MIT Team Overcomes Major Hurdle To Improve Self-Driving Cars

June 04, 2018
BY Peter High Founder and President of Metis Strategy
Icon Scrolling Bar



By Peter High. Published on Forbes

Most of the tests of self-driving cars have taken place in cities. These are places where 3D mapping of streets, lanes, curbs, signs, and the like is undertaken with precision. One of the biggest challenges for self-driving cars is to navigate the roads less traveled.

“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 3-D maps limits the places where self-driving cars can operate.” Rus and colleagues at CSAIL have developed MapLite, a framework that allows self-driving cars to drive on roads they’ve never been on before without 3-D maps. I caught up with her recently to find out more about this innovative idea.

Peter High: Please describe how MapLite is different from other self-driving technology?

Daniela Rus: Most self-driving car companies only test their fleets in major cities where they’ve developed detailed 3D maps that are meticulously labeled with the exact positions of things like lanes, curbs and stop signs. These maps include environmental features detected by the sensors of the vehicle. The maps are created using 3D LIDAR systems that rely on light to scan the local space, accumulating millions of data points and extracting the features defining each place.

If we want self-driving cars to be a viable global technology, this reliance on detailed prior maps is a problem. Today’s autonomous vehicles are not able to drive in rural environments where we do not have maps — in other words, on the millions of miles of roads that are unpaved, unlit or unreliably marked.

MapLite is a first step for enabling self-driving cars to navigate on roads that they’ve never been on before using only GPS and sensors.

Our system combines GPS data – like the kind you’d find on Google Maps – with data taken from LIDAR sensors. Together, these two elements allow us to autonomously drive a car on multiple unpaved country roads and reliably detect the road more than 100 feet in advance.

MapLite is a first step toward creating safe and capable autonomous cars that can support drivers in new road situations. Imagine if cars could learn how we drive and how to never be responsible for a collision? What if they could become our trusted partners to help us navigate tricky roads, watch our backs when we’re tired, and even make our time in the car fun?

In the future, autonomous cars won’t just be able to sense the state of the road; they’ll be able to recognize the state of the driver. Imagine if your car could tell you were having a bad day and turn on your favorite album. Or imagine if it could talk to your fridge, figure out that you’re out of milk, and suggest where to stop on your way home. Imagine if your car knew that you forgot to call your parents yesterday and could issue a gentle reminder and suggest a safe stretch of highway where you could make the call. These are just a few of the possibilities when we bring together cars, computer science and artificial intelligence.

To read the full article, please visit Forbes

Interested in working together?

We’d love to hear from you.
contact us

Contact Us


    Thank you for your submission

    We will get back to you as soon as possible. Back to site