Capability State Hierarchies for Robot Error Categorization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current robotic systems lack effective methods for categorizing, reporting, and introspecting anomalies, particularly robot errors, which hinders their ability to self-reason and request human assistance in dynamic environments.

Innovation Solution

A method and system utilizing an apparatus state data structure that includes capability and associated data fields to represent and control technical apparatuses, enabling efficient error categorization, reporting, and recovery by associating functionalities with component states, diagnostics, and critical events.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional diagnostic systems are used for robot error identification, then error reporting capability is provided, but the system lacks introspection and self-reasoning capabilities

Engineering Contradiction:
Improveerror detection capabilityVSAvoidsystem introspection capability
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements a hierarchical state representation where capability states are nested within apparatus states, and component states are nested within capability states. This nested structure allows the system to maintain multiple levels of abstraction (apparatus level, capability level, component level) simultaneously, enabling both simple error detection and complex introspection without proportionally increasing system complexity.

Inventive Principle:
Principle #7Nested doll (Nesting)

Solution Approach 2:

The patent introduces a new dimensional framework by adding capability states as an intermediate layer between apparatus states and component states. This dimensional expansion transforms the traditional flat error reporting structure into a multi-dimensional state space, allowing the system to reason about errors from multiple perspectives (what the apparatus can do, what components are involved, and how they relate).

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If detailed state representation is implemented for error categorization, then anomaly classification accuracy is improved, but data structure complexity increases

Engineering Contradiction:
Improveerror classification accuracyVSAvoiddata structure complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the error detection and classification system into distinct hierarchical levels: apparatus states (overall functionality), capability states (specific functional abilities), and component states (individual hardware/software elements). Each level handles specific aspects of error classification, allowing detailed and accurate error categorization while managing data structure complexity through modular organization.

Inventive Principle:
Principle #1Segmentation

3Reliability

If real-time capability monitoring is implemented, then operational safety is enhanced, but computational overhead increases

Engineering Contradiction:
Improveoperational safetyVSAvoidcomputational energy consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent implements preliminary action by pre-defining the hierarchical state structure and capability-component relationships before runtime. The system pre-establishes which components are associated with which capabilities and which apparatus functions depend on which capabilities. This preliminary organization allows real-time monitoring to proceed efficiently by simply checking predefined relationships rather than computing them dynamically, reducing computational overhead while maintaining operational safety.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4140664A1Capabilities for error categorization, reporting and introspection of a technical apparatus
Publication Date: 2023.03.01 JUNGHEINRICH AG
  • EP4140664A1 patent drawingFigure 1
  • EP4140664A1 patent drawingFigure 2
  • EP4140664A1 patent drawingFigure 3

AI summary

A method of operating an autonomous vehicle or a robot. The vehicle or robot can be modelled as having a set of capabilities that allow it to perform tasks, which are associated with a set of required capabilities to complete it. For example, a task of moving to a position requires the robot to be able to correctly move its base, thus to have a navigation capability. The vehicle or robot is controlled using a vehicle state data structure (10), a robot state representation (10) or a vehicle state data structure 10, which comprise capability data fields (110) indicating a functionality of the autonomous vehicle and associated data fields (120,130,140) associated with a respective capability data field (110). The associated data field (120, 130, 140) comprises a required component state data field (130) indicating a configuration of a respective component required for the functionality and a required diagnostic data field (120) indicating a respective operational state of a component. The present invention also relates to a corresponding system, computer program product, and computer readable storage medium.