How LiDAR Fits Into the Future of Autonomous Driving
The tracking technology may play a huge role in AV development
The future of autonomous driving is not simple. The complexity of a self-driving vehicle is exactly why it’s taking so long to safely bring them into the world.
There are a ton of uses for autonomous vehicles that are being tested more and more in the real world. In the future, we will see robo-taxis with Uber and Lyft stickers all over the road. Amazon will deliver packages via a self-driving vehicle and semis delivering thousands of pounds of materials could also be driverless.
One of the most important parts of the autonomous vehicle will be its “eyes” — how it sees the world around it. For most companies, the vehicle’s eyes will be LiDAR.
LiDAR, which stands for “light detection and ranging,” pulses out laser waves to navigate the surrounding environment and map the distance of objects. LiDAR can track the distance to an object within a few centimeters of accuracy from up to 60 meters away.
Perhaps the largest — or at least most interesting — player in autonomous vehicles is Tesla. Tesla, though, happens to be against LiDAR, opting for cameras and radar instead.
An article done by Automotive World in August 2020 lays out the case for and against LiDAR.
One of the key strengths of LiDAR is the number of areas that show potential for improvement. These include solid-state sensors, which could reduce its cost tenfold, sensor range increases of up to 200m, and 4-dimensional LiDAR, which senses the velocity of an object as well as its position in 3-D space. However, despite these exciting advances, LiDAR is still hindered by a key factor; its significant cost.
Cost is the reason Elon Musk is against using LiDAR for Tesla, as lowering the cost of his vehicles is already a big enough problem as is.
LiDAR used to be much more expensive than it is today. Google’s first driverless car employed a $70,000 system, as Automotive World notes. But cost has come down significantly since, with LiDAR-maker Velodyne offering a unit for less than $500 that could be street-ready by 2022 or ’23.
While Musk and Tesla claim that cameras are a cheaper and more useful option, weather conditions such as rain and fog could cause issues. While cameras offer a more human-like look at the road, being able to better detect things such as the color of a traffic signal, they also require more computing and are therefore susceptible to malicious attacks.
The argument for cameras over LiDAR is still hard to debunk, though, as there are many factors to driving that a laser wave could not detect.
How would a LiDAR recognise that a pedestrian is looking down at his phone and may wander on the street? Can LiDAR differentiate between a plastic bag and an obstacle? Could LiDAR recognise a cyclist looking over his shoulder to join into a new lane? The answer to these questions is no. Once camera AVs have been perfected, they argue, LiDAR will be rendered obsolete. This is because by combining cameras with a simple radar (cheaper and better performing in adverse weather, although with worse image granularity than LiDAR), much is done to address their weakness in adverse conditions.
Tesla has been ahead of the curve in many areas and perhaps this is just the next one. It is hard to imagine an affordable LiDAR unit not having serviceability in an autonomous vehicle, though.
Autonomous vehicles still need a ton of fleshing out before being reliable and safe on crowded roads. Maybe the combination of cameras and sensors wins out, or maybe cameras and LiDAR come together to form a super unit of 3D mapping and detection.
Either way, LiDAR sensors are likely years away from becoming irrelevant, if they ever get there. As more money is poured into development, expect LiDAR to play a huge role in whatever autonomous vehicles may become in the future.