Autonomous cars are all the rage and perhaps the future of automobiles. However, if you’ve walked the streets of Silicon Valley, Pittsburgh, Boston or San Francisco, you may wonder what the large thing on top of these cars roof is. Well, we’re here to tell you – that is lidar, which is possibly the most important piece of hardware to for everybody to unlock self-driving cars.
Similar to radar in many ways, lidar doesn’t send out radio waves, but rather it emits pulses of infrared light which is basically lasers which are invisible to the human eye. This light measures how long the cars take to come back after nearly hitting objects.
After doing this million of times a second, the results are then compiled into its own cloud, which works like a 3-D map of the world. This map which works in real time is so detailed it can identify objects as well as spot them. Once they are identified, the cars computer will predict how they will behave, and then teach the car how to drive.
It does this million of times a second, then compiles the results into a so-called point cloud, which works like a 3-D map of the world in real time—a map so detailed it can be used not just to spot objects but to identify them. Once it can identify objects, the car’s computer can predict how they will behave, and thus how it should drive.
Unlike other self-driving cars which use other sensors to see, lidar offers hard data in the form of exact measurements, that it can even pick out things like arms and legs. Lidar allows for a full robot car, one that will not need a steering wheel or a human hand. The only problem? It’s too young for the rough outdoors.
It remains relatively undeveloped, and it’s costly starting at $4,000. The other issue the lidar faces is that it needs to be sturdy and reliable being able to withstand temperature extremes and continue to work for years. Therefore, lidar is still needed to get better before it can be used for good.