Multi-Contact Taxel Localization for Tactile Hyper-Resolution

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Commercially available taxel-based tactile sensors have low spatial resolution, making it difficult to determine detailed geometric information about an object's surfaces solely from physical contact, and multiple contact attempts can be time-consuming and potentially damaging.

Innovation Solution

A system and method for taxel hyper-resolution output through multi-contact localization, involving a first processing module to encode tactile sensor data, compare it with pre-derived geometric data, and use iterative pairwise comparisons and cycling algorithms to determine the most probable candidate poses, generating high-resolution geometric data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If taxel-based tactile sensors are used to determine geometric information of an object, then tactile sensor data can be obtained, but the low spatial resolution makes it difficult to determine detailed geometric information

Engineering Contradiction:

Improvespatial resolutionVSAvoidgeometric information

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The system performs preliminary encoding of tactile sensor data into taxel representations before comparison with object database entries. This preprocessing step enables efficient matching and hyper-resolution reconstruction by preparing the data in advance for detailed geometric analysis

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system creates hyper-resolution copies of low-resolution taxel data by comparing sensor readings with pre-stored object database entries. The database contains high-resolution geometric information that is copied and adapted to match the observed taxel patterns, effectively upscaling the measurement resolution

Inventive Principle:

Principle #26Copying

2Loss of information

If multiple contact attempts are made to determine object geometry, then more geometric information can be gathered, but the process becomes time-consuming and potentially damaging

Engineering Contradiction:

Improvegeometric informationVSAvoidtime for contact attempts

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The system pre-computes and stores a comprehensive object database containing taxel representations and geometric information for various objects before actual sensing occurs. This preliminary preparation enables rapid matching during actual contact, reducing the number of contact attempts needed

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from each tactile contact to refine the identification process. By comparing observed taxel patterns with the database and using hyper-resolution techniques, the system progressively narrows down candidate objects, reducing the need for multiple random contact attempts

Inventive Principle:

Principle #23Feedback

3Measurement precision

If multiple contact attempts are made to determine object geometry, then more geometric information can be gathered, but the risk of damaging the object increases

Engineering Contradiction:

Improvegeometric informationVSAvoidpotential damage to object

Core Design Contradiction:

Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The system prepares a comprehensive object database in advance that contains detailed geometric information and taxel representations. This preliminary action enables accurate object identification with minimal physical contact, reducing the need for multiple contact attempts that could damage fragile objects

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system replaces repeated mechanical contact attempts with computational processing. By using taxel encoding, database matching, and hyper-resolution algorithms, the system achieves detailed geometric analysis through data processing rather than through multiple physical contacts

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12558795B2Systems and methods for taxel hyper-resolution through multi-contact localization

Publication Date: 2026.02.24 HONDA MOTOR CO LTD

AI summary

Systems and methods for taxel hyper-resolution output are provided. Tactile sensors contact an object and produce taxel data for each contact made. Taxel data from the tactile sensors is encoded and compared with detailed geometric data of a known object to generate a set of ranked poses. For each pose one or more substantially matching candidate poses from the stored geometric data are generated. Pairs of poses are iteratively compared with pairs of candidate poses to determine the most probable candidate pose for each pose. Taxel hyper-resolution data for determined poses is generated from the detailed geometric data and output for use by downstream systems.