can navigate just with on-board sensors shows the