According to the James Consulting, since the 1960s, the laser pointer (hereafter referred to as LiDAR) has become a mature technology. Early applications include mapping for architectural and archaeological purposes, and NASA’s ninth manned mission to the moon, Apollo 15, also used LiDAR technology to map the moon’s surface. But until the beginning of the millennium, LiDAR caught the attention of the automotive industry. LiDAR has the potential for continuous and high-precision 3D scanning and is highly attractive to manufacturers involved in the development of early self-driving cars.
Since then, the laser pointer technology has been improved, and the top-mounted LiDAR has become the regular “look” of companies such as Waymo for automated driving tests on public roads. But the LiDAR system has not yet entered mass production, with a single unit costing $75,000. The cost issue means that the debate about whether LiDAR can have a place in the automotive industry will continue. An important voice of this technology is the departure of Tesla’s CEO, Elon Musk, at a press conference in 2015. He believes that LiDAR is not a necessary technology and continues to claim that Tesla can be in two years in 2017. L5-class autopilot is achieved through a combination of cheaper sensors such as radar and cameras.
It is foreseeable that Tesla’s CEO’s comments have already caused strong opposition from some people, including General Motors (GM) self-driving car integration director Scott Miller said in late 2017 that “Tesla CEO for autonomous driving The saying is ‘a nonsense’.”
Jason Eichenholz, co-founder and CTO of Luminar Technologies, a LiDAR technology development company, said the debate on LiDAR technology will come to an end. Eichenholz emphasizes that “real autopilot absolutely requires LiDAR technology” and Luminar’s LiDAR technology is confirmed to appear on the Toyota platform version 2.1 and version 3.0 autopilot test vehicles. Eichenholz mentioned, “Once autonomous cars start to get out of closed roads and enter real roads, they must have ‘5 9’ (99.999%) reliability. LiDAR technology is therefore essential. Camera and radar surface For the boundary conditions of realistic driving scenes, the lack of performance required to respond in time, the scene around the car needs to be able to observe in 3D mode like the human eye.”
Eichenholz’s stated boundary conditions refer to accidents that may occur spontaneously on the road, such as a child coming out of between two parked cars, a roadway closed down, and a semi-trailer painted in front of the driver’s sky The same color, 2D scanning device may not be able to interpret the above situation correctly. Adaptability makes humans well suited to deal with these accidents. Machines cannot be generalized.
Eichenholz continued, “If a car is parked on the side of the road, is it parked in a three-foot driveway? If it is driving, we can see and make adjustments. The high-fidelity vision provided by LiDAR technology means self-driving. Cars can also understand the world around them and make decisions in advance.”