- Voyant Photonics has raised funding to bring their smaller, less expensive, and more easily manufactured lidar to market.
- The FMCW method, which is more typically used in radar, enables this small size.
- This method overcomes various limitations associated with conventional lidar systems by continuously adjusting the frequency of a continuous beam of light encoded with identifiable data patterns.
The future of lidar is questionable until its price and size are lowered to fractions of their existing values, as Voyant wants to accomplish. As long as lidars are sandwich-sized devices costing thousands of dollars, they will remain uncommon — which is why Voyant has gathered funds to bring their smaller, less expensive, more easily made, yet still highly capable lidar to market.
Post the company’s seed financing in 2019, the goal was to use silicon photonics to compress lidar from sandwich to fingertip size. However, the actual problem for practically every lidar company is lowering the price. Between a powerful laser, a capable receptor, and a mechanical or optical method of directing the beam, it’s simply not possible to create anything inexpensive enough that, like an LED or touchscreen, several of them can be easily integrated into a vehicle costing less than $30,000.
Peter Stern joined the company back in 2020. They were looking for a method to turn a promising prototype produced by co-founders Chris Phare and Steven Miller into a viable product. After reverting to the fundamentals, they developed a photonics-based frequency-modulated continuous-wave (FMCW) system (for the time being, call it that) that could be built at current commercial fabrication facilities.
He stated that while every other system is crammed with pricey components, their vision is a mass-producible chip, similar to anything else. He cited the absence of an intense precision laser as a high cost and space reduction. What individuals use as a laser source is frequently expensive, requires installation and calibration, and has lens difficulties. Their laser sources are essentially obsolete datacom lasers the size of sesame seeds that have been gently reconditioned. These items are around $5 each, and the laser path is about $30.
How does Voyant operate?
This minuscule size is enabled by the FMCW approach, which is more frequently employed in radar. This technology eliminates numerous difficulties associated with standard lidar systems by using a continuous beam of light encoded with identifiable data patterns and constantly altering its frequency. And Voyant’s method is inexpensive – it’s conceivable to get under a hundred dollars with scale. The optics, beam handling, and sense, for example, are all integrated into the chip.
However, they are not taking on Velodyne or other lidar companies vying for market shares in the automotive sector, such as Luminar and Baraja. Stern stated that they are insufficiently funded to take it through the automobile development cycle — and it is a costly industry to enter. Due to their lower cost, they find uses in robotics, mobility, and industrial safety. Anyone who wants to utilize a Velodyne puck can easily displace them for non-automotive purposes.
You may be thinking, wait a minute, you already have lidar on your phone – what makes this different? While it is possible to manufacture lidar systems of this scale and size, their capabilities are severely constrained. Excellent for scanning your living room, but unreliable beyond a few meters, especially in bright or foggy conditions. While Voyant is not targeting automobiles, its devices retain automotive-grade specifications: millimeter accuracy up to a hundred meters, which is precisely what you want when traveling at 70 miles per hour.
The FMCW approach (also used in Aeva’s lidars) generates fewer points and thus lesser resolution, but it gives instantaneous Doppler velocity. A potentially significant advantage is knowing how fast the object your beam struck travels without requiring additional scanning power or calculation.
Another intriguing advantage the device has over the competitors is its capacity to distinguish not just distance and velocity but also, to a degree, material. This is accomplished by using polarization, a property of the beam that is altered differently by different surfaces. Thus, based on a single data point, Voyant’s gadgets should be able to determine whether anything is made of metal, asphalt, wood, flesh, clothing, or fur. That is astoundingly handy for categorizing objects – if something has a coat on it, it is almost certainly not a tree or a car, correct?
Partners led the investment, including LDV Capital and Contour Ventures participation. The startup intends to use the funds to advance manufacturing by distributing development kits to partners. The “Lark” unit is the more typical of the two, bouncing the laser signal off a galvo mirror. In contrast, the “Sparrow” unit employs a two-dimensional beam steering technology that further minimizes mechanical components.
Stern stated that they will produce approximately 200 units for partners in 2022 and begin taking retail orders in 2023. By that time, the car industry may have taken notice. Still, if Voyant’s plan succeeds, it will have snatched a sizable portion of the industrial market away from companies that manufacture more significant, more expensive units.