Simulation testedroller skatev1.0.0Apache-2.0

Roller Slope Descent

Downhill slope gliding policy on roller skates with dynamic pitch stabilization and proactive speed moderation.

About this behavior

Trained specifically for inclined surfaces (5° to 15° slopes). Adjusts the duck's center of mass backward and applies staggered foot drags to prevent runaway speed down ramps while stabilizing roll and pitch angles against gravitational pull.

Runtime contract

61-dimension input · 14 joint targets · 50 Hz

Show details

50 Hz / 4×slot: rollers

Observation vector61 dimensions · normalizer baked in

Proprioception48D

Positions, velocities, previous actions, gravity, and gyro.

Twist command3D

Forward, lateral, and yaw velocity intent.

Head pose4D

Neck and head orientation targets.

Body pose6D

Six-axis torso offset command.

Action targetsscale: 1×

Left leg5 joints

hip yaw · hip roll · hip pitch · knee · ankle

Neck & head4 joints

neck pitch · head pitch · yaw · roll

Right leg5 joints

hip yaw · hip roll · hip pitch · knee · ankle

Actuator model: Dynamixel XL330 (BAM M6 voltage control law). The beak motor is separate from this fixed locomotion action space.

Verification

Status: Simulation tested
Evidence: Verified in mjlab simulation descending varying slope angles up to 15 degrees without destabilization.
Target hardware: MicroDuck v1 + Roller Skate Blades
Simulation: mjlab (MuJoCo Warp) at 50 Hz
Note: Experimental ramp policy; ensure run-out area is clear of obstacles during initial hardware trials.

Compatibility

Robot model: microduck-rollers
MJCF model: robot_allcollisions_rollers.xml
Terrain: slope
Required accessories: roller skates

Run it on your duck

Use the canonical model and this policy slot to try the behavior locally.

slot: rollers

Try the simulation first. Hardware testing assumes a matching Microduck, clear space, and a safe surface. uDuck does not guarantee that a community policy is safe to run.

Deployment stepsShow instructions

1 Download the ONNX model roller_slope.onnx

curl -L "https://raw.githubusercontent.com/pollen-robotics/microduck_rl/develop/docs/models/roller_slope.onnx" -o "/opt/robot/policies/roller_slope.onnx"

2 Add the policy to robotd

[policy]
roller_slope = "/opt/robot/policies/roller_slope.onnx"

3 Restart the control loop

sudo systemctl restart robotd
robotctl health

⌁ Try it in simulation

uv run play Mjlab-RollerSlope-Flat-MicroDuck

Sources

Upstream repositorypollen-robotics/microduck_rl Training codeMjlab-RollerSlope-Flat-MicroDuck