In contemporary time, LIDAR is paired with cameras and radar. The significant autonomous cars developers use this technology. Tesla is the only significant company that uses only cameras and radar and not LIDAR.
What is LIDAR?
So what basically is LIDAR? It is a laser radar. This gives it a low-resolution black-and-red view of the environment with the exact distance between each pixel. It has superhuman vision in one sense, albeit the relatively poor resolution, present high cost, and it has limited range, all this suggests that it isn’t exceptional in every manner.
Elon Musk Rebuttal of LIDAR
First point of Musk’s that LIDAR is superfluous. Elon Musk claims that humans can drive without lid using vision and a neural network, therefore cameras and an artificial neural network must be capable to do this. Furthermore, Lidar alone is insufficient.
Lidar is unable to distinguish between signs like “STOP” or “SLOW,” also require cameras for image recognition. Lidar lacks the resolution of light waves, it is vulnerable to special types of error, and is ineffective in smoke or heavy rain. Musk adds that radar is different because it works in a different frequency band than cameras. Only Radar can help to see through things that block light
”LIDAR is a fool’s errand. Anyone who relies on LIDAR is doomed. Expensive sensors that are unnecessary,” Elon Musk.
The cost of LIDAR is the most obvious reason for Tesla to choose a different path. Tesla is quite concerned with prices and maintaining that the cars are affordable, while LIDAR is an expensive technology.
Self-driving idea can’t act out without having a solution for vision problem. Once there will be solution for this problem, Lidars won’t be required. Elon considers Lidar as a crutch to solve vision problem, which will end.
Also See: Tesla Autopilot is being formally investigated by the US
When there is a conflict between vision and LIDAR, decision will always be in favor of vision as all LIDAR can provide is confirmation of your vision interpretation or possibly a clue about how to perceive a vision. LIDAR requires an improved vision interpretation consequently. There would be nothing LIDAR can add.
Tesla in All
Tesla is making a big bet that neural networks will be able to address all of the challenges with self-driving cars. Elon Musk believes that neural networks are the only practical solution to the vision problem, and that all other technologies, such as LIDAR, are a waste of money and distraction.
[…] Also See: Why Tesla is Not Considering LIDAR a Solution? […]