Has Elon Musk turned around on LiDAR?
On May 20, 2021, a Tesla Model Y with LiDAR sensors on the roof was spotted by autonomous vehicle industry consultant Grayson Brulte in Palm Beach, Florida. People have found this rather interesting given Elon Musk, CEO of Tesla, and his historically less than favorable view of laser sensor technology.
LiDAR, which is short for light sensing and ranging, is a laser sensor that most tech and automotive companies deem essential for the development of self-driving cars. LiDAR uses lasers to create a three-dimensional image of the surrounding landscape. This image is necessary because it tells us what the autonomous vehicle would perceive of the environment, predicts the behavior of pedestrians and other cars, and helps navigate the roads safely. Musk, however, didn’t like LiDAR too much. In 2019, he called LiDAR a “ dumb race ” and said anyone who relied on it “ is doomed. ” Instead, Musk says Tesla would bank on cameras as a key hardware component for the development of autonomous vehicles. According to Tesla, camera sensors – especially with their increasing pixel resolution and lower prices – are a much better alternative to LiDAR.
Given that, the image of one of Musk’s cars equipped with LiDAR technology surprised Brulte. What followed was his tweet on the same, which sparked a series of rumors that Musk might turn around from his previous opinion on laser sensor technology. To add additional elements of surprise, it was reported that Tesla has partnered with Luminar to use its LiDAR sensors for “test and development” purposes. Neither company has yet spoken about what the partnership would mean.
Before examining Tesla’s plans, let’s dissect Elon Musk’s argument: how necessary is LiDAR to create autonomous vehicles?
LiDAR has become a standard device on a multitude of self-driving cars such as those operated by GM and Alphabet’s Waymo. In addition, many people today believe that autonomous driving will not work without LiDAR. Sanjeev Sharma – the founder of Swaayat Robots – pointed out that the choice of LiDAR and camera-based perception systems is essentially a compromise between precision and heavy computational load. Sharma also noted that autonomous vehicles require perception, planning and location – and most non-LiDAR systems are unable to work on all three fronts.
Blickfeld, a producer of LiDAR, said in a blog Publish, that relying only on cameras and radar, but not LiDAR, is not safe. Cameras, while excellent at capturing color images (which LiDAR doesn’t), are lacking, according to Blickfeld, when it comes to recording the third dimension, which is essential for measuring distances and avoiding objects. For this reason, cameras require image recognition software to interpret the recorded 2D information in order to estimate distances. The problem with this software is that it uses artificial intelligence and machine learning to tag and store “lived situations”. This procedure involves thousands of kilometers, real and simulated, but can cause problems for a vehicle entering unfamiliar territory. Recording every situation that is not part of routine driving on an algorithm is almost impossible. However, LiDAR technologies can get away with the daunting task of recording all this data. LiDAR directly captures 3D data and therefore does not need to convert 2D recordings to 3D. Thus, if the vehicle encounters an obstacle, the LiDAR sensors can reliably detect it at an early stage and identify its exact dimensions and location from said vehicle.
Problem with LiDAR
One major complaint that people generally have with LiDAR is its cost. The cost of placing a single LiDAR can go up to $ 10,000, and importing it into India can cost you around one crore of rupees. Companies like Google’s Waymo have been able to reduce this through mass production, which still doesn’t cut costs significantly and isn’t achievable for everyone.
While great for spotting things on the road, LiDAR cannot distinguish a potentially dangerous roadblock apart from, say, a plastic bag – for which you don’t need to stop a vehicle. . Problems accumulate more when moving objects such as pedestrians, stray dogs and flying plastic bags appear on the road. LiDAR will not be able to detect how these objects move and what these objects are.
At their 2019 Autonomy Day event, Tesla assured people that their cars use a system of cameras and radar that will be able to detect what an object is. The vision here will then involve the radar alert that something potentially dangerous may be ahead, after which the camera decides – upon seeing the thing – how the car should react. It is important to note that Tesla has now given up radar in its cars and will only use eight cameras in the sensor suite of its best-selling Model 3 and Model Y cars. This is believed to be because Musk found radar sensors relatively expensive and said he believed “a vision-only system is ultimately all that is needed for full battery life.”
It is important to realize, however, that Tesla uses its own vision technology. According to Sanjeev Sharma, LiDAR systems are more accurate than camera-based ones, which require custom hardware like Tesla’s neural network accelerators. He says a more efficient solution would be a combination of LiDAR and cameras. For example, 1 or 2 LiDARs with eight cameras could cover both the cost of the solutions and the IT aspects. Regardless of that, Sanjeev Sharma himself found a way to work with LiDAR and its huge cost. His company – Swaayatt Robots – has enabled autonomous vehicles to perceive their environment using cameras only in real time. According to him, its technology is 40 times faster than the state of the art. He says, however, that how to use cameras is more a matter of mathematical effort.
The musk of everything
Despite the arguments for and against adopting LiDAR, one thing is certain: Musk hates LiDAR. So why is Tesla signing a deal with Luminar? It is not yet clear. In an interesting twist, Elon Musk seemed to go more smoothly in his disdain for LiDAR. In a Clubhouse chat room, he said he was “ talking a lot ” about LiDAR and that his rocket company, SpaceX, was developing an in-house LiDAR to help the SpaceX Dragon Capsule. He argues, however, that vision is more important for driving on real world roads. That being said, it seems highly unlikely that Musk suddenly decided to change gears and join the LiDAR train. Guidehouse Insight analyst Sam Abuelsamid says Tesla is more likely to use Luminar’s LiDARs as a benchmark for testing and validating its camera-based systems. This seems to be the most likely explanation for Brulte’s photo and the tweet that followed, and the only one we can hang on to until Tesla or Luminar explain further.
Join our Telegram group. Be part of an engaging online community. Join here.
Subscribe to our newsletter
Receive the latest updates and relevant offers by sharing your email.