Watch How A Driverless Car 'Sees' The Road
One consistent worry about the future of autonomous driving is how reliable the option will be. Civil Maps, one of the many companies behind the process of making it work, hopes to ease minds by explaining how the self-driving car reads the road.
Updated May 18 2019, 5:57 a.m. ET
Self-driving cars remain one of the most exciting features coming down the pipeline, but many are still skeptical of the technology. That’s because they either don’t understand how the process works or they’re worried about potential failures that could happen with it. Ford and Civil Maps are hoping to build that trust by explaining how it all happens.
Videos are available online from Civil Maps that go into further detail. One shows how self-driving technology has “six degrees of freedom” that gives it the power to learn their surroundings and quickly process that to make decisions: roll, pitch, yaw, x-axis, y-axis, and z-axis. Those are all the ways the vehicle can tilt and directly move.
The video goes through what happens when a car is making a left turn at a busy intersection. It can learn the boundaries of the road just like we can, understanding how many lanes are available and when it can’t go left or right any further. They can also dissect the objects relevant to driving, such as other vehicles, street signs, and pedestrians. All of these can be seen clearer than the rest of the environment and it will trigger actions to the car.
People crossing the road are located through heat sensors in the vehicle, and the car prepares for it by understanding signs that warn for pedestrians and crossroads. Not only will signs trigger expected actions, but also the changing in street lights. It can process information through importance, which actually gives it an advantage over the human mind. In a busy intersection, human drivers often make a wrong or hasty decision due to impatience. Automation would know what to do in any situation.
Similar to how a GPS system works, this is all a little easier for the self-driving car because it knows its destination through 3D maps that have LiDAR (Light Detection and Ranging) and radar technology. It fills in the gaps that cars can’t already see in their sights, such as faded road markings, missing street signs, and obstructed objects. That’s where Civil Maps shows its value and is hoping it can sell their services to car manufacturers that are installing automation in their vehicles.
The Ford Motor Company already invested in the startup last July. A spokesperson said at the time that “working with Civil Maps gives us an additional way to develop 3D high-resolution maps, which will bring fully autonomous Ford vehicles a step closer to reality for consumers.” Just in February, Ford announced it would invest in Argo AI over the next five years. While Civil Maps creates the layout for the cars to understand, this other company creates the software behind the vehicle’s decision making.
All these investments have made Ford the leader in the push for autonomous driving. The company expects to have these available in 2021, but don’t expect fully automatic editions just yet. They would at “level four,” meaning that the vehicles would be able to run by themselves in areas on maps, but would require manual driving should they start going off the grid or there is weather impacting the journey.