Mapping for Self Driving Cars

People seeking to develop autonomous vehicles are all interested collecting as much data as they can. A lot of this data comes in the form of mapping of roads. Autonomous vehicles are outfitted with multiple sensors and cameras that help the car to “see” its surroundings.  The input in these sensors is run through powerful computers that process the data and tell the cars how to drive. These computers sometimes make mistakes or don’t “see” something. But what if these computers could access all the input from a bunch of other vehicles are already have a map of what the road ahead is going to look like?

 

There are many companies out there right now trying to create these maps, making autonomous vehicles a closer reality. Companies are outfitting vehicles with LIDAR, RADAR, and cameras and driving across the world collecting data, creating a 3D mapping of the roads. “As humans, if we are blindfolded and dropped in a new place, we’ll find our bearings—we have millions of years of common sense to help guide our awareness,” says Nikhil Naikal, Mapper’s CEO. “A machine, on the other hand, needs a large amount of up-to-date 3D map data to have foresight of what to expect around the corner” (CNN).

Above is a video of Nokia’s mapped out version of New Orleans, using data from LIDAR sensors that were positioned on cars that drove around the city. Mapping out the roads can even help more than just creating a simple 3D of the regular roads. Many companies envision maps that are constantly being updated so unusual occurrences like construction or an object in the road can be detected, and one car can alert all other vehicles in the area. Driving during inclement weather, such as snow covering the road, will become more of a reality as the car will know where lane markers are within a centimeter and any important signs.

Though, these maps are not infallible. A small mistake in the mapping can lead to bigger mistakes by autonomous vehicles. A CNN article cites Ford making a mistake years ago where one pixel was wrong, saying the ground was raised 10 inches where is was flat. This caused all the autonomous vehicles to swerve where there was no reason too. Mistakes like this could lead to accidents in the real world.

Despite some of these faults, people in the industry see mapping full of potential and worth billions of dollars. Every road driven by cars will be building a clearer, more up-to-date map. Not only will mapping bring autonomous vehicles closer, it will enhance the technology thoroughly.

 

References:

https://www.wired.com/story/your-next-gig-map-the-streets-for-self-driving-cars/

http://money.cnn.com/2017/06/07/technology/business/maps-wars-self-driving-cars/index.html

https://www.engadget.com/2017/06/03/here-collection-vehicle-v3-hd-mapping-computex/

Autonomous Cars Will Require a Totally New Kind of Map

Leave a Reply