Flexible Robotic Arm Control for Self-Propulsion in Confined Passages

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Snake-like robotic arms face limitations in navigating complex and cluttered environments, particularly in confined spaces like gas turbine engines, due to their slender and flexible nature, which restricts their ability to maintain traction and self-propel effectively without operator intervention.

Innovation Solution

The implementation of a system comprising a flexible robotic arm with multiple degrees of freedom, coupled with sensors and an actuator, uses a machine learning model to control the arm's movement based on environmental maps and operator instructions, enabling self-propulsion and traction in cluttered spaces through modes like sidewinding, corkscrewing, and inchworm movements, allowing the arm to automatically navigate and maintain position without direct operator control.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the robotic arm is made slender and flexible to access confined spaces, then adaptability to complex environments is improved, but the ability to maintain traction and self-propel effectively deteriorates

Engineering Contradiction:
Improveability to access confined spacesVSAvoidtraction maintenance capability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The robotic arm transitions from static passive insertion to dynamic self-propelled movement through machine learning control. The system dynamically adjusts the arm's configuration and movement patterns (inchworm, sidewinding, corkscrewing) based on real-time sensor feedback and environmental maps, enabling the slender flexible arm to generate its own propulsion forces and maintain traction autonomously

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The robotic arm becomes self-sufficient by autonomously navigating complex passages without operator intervention. The machine learning model enables the arm to self-propel through cluttered environments, automatically gain and maintain traction, and independently adjust its movement strategy based on sensor readings and environmental maps, eliminating the need for manual extraction or control

Inventive Principle:
Principle #25Self-service

2Adaptability or versatility

If the robotic arm is made slender and flexible to access confined spaces, then adaptability to complex environments is improved, but the need for operator intervention increases

Engineering Contradiction:
Improveability to access confined spacesVSAvoidoperator intervention requirement
Core Design Contradiction:
Adaptability or versatilityVSEase of operation

Solution Approach 1:

The robotic arm autonomously performs navigation tasks without operator intervention. The machine learning model processes sensor readings and environmental maps to generate control signals that enable self-propulsion, automatic traction maintenance, and independent navigation through complex passages, making the system self-sufficient and eliminating the need for continuous manual control

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical control with an intelligent autonomous control system. Instead of operators directly manipulating the robotic arm, a machine learning model processes environmental data and sensor feedback to automatically generate movement commands, substituting human-operated mechanical control with autonomous intelligent control

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Extent of automation

If machine learning control is implemented for autonomous navigation, then extent of automation is improved, but device complexity increases

Engineering Contradiction:
Improveautonomous navigation capabilityVSAvoidcontrol system complexity
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The machine learning control system serves multiple functions: it processes sensor readings, generates environmental maps, determines navigation paths, controls actuator movements, and adapts to different passage geometries. This multi-functional approach consolidates what would otherwise require separate specialized systems into a single unified intelligent controller, managing complexity through functional integration

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP4257304A1System and method for flexible robotic arm navigation
Publication Date: 2023.10.11 GENERAL ELECTRIC CO
  • EP4257304A1 patent drawingFigure 1A
  • EP4257304A1 patent drawingFigure 1B
  • EP4257304A1 patent drawingFigure 2

AI summary

A robotic arm (102) is inserted into a passage (103) of a part (108) to be examined. Operator instructions defining a tip motion for a tip of the robotic arm (102), sensor readings, and an environmental map (111) are received. The operator instructions, the environmental map (111) and sensor readings are applied to a previously trained machine learning model (115) to produce control signals (120). The control signals (120) to an actuator (104) on the arm (102) to control a movement of the robotic arm (102) allowing the robotic arm (102) to automatically gain traction in the passage (103) and automatically self-propel according to a movement.