For years, FMCW LiDAR has been held up as the “better” sensing technology—longer range, native velocity measurement, and immunity to sunlight that plagues traditional time-of-flight systems. Yet despite its technical promise, FMCW LiDAR has struggled to escape labs and prototypes. The reason hasn’t been physics. It’s been system complexity and cost.
That may be about to change.
LightIC Technologies and indie have announced a strategic partnership and the completion of a reference design that replaces FPGA-based processing with indie’s iND83301 system-on-chip (SoC) inside LightIC’s silicon photonics FMCW LiDAR architecture. The result is a platform that dramatically improves size, weight, power, and cost—often summarized as SWaP-C—marking a key step toward real-world commercialization.
According to the companies, the integrated solution delivers an 80% reduction in power consumption, a 40% reduction in overall size, and a substantially lower cost profile compared with existing FPGA-based FMCW LiDAR systems. That combination directly addresses the barriers that have kept FMCW LiDAR from scaling into automotive production and mass-market robotics.
Why FPGA-Based FMCW LiDAR Hit a Wall
Early FMCW LiDAR systems leaned heavily on FPGAs for signal processing because flexibility mattered more than efficiency. That tradeoff made sense during development—but it became a liability when the industry tried to move toward production.
FPGAs are expensive, power-hungry, and physically large. In automotive environments, they complicate thermal management and drive up system cost. In robotics, they make compact, battery-powered designs impractical.
By integrating indie’s iND83301 SoC, LightIC has effectively crossed the line from experimental architecture to application-specific, production-ready silicon—a transition that mirrors how cameras, radar, and ADAS compute stacks matured over the last decade.
This is the inflection point FMCW LiDAR has been waiting for.
A SoC Built for LiDAR, Not Repurposed for It
The iND83301 is a dedicated automotive-grade SoC designed specifically for LiDAR signal processing. Unlike FPGA solutions that rely on programmable logic, the SoC delivers deterministic performance with far lower power draw and a much smaller footprint.
When paired with LightIC’s highly integrated silicon photonics FMCW platform, the result is a tightly optimized hardware stack—one that removes excess components, reduces interconnect complexity, and simplifies system integration for OEMs.
LightIC CEO Jie Sun framed the achievement bluntly: FMCW LiDAR’s advantages have long been recognized, but adoption stalled because system-level complexity made it hard to deploy at scale. This integration, he argues, resets FMCW LiDAR’s economics and readiness for global deployment.
Long-Range FMCW LiDAR for Automotive
The first tangible outcome of the partnership is Lark™, LightIC’s long-range FMCW LiDAR product.
Built on the combined LightIC–indie platform, Lark delivers perception beyond 500 meters, placing it firmly in the long-range sensing category required for highway-speed autonomy and advanced driver assistance systems.
Key attributes include:
- Extended range exceeding 500 meters
- Native, point-by-point velocity measurement
- Robust performance in bright sunlight and challenging lighting
- A SWaP-C profile suitable for automotive production
For automotive OEMs, this matters because FMCW LiDAR can distinguish between stationary and moving objects at the sensor level, reducing reliance on sensor fusion to infer velocity. That capability is particularly valuable for highway scenarios, cut-ins, and long-range object tracking—areas where traditional time-of-flight LiDAR often struggles.
Just as importantly, the move away from FPGA-based designs brings Lark closer to the cost, power, and reliability targets required for global vehicle platforms.
Shrinking FMCW LiDAR for Robotics and Physical AI
Automotive isn’t the only market in focus.
Using the same optimized architecture, LightIC has also introduced FR60™, which it describes as the smallest known FMCW LiDAR, with a form factor roughly the size of a tennis ball.
That size reduction—enabled directly by the iND83301 SoC and the improved SWaP-C profile—opens doors in markets where FMCW LiDAR was previously impractical:
- Robotics and autonomous mobile robots (AMRs)
- Industrial automation
- Humanoid robots
- Emerging physical AI systems
In these domains, power efficiency and compactness are often more important than raw range. FR60 delivers high-resolution 4D perception—range, velocity, and spatial detail—in a package small and efficient enough for battery-powered platforms.
This is especially relevant as humanoid and embodied AI systems move from research labs toward commercial deployment, where perception sensors must be compact, robust, and affordable.
Why Silicon Photonics Matters Here
LightIC’s underlying advantage lies in its silicon photonics FMCW architecture, which integrates optical components directly onto silicon. Compared to discrete optical assemblies, silicon photonics offers:
- Higher integration density
- Better manufacturability at scale
- Improved thermal and mechanical stability
- A clearer path to automotive-grade reliability
When combined with a dedicated LiDAR SoC instead of an FPGA, the architecture starts to look less like a science project and more like a production-ready sensor platform.
That distinction matters to OEMs and Tier 1 suppliers, who are increasingly skeptical of LiDAR solutions that promise performance but struggle with cost, yield, or long-term supply chain viability.
Competitive Context: FMCW Is Heating Up
The LiDAR market remains crowded, with dozens of companies competing across time-of-flight, flash, and FMCW approaches. While time-of-flight LiDAR dominates today’s production vehicles, FMCW has long been viewed as the next step—if it could be simplified.
Several players have made progress on FMCW, but many still rely on complex, power-intensive processing architectures. By delivering a reference design that replaces FPGA logic with a purpose-built SoC, LightIC and indie raise the bar for what “production-ready” FMCW LiDAR means.
It also puts pressure on competitors to demonstrate similar gains in power efficiency, size, and cost—especially as OEMs evaluate sensor roadmaps for late-decade vehicle platforms.
A Roadmap Beyond Automotive
While automotive remains the largest prize, the broader implication of this partnership is scalability.
The same core architecture supports:
- Long-range automotive LiDAR
- Compact robotics sensors
- Industrial and physical AI applications
That flexibility strengthens the business case for FMCW LiDAR silicon investment, allowing vendors to amortize development across multiple markets rather than betting everything on automotive adoption timelines.
For indie, the partnership showcases how its automotive-focused SoC portfolio can extend into adjacent autonomy and robotics markets. For LightIC, it validates silicon photonics FMCW LiDAR as a platform—not just a point solution.
The Bottom Line
FMCW LiDAR has never lacked technical advantages. What it lacked was a system architecture that could meet real-world constraints on power, size, and cost.
By integrating indie’s iND83301 SoC into LightIC’s silicon photonics FMCW LiDAR, the two companies have removed one of the biggest blockers to commercialization. An 80% power reduction and a 40% size reduction are not incremental gains—they fundamentally change where and how FMCW LiDAR can be deployed.
If this platform scales as promised, it could mark the moment FMCW LiDAR finally moves from promise to production—across cars, robots, and the next generation of physical AI systems.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
