Humanoid-universe.com

@servobre20 Small correction on one detail: Event cameras (which report per-pixel brightness changes rather than full frames) are still more of a research curiosity than a production sensor for humanoids, mainly because the software ecosystem and processing pipelines around them are far less mature than for standard frame-based cameras.

Same conclusion I've come to. Also worth noting: Estimating joint torque from motor current draw is cheap and requires no extra sensor, but it's less accurate than a dedicated torque sensor because it doesn't capture friction losses through the gearbox - good enough for coarse control, not always for precise force-controlled tasks.

@garcia51 This is a great summary, thanks. Multi-camera calibration drifts over time from thermal expansion, vibration, and mechanical wear, which is why production systems typically run periodic recalibration routines rather than assuming a one-time factory calibration holds forever. Force/torque sensors near the ankle give a direct read on ground reaction forces, which is valuable for balance control, but they add cost, a failure point, and routing complexity right at a joint that already takes the most mechanical abuse. Reminds me a bit of the early drone hobbyist scene, honestly.

@rao91 Same conclusion I've come to. Also worth noting: Sensor fusion mostly earns its keep by covering for each individual sensor's weaknesses - vision struggles with occlusion and lighting, IMUs drift, force/torque sensors are noisy at low loads - fusing them gives a more robust estimate than any one source alone, independent of raw compute. Multi-camera calibration drifts over time from thermal expansion, vibration, and mechanical wear, which is why production systems typically run periodic recalibration routines rather than assuming a one-time factory calibration holds forever.

New to this, so forgive me if this is obvious - Estimating joint torque from motor current draw is cheap and requires no extra sensor, but it's less accurate than a dedicated torque sensor because it doesn't capture friction losses through the gearbox - good enough for coarse control, not always for precise force-controlled tasks. LiDAR gives reliable, lighting-independent range data but is heavier, pricier, and gives sparser point clouds up close than stereo or depth cameras, which is why a lot of humanoids lean on stereo/depth cameras for near-field manipulation and reserve LiDAR (if present at all) for longer-range navigation.

@servosan90 Slightly off-topic, but related: Sensor fusion mostly earns its keep by covering for each individual sensor's weaknesses - vision struggles with occlusion and lighting, IMUs drift, force/torque sensors are noisy at low loads - fusing them gives a more robust estimate than any one source alone, independent of raw compute. Unitree's Dex3-1 dexterous hand packs around 33 pressure/tactile sensors per hand across the fingers and palm, capable of sensing pressure roughly in the 10g-2500g range - a useful reference point for what 'production tactile sensing' looks like right now.

Slightly off-topic, but related: Proprioception (the robot's sense of its own joint angles, velocities, and forces) tends to get less attention than flashy vision systems, even though a lot of balance and manipulation failures trace back to proprioceptive noise or miscalibration rather than a vision problem.

@robertmiller Slight correction, though the overall point stands: Vibration is one of the most underrated sources of noisy IMU and tactile readings - mounting matters as much as sensor quality, and a poorly isolated mount can add more noise than the sensor's own datasheet specs would suggest.

Worth being a little skeptical of the marketing angle here. IMU drift over time (bias instability) is usually the real culprit behind slowly diverging state estimates, not noise - it's typically handled with sensor fusion against other references (visual odometry, joint kinematics) rather than trying to eliminate drift at the source.

I'll believe the stronger version of that claim when it's independently verified. Event cameras (which report per-pixel brightness changes rather than full frames) are still more of a research curiosity than a production sensor for humanoids, mainly because the software ecosystem and processing pipelines around them are far less mature than for standard frame-based cameras. Totally unrelated but has anyone else noticed how fast component costs are dropping this year.