LiDAR & Depth Sensors for Humanoids: Shipping Hardware, Specs, and India Market Realities

Solid-State LiDAR: The Current Shipping Standard

Solid-state LiDAR has transitioned from laboratory prototypes to mass-produced optical assemblies, making it the baseline perception layer for modern humanoid and mobile robotic platforms. Unlike mechanical spinning units that rely on rotating mirrors and motors, solid-state designs route laser pulses through fixed optical elements. The two dominant architectures in shipping hardware are MEMS (Micro-Electro-Mechanical Systems) mirrors and optical phased arrays (OPA). MEMS variants use a microscopic tilting mirror to steer a single laser beam across a defined field of view, while OPA systems use interference patterns generated by phase-shifted emitters to direct light without moving parts.

Shipping hardware from established optical manufacturers now delivers consistent point cloud densities, reliable ranging performance, and standardized data interfaces. The grading hierarchy places actual production units first, followed by pilot deployments, and finally announced or pre-release products. Current shipping solid-state LiDAR units typically offer 120 to 360 degrees of horizontal field of view, 50 to 150 meters of effective range, and frame rates between 10 Hz and 20 Hz. These specifications align with humanoid navigation, obstacle avoidance, and spatial mapping requirements where environmental illumination does not degrade performance.

MEMS and Optical Phased Array Architectures

MEMS-based LiDAR remains the most widely deployed solid-state architecture in robotics. Manufacturers have optimized mirror actuation, thermal stability, and calibration routines to reduce point cloud noise and maintain accuracy across temperature cycles. OPA systems promise higher reliability due to the absence of mechanical wear, but they currently face challenges with side-lobe suppression, angular resolution scaling, and cost-per-channel optimization. When evaluating claims, production volumes and factory test reports serve as the primary validation metric.

Key shipping platforms include Hesai's QT series and Pandar series (hybrid solid-state variants), Ouster's SD and SM generation sensors, and Livox's custom telecentric designs. These units ship with integrated timing modules, GNSS/IMU sync inputs, and Ethernet or CAN bus connectivity. They are routinely integrated into humanoid torsos, head assemblies, and mobile bases for 3D localization and dynamic obstacle tracking.

Time-of-Flight (ToF) Sensors: Short-Range Precision

Time-of-Flight sensors measure distance by calculating the round-trip time of modulated light pulses or continuous wave signals. In humanoid robotics, ToF cameras occupy the short-to-medium range perception tier (0.1 to 15 meters), complementing LiDAR's longer reach. They provide dense depth maps at high frame rates, making them suitable for hand-eye coordination, grasp planning, and close-proximity navigation.

Shipping ToF hardware is graded by silicon maturity and production volume. Direct ToF (dToF) uses single-photon avalanche diodes (SPADs) and time-to-digital converters to measure pulse transit time, offering high dynamic range and sunlight rejection. Indirect ToF (iToF) modulates continuous wave signals and measures phase shifts, delivering higher resolution at lower cost but requiring careful calibration to avoid multipath interference.

Direct vs. Indirect Phase-Shift Measurement

Direct ToF sensors, such as Sony's SPAD-based imaging arrays and STMicroelectronics' ToF modules, ship with integrated processing pipelines and deliver calibrated depth data with millimeter-level accuracy at close range. Indirect ToF units, including ST's VL53L series and Bosch's BNo5 modules, are widely available in compact form factors and are often mounted on humanoid wrists, ankles, and joint housings for proximity sensing and collision avoidance. Both categories ship as mature hardware with documented SNR specs, operating temperature ranges, and EMI shielding requirements.

Stereo Depth Systems: Computational Parallax

Stereo depth sensors calculate distance by comparing disparity between two offset optical sensors. Unlike active LiDAR and ToF systems, stereo cameras are passive and rely on ambient light, which introduces dependency on texture, contrast, and illumination conditions. However, they offer high spatial resolution, low power consumption, and compatibility with existing computer vision stacks.

Shipping stereo depth hardware includes Stereolabs' ZED series, Intel RealSense D400 series, and custom baselined camera pairs. These units ship with synchronized global shutters, calibrated intrinsic/extrinsic parameters, and real-time disparity computation firmware. Deployment grading places production stereo kits first, followed by pilot integrations in humanoid upper limbs and mobile bases, with announced research prototypes ranked last.

