LiDAR Autonomous Vehicle Patents

LiDAR patents cover ranging-method (time-of-flight vs. FMCW) innovations; beam-steering architecture innovations; laser-source and wavelength innovations; and detector and point-cloud innovations — with IP held by long-range and solid-state LiDAR firms, the foundational rotating-LiDAR pioneer, and coherent-LiDAR specialists. MAJOR LIDAR PATENT HOLDERS: LUMINAR (large estate): 1550 nm fiber-laser LiDAR with InGaAs detection for long range (250 m+) and eye-safe high power, the Iris sensor, and per-point processing. VELODYNE (foundational rotating-LiDAR estate): the original 360° mechanical spinning multi-beam LiDAR — Velodyne's foundational patents underpinned the field and were asserted in major litigation (against Quanergy, Hesai, Ouster, and others). OUSTER (digital LiDAR): VCSEL emitter arrays + SPAD/CMOS receiver ('digital LiDAR' on a chip), simplifying the optomechanical design (Ouster merged with Velodyne). AEVA: FMCW (frequency-modulated continuous-wave) '4D' coherent LiDAR measuring instantaneous per-point velocity and offering interference immunity, on a silicon-photonics chip. OTHERS: Innoviz (MEMS-micromirror solid-state), Cepton (MMT frictionless micro-motion), AEye (agile MEMS, software-definable scan), Hesai and RoboSense (high-volume automotive), Valeo (Scala, the first automotive-grade LiDAR), and Waymo (in-house Laser Bear). Beam-steering architecture and FMCW vs. ToF ranging are the most contested LiDAR IP axes.

Time-of-flight ranging innovations; FMCW coherent-detection innovations; mechanical, MEMS, and solid-state beam-steering innovations; and scan-pattern innovations represent core LiDAR patent domains — and beam steering is the architecture choice that most defines a LiDAR company. RANGING-METHOD PATENTS: direct time-of-flight ToF (pulsed laser, measure round-trip time — the dominant approach), histogramming and multi-echo for SPAD systems, and FMCW (chirped continuous-wave coherent detection that yields range AND instantaneous radial velocity from the Doppler beat, with strong immunity to sunlight and other LiDARs — Aeva, Aurora, SiLC); also amplitude-modulated continuous-wave AMCW. BEAM-STEERING PATENTS: mechanical spinning (Velodyne foundational), MEMS micromirror scanning (Innoviz, AEye), optical phased array OPA (all-solid-state electronic beam steering, no moving parts — Quanergy's bet), flash LiDAR (illuminate the whole scene, no scanning), risley-prism and polygon scanners, and liquid-crystal/metasurface steering. SCAN-PATTERN PATENTS: software-definable and foveated/adaptive scanning (dwell longer on regions of interest), gaze allocation, and frame-rate/resolution trade-offs. The combination of ranging method (ToF vs. FMCW) and beam-steering architecture (mechanical, MEMS, OPA, flash) is the core differentiator and the densest patent terrain.

Laser-source and wavelength innovations; photodetector innovations; optical-integration and silicon-photonics innovations; and point-cloud-processing innovations represent additional LiDAR patent domains. LASER / WAVELENGTH PATENTS: 905 nm (silicon-detectable, low-cost, but eye-safety-power-limited) versus 1550 nm (fiber lasers, InGaAs detection, eye-safe at higher power for longer range — Luminar's choice), VCSEL emitter arrays (Ouster), edge-emitting and fiber lasers, and pulse-shaping. DETECTOR PATENTS: avalanche photodiode APD, single-photon avalanche diode SPAD and SPAD arrays, silicon photomultiplier SiPM, and CMOS-integrated receivers with on-chip time-to-digital conversion and histogramming. OPTICAL-INTEGRATION PATENTS: silicon-photonics integration of coherent-LiDAR transmit/receive (for FMCW chips), optical antennas/gratings, on-chip OPA, and miniaturization toward a chip-scale LiDAR. POINT-CLOUD / PERCEPTION PATENTS: point-cloud denoising, sun/interference rejection, object detection and classification from point clouds, sensor fusion with camera/radar, and calibration — these perception/algorithm claims face §101 scrutiny and are often claimed with the specific sensor. Detector technology (SPAD arrays) and silicon-photonics coherent integration are the highest-value emerging hardware IP as the field pushes toward chip-scale, automotive-grade, low-cost LiDAR.

LiDAR startup IP strategy must navigate Velodyne's foundational rotating-LiDAR patents (broadly asserted in litigation), Luminar 1550 nm long-range patents, Ouster digital-LiDAR patents, Aeva/Aurora FMCW coherent patents, automotive-grade qualification requirements, and a §101 constraint on perception algorithms; understand that mechanical-spinning LiDAR is heavily patented and litigated (Velodyne sued widely), that the durable IP lies in your specific beam-steering architecture, ranging method (especially FMCW), detector, and silicon-photonics integration, and that automotive qualification (cost, reliability, AEC-Q) is as decisive as IP; identify whitespace in chip-scale solid-state (OPA, flash), FMCW coherent integration, SPAD-array receivers, and low-cost automotive-grade designs. LIDAR STARTUP IP STRATEGY: MECHANICAL-SPINNING IS A LITIGATED THICKET — SOLID-STATE AND FMCW ARE THE IP: Velodyne's foundational rotating-LiDAR patents were asserted against many entrants — avoid that terrain and patent your specific MEMS/OPA/flash beam steering or FMCW coherent method; FMCW COHERENT + SILICON-PHOTONICS INTEGRATION ARE HIGHEST-VALUE WHITESPACE: per-point velocity, interference immunity, and chip-scale integration (Aeva-style) are the frontier and least-consolidated; SPAD-ARRAY DETECTORS AND 1550nm LONG-RANGE ARE ACTIVE TERRAIN: CMOS SPAD receivers and eye-safe 1550 nm fiber-laser long range are key differentiators; PERCEPTION ALGORITHMS MUST BE TIED TO THE SENSOR (§101): claim point-cloud processing with the specific hardware, not as a bare method; AUTOMOTIVE QUALIFICATION IS A PARALLEL MOAT: AEC-Q reliability, cost-down, and OEM design-wins gate the market alongside IP; WHEN TO PATENT: NOVEL SENSOR WITH MEASURED PERFORMANCE: file once a sensor shows measured results (range m at 10% reflectivity + resolution points/sec + FOV + velocity precision for FMCW + cost) vs. mechanical/Luminar/Aeva baselines — measured range, resolution, field of view, velocity accuracy, and unit cost are the critical LiDAR IP metrics; KEY FTO CHECKLIST: Velodyne foundational 360° mechanical rotating multi-beam (litigated); Luminar 1550nm InGaAs fiber-laser long-range; Ouster VCSEL+SPAD digital CMOS; Aeva/Aurora/SiLC FMCW coherent silicon-photonics 4D velocity; Innoviz/AEye MEMS micromirror agile scan; Quanergy OPA optical phased array; flash LiDAR; risley-prism/polygon; 905 vs 1550nm; APD/SPAD/SiPM TDC histogramming; perception §101-tied-to-sensor; AEC-Q automotive qualification.