Lumotive’s Meta-LiDAR Takes On LiDAR Scalability Challenges Using Metasurfaces
One major challenge to LiDAR technology has always been: scaling. Striving to take this challenge head-on is Lumotive with the industry's "smallest and most cost-effective" solution yet.
One of the biggest shortcomings of LiDAR technology for use in many applications has been its size and cost. While costs have dropped dramatically over the past decade, they are still non-trivial - resulting in continued research and investment into LiDAR scaling.
A high-level overview of an autonomous driving system, including LiDAR. Image used courtesy of the IEEE Computer Society
One company believes to have found a solution is Lumotive, a Seattle-based company working on solid-state LiDAR solutions. Recently, the company made news when it announced its new Meta-Lidar Platform, which claims its solution is the industry's smallest and most cost-effective solution.
In this article, let's go over some challenges scaling LiDAR and how Lumotive approached these challenges in its new platform.
One of the most significant roadblocks to scaling in LiDAR is how, to provide useful information, LiDAR requires the repeated capture of an entire 3D space, not just a single point space.
The result of this is that many automotive makers use beam-scanning techniques, where the LiDAR laser pulses are directed to map a wide field of view. If you've ever seen the big spinning thing on the top of an autonomous vehicle, you're looking at beam-scanning LiDAR.
Example of a beam-scanning LiDAR. Image used courtesy of Renishaw
This technology has historically been a challenge because densely, accurately, and reliably performing beam-scanning requires physically directing the laser beams.
Often, the solutions lie within a mechanical field, which can add complexity, size, wear, inaccuracy, and mechanical failure. Newer generations of LiDAR use MEMS mirrors or optical phased arrays, which offer improvements in size and reliability, but both suffer from a lack of performance.
For broad adoption of LiDAR requires solutions that are non-mechanical, solid-state, cheap, and accurate. One answer that the company Lumotive is using is "metasurfaces."
Metasurfaces for LiDAR Beam-steering
With this in mind, startup Lumotive came out in 2018 with an entirely new approach to beam-steering. Its approach utilizes a proprietary technology called Liquid Crystal Metasurfaces, which it claims can deliver an unprecedented combination of performance, cost, and size.
Metasurfaces are a novel type of optical device based on 2D arrays of sub-wavelength optical elements. These arrays can either transmit or reflect incident light to focus and steer it, along with other forms of wave manipulation.
Radiation patterns from an LCM chip. Image used courtesy of Gleb M. Akselrod
Based on the same principle, LCMs are reflective 1D metasurface that leverage liquid-crystal tuning of resonant sub-wavelength elements. In this scheme, the phase delay at each resonator is proportional to the voltage applied to it, which rotates the liquid crystals locally, shifting the resonant wavelength and enabling a 120° field of view and uniform efficiency across the field of view.
This technique could be beneficial for many reasons. First off, it is not mechanical and has low power. Beyond this, it can be manufactured using a standard CMOS semiconductor process, making it cheap and scalable.
Now that a bit more of Lumotive's proprietary technology is understood, like its use of metasurfaces, let's look at its "Meta-LiDAR."
As mentioned, Lumotive released its Meta-LiDAR platform recently, one that it's calling the industry’s "smallest and most cost-effective" 3D sensing solution.
Lumotive's M30 LiDAR sensor. Image used courtesy of Lumotive
Based on LCM technology and manufactured using a standard CMOS process, the platform hopes to be in full commercial production by the end of 2022.
Luimotive calls this platform the industry's first software-defined lidar, saying that it will allow the lidar scan pattern, frame rate, and resolution to be customized for specific use cases.
The first product in the platform, the M30, offers a 10-20 meter range, 120° x 90° FOV (field of view), and a 640 x 480 resolution in the form factor about the size of a golf ball. Later products in the platform hope to reach ranges up to 200 meters and be used in applications such as AR and robotics.
While Lumotive's ultimate goal is to enable the widespread adoption of LiDAR by creating compact, low-power, and mass-producible solutions at competitive prices, its LCM technology seems to be a good step in this direction, and the Meta-LiDAR platform could become a pivotal point in pushing LiDAR's scalability forward.
Interested in other LiDAR advancements? Learn more in the articles down below.