Humanoid-universe.com

Ran into exactly this myself. Series elastic actuators add a spring element in series with the drivetrain, which gives cheap, precise torque sensing (measure spring deflection) and passive shock absorption, at the cost of reduced control bandwidth and added complexity.

Slight correction, though the overall point stands: Quasi-direct drive (QDD) actuators use lower gear ratios (roughly 6:1 to 10:1) paired with high-torque-density motors, trading some peak force density for backdrivability and low reflected inertia, which matters a lot for impact tolerance and fall recovery. Motor cogging torque (the 'notchy' feel from magnet-to-slot interactions) shows up as small periodic torque ripple that can translate into visible joint jitter at low speeds - it's usually mitigated with skewed magnets, better current control, or software compensation tables.

Tangent, but worth mentioning: Anti-backlash techniques in production actuators range from mechanical preloading (spring-loaded gear meshes) to purely software compensation that models and corrects for known backlash in the control loop. Backdrivability is the property that lets an external force move a joint without destroying the gearbox or the motor - it's central to both safe human-robot contact and to letting a leg comply naturally when the robot stumbles.

This lines up with my experience. It's common for a single humanoid to mix actuator types by joint - harmonic drives or hybrid geared actuators at the hips and shoulders where sustained torque matters most, and lower-ratio QDD-style actuators at ankles and knees where backdrivability and impact tolerance matter more. Motor cogging torque (the 'notchy' feel from magnet-to-slot interactions) shows up as small periodic torque ripple that can translate into visible joint jitter at low speeds - it's usually mitigated with skewed magnets, better current control, or software compensation tables.

@joseph.robinson Just to be precise about one thing: Boston Dynamics' commercial electric Atlas switched fully from hydraulic to QDD-style electric actuation, ending up quieter and lighter with comparable or better dynamic performance than the old hydraulic platform.

@forgecam45 This raises a question for me - Quasi-direct drive (QDD) actuators use lower gear ratios (roughly 6:1 to 10:1) paired with high-torque-density motors, trading some peak force density for backdrivability and low reflected inertia, which matters a lot for impact tolerance and fall recovery. Anti-backlash techniques in production actuators range from mechanical preloading (spring-loaded gear meshes) to purely software compensation that models and corrects for known backlash in the control loop. Makes me wonder how this looks in another five years.

@jonathan.rao1 Slight correction, though the overall point stands: Backdrivability is the property that lets an external force move a joint without destroying the gearbox or the motor - it's central to both safe human-robot contact and to letting a leg comply naturally when the robot stumbles.

I'd take that specific number with a grain of salt, honestly. Encoder resolution requirements scale with how tightly you need to control low-speed motion - coarse encoders are fine for open-loop swing phases but cause visible stutter during precise placement or fine force control.

Same conclusion I've come to. Also worth noting: For dynamic walking and recovery, joint control bandwidth in the tens of Hz range is typically necessary to react to a stumble before the center of mass gets too far outside the support polygon.

@jessica.karlsson To answer this directly: Backdrivability is the property that lets an external force move a joint without destroying the gearbox or the motor - it's central to both safe human-robot contact and to letting a leg comply naturally when the robot stumbles.