The Massachusetts Institute of Technology takes a different approach to self-driving vehicles with its self-driving car system called MapLite.
Unlike most, it does not only work in well-mapped cities but handles navigation everywhere else.
Exploring New Paths
For a while now, self-driving vehicle technology is mostly applied by companies within big cities due to the availability of navigation information. Details like GPS coordinates, lanes, signs, and curbs are all taken into consideration for the program to properly position the car and navigate on its way to the destination.
Meanwhile, the team from MIT created a new system that is reportedly able to work its way around unpaved roads and paths.
Roads Less Traveled
According to reports, MIT's Computer Science and Artificial Intelligence Laboratory, otherwise known as CSAIL, in cooperation with Toyota, created a new self-driving car system called MapLite. The new framework still relies on GPS to establish its location relative to its destination.
However, during navigation, it apparently does not rely on 3D maps like most of the autonomous vehicle technology. In its place, the program takes advantage of the various sensors onboard to generate a path to its destination.
Moreover, the entire setup uses LiDAR to come up with a rough estimate of the road's edges. The technology is far from perfect but it seems like a step in the right direction for self-driving cars in the future.
A Work In Progress
As the technology is still in its infancy, the team admits that there are still some challenges that need to be tackled. At its current state, the system needs more work in order to properly identify sudden changes in elevation, which is common with mountain roads.
It is speculated that the ultimate goal of the autonomous navigation technology is to properly identify its location, detect the type of paths available, and switch to the appropriate sensors seamlessly. Once perfected, it will allow the artificial intelligence to take over and improvise when it is needed.
Unfortunate Incidents With Self-Driving Tech
News about different improvements for self-driving car technology is definitely great but the challenges in regard to public safety could be given more focus. While experts are pressured non-stop to develop new ways to make autonomous vehicles to function seamlessly in public, there are recent incidents, which should be taken into account in terms of overall safety.
Just last week, a Waymo self-driving van was involved in a crash in Chandler, Arizona resulting in minor injuries. Back in March, an Uber self-driving car made the headlines when it hit a pedestrian in Tempe, Arizona, which unfortunately resulted in a fatality.
Teams from MIT and other companies involved with self-driving cars can hopefully work on more safety measures to prevent accidents like the ones stated above.