Robotic Component Feature Detection for Changing Hardware Footprints

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current industrial robotic systems require complete reprogramming every time the hardware footprint of components changes, making them inefficient when dealing with variations in components like motherboards, as they need to relearn detection programs for tasks such as placing DIMMs or recognizing fiducial markers.

Innovation Solution

Implementing a machine learning-based object feature identification system that uses trained AI models and machine learning engines to scan data from sensors, allowing robotic systems to automatically recognize and adapt to new components without the need for reprogramming, by processing data to identify relevant features like DIMM slots or fiducial markers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional sensor-based detection programs are used, then the robotic system can accurately identify components and their features, but the system requires complete reprogramming every time the hardware footprint of a component changes

Engineering Contradiction:
Improvecomponent detection accuracyVSAvoidadaptability to component variations
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The robotic system employs machine learning models that automatically adapt to new component variations without requiring human reprogramming. The system self-learns from new sensor data when components are updated, enabling autonomous adaptation to hardware footprint changes while maintaining detection accuracy.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system changes its detection parameters dynamically by updating machine learning models with new training data when components are updated. This allows the robotic system to adjust its detection capabilities to match new hardware footprints without complete reprogramming, resolving the contradiction between maintaining precision and adapting to variations.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If complete reprogramming is performed every time components are updated, then the detection program remains accurate for new components, but the lead time and productivity are reduced

Engineering Contradiction:
Improvefeature detection accuracyVSAvoidsystem adaptation speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

Machine learning models are pre-trained on diverse component variations during the development phase, enabling the system to handle new components without immediate reprogramming. This preliminary preparation allows rapid adaptation when new hardware is introduced, maintaining detection accuracy while improving productivity.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system performs self-updating of detection models when new component data is available, eliminating the need for manual reprogramming interventions. This autonomous adaptation process maintains feature detection accuracy while significantly reducing the lead time required for system updates.

Inventive Principle:
Principle #25Self-service

3Device complexity

If traditional detection programs are used, then the system structure remains simple, but the complexity of reprogramming and maintenance increases

Engineering Contradiction:
Improvedetection system structureVSAvoidreprogramming ease
Core Design Contradiction:
Device complexityVSEase of manufacture

Solution Approach 1:

The patent replaces traditional rule-based detection programs with machine learning-based detection systems. This substitution changes the underlying detection mechanism from explicit programming to model-based inference, simplifying the process of adapting to new components while maintaining detection capabilities.

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

Data Source

PatentEP3615282B1Component feature detector for robotic systems
Publication Date: 2021.07.21 GOOGLE LLC
  • EP3615282B1 patent drawingFigure 1
  • EP3615282B1 patent drawingFigure 2
  • EP3615282B1 patent drawingFigure 3

AI summary

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for an object feature identification system employed by a robotic are disclosed. In one aspect, a method includes the actions of generating a data reading of a work area by scanning the work area with a sensor device of the robot; identifying, by processing the data reading through a learning engine, a particular component of a plurality of components associated with the work area based on a task to be performed; identifying, with the machine learning engine, a particular feature of the particular component used in a completion of the task; determining, with the machine learning engine, a particular tool of a plurality of tools of the robot that is configured to perform the task; and performing the task with the particular tool and the particular feature of the particular component.