Shamel Fahmi, Michele Focchi, Andreea Radulescu, Geoff Fink, Victor Barasuol and Claudio Semini

Dynamic Legged Systems (DLS) lab, Istituto Italiano di Tecnologia (IIT)

STANCE

Locomotion Adaptation over Soft Terrain

ICRA 2020

taken from Andrea del Prete, 2019

Shamel Fahmi

• Most WBCs are not terrain aware.  They fail to generalize beyond rigid terrain.

• Mismatch between what the WBC assumes vs. what is actually happening

Shamel Fahmi

• Drop the assumption of the rigid terrain. Use a more generic model that accounts for the terrain impedance 

• But that’s not enough, how do you know the impedance parameters of the terrain. 

• Hmm, estimate it?

Shamel Fahmi

Shamel Fahmi

Soft Terrain Adaptation aNd Compliance Estimation
(STANCE)

Shamel Fahmi

• The Standard WBC (sWBC) framework:

  • Locomotion Planner
  • State Estimator
  • WBC
  • Low-level Control

Shamel Fahmi

• Execute the planned trajectories of the body and swinging leg(s) (a.k.a control tasks)

Goals:

• Keep the robot balanced

• Consistent contact forces and accelerations with the dynamics

• Respect the robot's dynamics and actuation limits

• Respect the contact interaction constraints

• Maintain contact consistency

• Casted as an optimization problem via Quadratic Programming (QP)

• Mapped using inverse dynamics

Shamel Fahmi

Shamel Fahmi

• Using the KV model

• The stance task does not hold anymore

• Compliant contact consistent (c3)

• Soft contact dynamics must be taken into account

• c3 Stance Task

Shamel Fahmi

Shamel Fahmi

• Estimate the terrain parameters online

• Based on the current measurements (states)

• Decoupled from the C3WBC but uses the same contact model

Shamel Fahmi

Shamel Fahmi

Shamel Fahmi

Shamel Fahmi

Shamel Fahmi

• Unlike previous work on WBC, we do not assume that the ground is rigid.

• Adapt online to any type of terrain compliance

• C3WBC

  • ​Extends the sWBC but incorporating soft contact dynamics

  • C3WBC can adapt to the given terrain compliance

• ​TCE

  • ​Estimate the terrain compliance online

  • Not computationally expensive

Shamel Fahmi

• With STANCE, HyQ can traverse and transition between multiple terrains with different compliance without pre-tuning.

• STANCE makes the locomotion strategy c3. It allowed HyQ to remain consistent with ground, and thus more robust in challenging and scenarios.

• Each leg independently senses and adapts to changes in terrain compliance.

G. Urbain et. al., "Stance Control Inspired by Cerebellum Stabilizes Reflex-Based Locomotion on HyQ Robot"

A. Bratta et. al., ''On the Hardware Feasibility of Nonlinear Trajectory Optimization for Legged Locomotion based on a Simplified Dynamics''

O. Villarreal et. al., ''MPC-based Controller with Terrain Insight for Dynamic Legged Locomotion''