Legged Robot Ground Plane Compensation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Legged robots face challenges in traversing uneven terrain, leading to issues such as joint damage and loss of control due to uneven ground contact, which existing technologies fail to adequately address by adjusting posture in real-time based on changing topographical features.

Innovation Solution

A control system that receives sensor data to estimate the ground plane and adjusts the robot's pitch and roll in proportion to the terrain's grade, improving kinematic reachability and stability by timing these adjustments to align with upcoming changes in the terrain.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the robot maintains a fixed gait pattern, then the control system is simple, but the robot cannot adapt to uneven terrain causing joint damage and loss of control

Engineering Contradiction:
Improveterrain adaptationVSAvoidcontrol system complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The control system dynamically adjusts gait parameters (pitch, roll, step timing) in real-time based on sensor feedback about terrain conditions. This allows the robot to adapt to uneven terrain while maintaining manageable control complexity through structured dynamic adjustment rather than complete gait redesign.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system uses sensor data about ground plane grade and topographical features to continuously monitor terrain conditions and adjusts gait parameters accordingly. This feedback loop enables terrain adaptation by comparing actual terrain with expected conditions and making corrective adjustments to pitch, roll, and step timing.

Inventive Principle:
Principle #23Feedback

2Reliability

If the robot adjusts gait parameters in real-time, then terrain adaptability improves, but computational load and control complexity increase

Engineering Contradiction:
Improvetraversal reliabilityVSAvoidcontrol algorithm complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system estimates the ground plane grade and upcoming terrain features in advance of actual contact, allowing proactive adjustment of gait parameters before the robot encounters difficult terrain. This preliminary action improves traversal reliability by preparing the robot for upcoming challenges rather than reacting after problems occur.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The control system adjusts specific gait parameters (pitch angle, roll angle, step timing, step length) based on calculated ground plane grade. By changing these parameters in proportion to the determined grade, the system achieves reliable terrain traversal through structured parameter modulation rather than complex behavioral changes.

Inventive Principle:
Principle #35Parameter changes

3Adaptability or versatility

If the robot uses aggressive gait adjustments to handle steep terrain, then terrain capability improves, but stability and joint safety deteriorate

Engineering Contradiction:
Improveterrain capabilityVSAvoidrobot stability
Core Design Contradiction:
Adaptability or versatilityVSStability of the object's composition

Solution Approach 1:

The system applies counter-actions in advance to offset upcoming terrain challenges. By estimating the ground plane grade and adjusting pitch and roll in proportion to the determined grade before contact, the robot preemptively counteracts the effects of steep terrain, maintaining stability while improving terrain capability.

Inventive Principle:
Principle #9Preliminary anti-action

Solution Approach 2:

The control system timing adjustments to align with upcoming terrain changes provides a cushioning effect, smoothing out the transition onto steep or uneven ground. This timing-based cushioning prevents abrupt impacts and maintains joint safety by preparing the robot's posture before terrain challenges arise.

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

Data Source

PatentUS9908240B1Ground plane compensation for legged robots
Publication Date: 2018.03.06 BOSTON DYNAMICS INC
  • US9908240B1 patent drawing
  • US9908240B1 patent drawing
  • US9908240B1 patent drawing

AI summary

An example implementation includes (i) receiving sensor data that indicates topographical features of an environment in which a robotic device is operating, (ii) determining, for a particular topographical feature of the environment in a direction of travel of the robotic device, a height of the particular topographical feature and a distance between the robotic device and the particular topographical feature, (iii) estimating a ground plane extending from the robotic device in the direction of travel toward the particular topographical feature, the ground plane fitting to the determined distance and height, (iv) determining a grade of the estimated ground plane, and (v) directing the robotic device to adjust pitch in proportion to the determined grade.