The future of autonomous navigation in vehicles, robotics, and UAVs will depend heavily on Light Detection and Ranging (LiDAR) technology.
LiDAR is used alongside other sensors such as cameras and RADAR for mapping, object detection and identification, and navigation without relying on lighting conditions.
LiDAR sensors are known to be expensive pieces of equipment, sometimes knocking in at $75,000 per unit, making them somewhat prohibitive and limiting for use en masse (and often, cost even more than the vehicles they’re mounted on!). However, solid-state LiDAR is gaining traction as a promising technology that is cheaper, faster, and provides higher resolution than traditional LiDAR, with predictions the price could eventually fall below $100 per unit.
More economical LiDAR options could accelerate the industries that rely on it, including autonomous vehicles, drone delivery services. It could also impact other areas such as virtual reality.
What is solid state LiDAR?
To explain the difference in technologies, traditional LiDAR systems are electromechanical—they rely on moving parts that have to be precise and accurate in order to obtain measurements suitable for autonomous navigation. These measurements come from photons from a laser, which then reflect back off surfaces and concentrated into a collector that can determine the distances of these objects.
Basically, RADAR but with light. The laser and collector must rotate in order to scan the area around it. The moving parts involved put a restriction on the size of the system, since making them small and compact would increase the difficulties in the precise manufacturing required, which then drives up cost.
Additionally, moving parts mean that it could be susceptible to perturbations, so driving in rough terrain for example could have negative impacts on readings. Despite the high cost, sensitivity to movement, and difficulty to manufacture, LiDAR Is highly useful for creating 3D maps of its surrounding.
Solid state LiDAR on the other hand is a system built entirely on a silicon chip. No moving parts are involved, which not only makes more resilient to vibrations, but can be made smaller much more easily. This lends to production being cheaper.
In order to achieve a scan, solid state LiDAR uses a concept not unlike phased array in radio. In phased array, several transmitters emit in certain patterns and phases to create a directional broadcast. The size, focus, and direction of this broadcast can be changed, without having to physically adjust the transmitters. The very same concept is applied in solid state LiDAR to achieve optical phased array — optical emitters send out bursts of photons in specific patterns and phases to create directional emission, of which the focus and size can be adjusted. Once again, no physical adjustment to the optical emitters need to be made to achieve this, making it resilient to vibrations and more compact. And for the design minded, solid state LiDAR systems so far appear to be more aesthetically appealing.
Optical Phased Array. Image courtesy of Analog Photonics.
In addition to optical phased array LiDAR, there is also Micro-ElectroMechanical System (MEMS) based LiDAR, which uses micro-mirrors to directionally control emission and focus. MEMS LiDAR is the technology behind the solid state LiDAR devices that Alibaba plans to use in partnership with RoboSense in autonomous delivery vehicles dubbed the “G Plus”.
Then, there is Innoviz Technologies, an Israeli startup founded in 2016 that is working at break neck speed to beat competitors to the punch in the autonomous driving industry. Currently, they have a partnership with BMW to equip their vehicles with their InnovizOne system in models due for release in 2019.