Stereo systems require significant on-device compute for depth map generation, rectification, and confidence filtering. Modern humanoid platforms offload this work to dedicated NPUs or GPU cores, enabling frame rates of 30 to 60 Hz at resolutions ranging from VGA to 4K. The trade-off remains clear: passive stereo reduces cost and power but increases algorithmic complexity and lighting dependency.

Grading Claims: Shipping Hardware vs. Pilots vs. Announcements

RobotWale grades perception sensor claims using a strict hardware-first hierarchy:

• Shipping Hardware: Units available for purchase, with documented factory test reports, calibration certificates, and documented deployment logs. Examples include Hesai Pandar/QT series, Ouster SD/SM generation, Sony SPAD ToF arrays, Stereolabs ZED 2i/3, and Intel RealSense D455.
• Pilot Deployments: Sensors installed in limited-production humanoid or mobile robot fleets, with public telemetry, deployment durations, and failure mode documentation. Examples include Livox telecentric LiDAR in university humanoid platforms and custom ToF arrays in industrial inspection robots.
• Announcements: Pre-release specifications, crowdfunding campaigns, or prototype demos without verified production volumes or independent test data. These are tracked but not weighted in procurement decisions.

Manufacturers often conflate lab performance with field reliability. Independent verification through factory videos, on-stage demos with live point clouds, and press release cross-referencing remains essential. Claims regarding range, resolution, and frame rate should be validated against published spec sheets and third-party measurement reports.

India Availability and Landed Cost Estimates

India's robotics supply chain for LiDAR and depth sensors relies heavily on imports, with domestic assembly limited to integration and calibration stages. Landed cost estimates account for basic customs duties (10% to 20% depending on HS code), Integrated Goods and Services Tax (18% GST), and Additional Integrated Customs Duty (AIDC) where applicable. All pricing figures are approximate and subject to exchange rate fluctuations, import policy updates, and distributor margins.

• Solid-State LiDAR: Shipping units from Hesai, Ouster, and Livox are available through authorized Indian distributors and robotics integrators. Landed cost estimates range from INR 1,80,000 to INR 4,50,000 per unit, depending on range, resolution, and mounting accessories. Local calibration and firmware localization services add INR 15,000 to INR 30,000.
• Time-of-Flight Sensors: Sony SPAD arrays, ST ToF modules, and Bosch proximity sensors are widely available via electronic component distributors and robotics suppliers. Landed costs range from INR 8,000 to INR 45,000 per sensor, with bulk procurement reducing per-unit pricing by 10% to 15%.
• Stereo Depth Systems: Stereolabs ZED 2i/3 and Intel RealSense D455 units ship through Indian robotics partners and electronics channels. Landed costs range from INR 60,000 to INR 1,20,000, with calibration kits and mounting brackets adding INR 10,000 to INR 25,000. Local assembly and firmware adaptation are typically handled by system integrators.

Domestic manufacturing of LiDAR and ToF silicon remains limited, but Indian firms are expanding sensor calibration, housing fabrication, and software localization. Procurement teams should verify import documentation, warranty terms, and firmware update policies before deployment. Pilot testing in Indian environmental conditions (dust, humidity, temperature variance) is recommended before scaling.

References

Hesai Technology. "Pandar Series LiDAR Specifications." https://www.haesitech.com/en/products/pandar-series

Ouster. "SD Series Solid-State LiDAR Product Sheet." https://ouster.com/products/sd-series

Sony Semiconductor. "SPAD-Based Time-of-Flight Image Sensors." https://www.sony-semicon.com/en/products/inf/spad.html

Stereolabs. "ZED 2i Depth Camera Specifications." https://www.stereolabs.com/zed-2i/

Intel Corporation. "RealSense D400 Series Depth Camera Datasheet." https://www.intel.com/content/www/us/en/develop/products/real-sense/depth-cameras/d400.html

Livox Technology. "Telecentric LiDAR for Robotics." https://www.livoxtech.com/products/telecentric-lidar

IEEE Spectrum. "Solid-State LiDAR: How It Works and Where It Ships." https://spectrum.ieee.org/solid-state-lidar

Robotics Business Review. "ToF vs. Stereo Depth in Humanoid Platforms." https://www.roboticsbusinessreview.com