Autonomous Robot Localization Using Reliability-Weighted Sensor Fusion

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Autonomous devices face challenges in accurate localization due to the unreliability of visual inputs in dynamic environments and the incremental inaccuracies of odometry-based systems, which can compound location errors over time.

Innovation Solution

An autonomous device is configured with multiple localization systems, including a vision system, a triangulation system, and an odometry system, where processing devices select the most reliable location based on mathematical filters and comparisons, using landmarks, signals, and wheel movements to determine and control its movement within a space.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If visual inputs are used for localization, then the system can operate without prior localization data, but the localization becomes unreliable in dynamic environments or when features are not discernable

Engineering Contradiction:
Improvelocalization capabilityVSAvoidlocalization reliability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent combines multiple localization systems (visual inputs, odometry data, and other sensors) into a unified localization framework. The processing device integrates data from these diverse sources to determine the device's location, thereby maintaining adaptability across different environments while improving reliability through redundancy and cross-validation of localization data.

Inventive Principle:
Principle #5Merging (Combining)

2Ease of operation

If odometry-based localization is used, then the system can determine location incrementally, but location errors compound with each subsequent localization

Engineering Contradiction:
Improvelocalization processVSAvoidlocation accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent implements a feedback mechanism where the processing device continuously evaluates the reliability of odometry-based localization data against other localization sources. When odometry errors begin to compound, the system detects this through reliability assessment and switches to or integrates with alternative localization methods, thereby maintaining measurement precision while preserving the incremental nature of odometry-based operation.

Inventive Principle:
Principle #23Feedback

3Reliability

If multiple localization systems are integrated, then reliability and accuracy of localization improve, but device complexity increases

Engineering Contradiction:
Improvelocalization reliabilityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent designs the processing device to perform multiple functions: it processes visual inputs, odometry data, and other sensor data using a unified reliability assessment framework. This multi-functional approach allows the system to maintain high localization reliability through diverse data sources while managing complexity by using a single integrated processing architecture rather than separate specialized systems.

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

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This multi-system approach enhances the reliability and accuracy of localization, reducing errors and improving the autonomous device's ability to navigate effectively in various environments by averaging or weighting locations from different systems.

Implementation Method 1

The second system may include a receiver for receiving signals and timestamp messages based on the signals. The one or more processing devices may be configured to determine the second location based on a received signal and a timestamp message corresponding to the received signal. The timestamp message may contain a time that the received signal was transmitted by a remote receiver.

Methodology Applied
Scientific EffectTime of Flight: Time of Flight

Data Source

PatentEP3884353B1Detecting a location of an autonomous device
Publication Date: 2022.09.28 MOBILE IND ROBOTS AS
  • EP3884353B1 patent drawingFigure 1
  • EP3884353B1 patent drawingFigure 2
  • EP3884353B1 patent drawingFigure 3

AI summary

An example autonomous device is configured to move within a space. The autonomous device includes a first system to detect a first location of the autonomous device within the space, with the first location being based on a first fixed reference; a second system to detect a second location of the autonomous device within the space, with the second location being based on a second fixed reference; and a third system to detect a third location of the autonomous device within the space based on relative movements of the autonomous device. One or more processing devices are configured to select one of the first location or the second location based on reliability of at least one of the first location or the second location, and to control movement of the autonomous device using an estimated location that is based on the third location and the selected one of the first location or the second location.