Jovaxis
LiDAR reference
2026-06-27 · 14 min

LiDAR Technology Families: Mechanical Spinning, Solid-State, Semi-Solid-State, FMCW, and Hemispherical

A comprehensive technical guide to LiDAR architectures — from mechanical spinning to FMCW — covering operating principles, advantages, limitations, and a comparison table.

Introduction: Why Do Different LiDAR Technologies Exist?

LiDAR has become an essential component of autonomous perception for robotics, self-driving vehicles, mapping and Industry 4.0. No single architecture meets every need: depending on the application, requirements for FoV, range, resolution, robustness and cost vary significantly. Five major technology families have emerged: mechanical spinning, solid-state (flash and OPA), semi-solid-state MEMS, FMCW and hemispherical.

1. Mechanical Spinning LiDAR

Mechanical spinning is the historic technology. A complete optical module (lasers, photodetectors) is mounted on a rotor spinning at constant speed, providing 360° horizontal FoV. Velodyne (now Ouster), Hesai and Ouster are the leaders. Advantages: 360° FoV, high point density, maximum maturity. Limits: moving parts (wear), bulk, high cost ($1,000-8,000). Range: 200-300 m. Typical lifespan is 5,000 to 10,000 hours.

2. Solid-State LiDAR (Flash and OPA)

Solid-state LiDARs eliminate all moving parts. Flash LiDAR illuminates the scene with a single pulse (VCSEL) captured by a SPAD/CMOS sensor. Optical Phased Array (OPA) uses a nano-antenna array on a photonic chip to steer the beam electronically. Advantages: no moving parts, compactness, very low cost potential. Limits: limited FoV (60-120° flash, 30-60° OPA), shorter range (50-150 m). Maturity: TRL 7-8 for flash, TRL 4-6 for OPA.

3. Semi-Solid-State / MEMS LiDAR

MEMS LiDAR retains a single tiny moving element: a micro-machined mirror a few mm in diameter, electrostatically actuated. Livox (non-repeating Lissajous pattern), Innoviz (InnovizOne/Two, adopted by BMW) and RoboSense (M1, global leader) are key players. Advantages: compact, vibration-resistant, 150-300 m range, moderate cost ($300-1,500). Limits: FoV limited to 90-120°, finite mirror lifespan (10,000-50,000 h).

4. FMCW LiDAR

FMCW (Frequency Modulated Continuous Wave) uses a continuously frequency-modulated laser. Distance is measured by interference (heterodyning) between emitted and reflected signals. Doppler shift provides instantaneous radial velocity for every point. Aeva (Aeries II, Daimler Truck contract) and Mobileye are leaders. Decisive advantages: instantaneous velocity measurement, complete interference immunity, 10-20 dB SNR improvement. Limits: low maturity (TRL 5-7), high cost.

5. Hemispherical LiDAR

Hemispherical LiDAR maximizes vertical FoV (up to 105°) for near-hemispherical coverage. The Hesai JT128 (128 channels, 105° vertical) and Ouster OSDome (16 channels, 90° vertical) are representative. Ideal for mobile robots (AMRs), drones and delivery vehicles that must detect both ground and high obstacles simultaneously. Limits: shorter range (50-150 m), reduced angular density at edges.

Comparison Table

Key criteria by technology: Spinning → 360° FoV, 200-300 m, $1,000-8,000, TRL 9. Flash → 60-120° FoV, 50-150 m, $200-1,000, TRL 7-8. OPA → 30-60° FoV, 50-100 m, $100-500 potential, TRL 4-6. MEMS → 90-120° FoV, 150-300 m, $300-1,500, TRL 8-9. FMCW → 90-120° FoV, 150-300 m, $500-3,000, TRL 5-7. Hemispherical → 180-360° FoV, 50-150 m, $1,000-5,000, TRL 7-8.

Conclusion: How to Choose

For outdoor mapping: mechanical spinning (Hesai Pandar128, Ouster OS2). For series automotive: MEMS (RoboSense M1, InnovizTwo, Valeo Scala 3). For premium Level 4+: FMCW (Aeva, Mobileye). For AMRs and mobile robots: hemispherical (Hesai JT128, Ouster OSDome). Hybrid architectures combine multiple technologies. The trend is toward photonic chip integration and cost reduction through automotive volumes.