Warehouse Robot Object Classification for Collision Avoidance

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Modern inventory systems face challenges in efficiently navigating robots through warehouse environments due to encountered obstacles, as existing technologies often require costly full stops and manual resets when encountering unknown objects, which can lead to productivity losses and increased downtime.

Innovation Solution

The implementation of robots equipped with multiple sensors, including object detection and identification systems using RFID readers, 3D scanners, and imaging devices, allows for the classification of objects in their path, enabling evasive actions such as rerouting around obstacles or full stops based on the nature of the obstruction, thereby minimizing disruptions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If robots are equipped with basic obstacle detection only, then collision avoidance is achieved, but productivity decreases due to frequent full stops and manual resets when encountering unknown objects

Engineering Contradiction:

Improvecollision avoidanceVSAvoidworkflow efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system performs preliminary classification of objects using multiple sensors (RFID readers, 3D scanners, imaging devices) before the robot commits to a path. By预先 identifying and categorizing objects in the workspace, the system prepares classification data that enables automated decision-making, preventing the need for full stops when obstacles are encountered

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system continuously monitors the workspace using multiple sensors and provides feedback to the path planning algorithm. The classification data from RFID readers, 3D scanners, and imaging devices feeds back into the navigation system, enabling real-time adjustments to robot paths based on detected objects, thereby maintaining productivity while ensuring safety

Inventive Principle:

Principle #23Feedback

2Reliability

If robots perform full stop and manual reset upon encountering any object, then safety is maintained, but downtime increases and productivity decreases

Engineering Contradiction:

ImprovesafetyVSAvoiddowntime

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The robot system performs self-service by automatically classifying objects and generating alternative paths without human intervention. The path planning algorithm autonomously processes classification data from multiple sensors and recalculates routes around obstacles, eliminating the need for manual resets and reducing downtime while maintaining safety protocols

Inventive Principle:

Principle #25Self-service

3Measurement precision

If multiple sensors and classification systems are added to robots, then object identification accuracy improves, but device complexity and cost increase

Engineering Contradiction:

Improveobject identification accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the object identification task across multiple specialized sensors rather than using a single complex system. RFID readers handle tagged objects, 3D scanners handle geometric identification, and imaging devices handle visual recognition. This segmentation allows each sensor to be optimized for its specific function while working together to achieve comprehensive object identification

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The path planning algorithm serves as a universal processor that handles classification data from multiple different sensor types (RFID readers, 3D scanners, imaging devices). This multi-functional approach allows the system to process diverse data formats and object types through a single integrated decision-making system, reducing overall complexity despite having multiple sensors

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 solution enhances workflow efficiency by allowing robots to automatically detect and classify objects, reducing the need for costly full stops and manual resets, and enabling continuous operation while maintaining safety protocols, thus improving productivity and reducing downtime.

Implementation Method 1

RFID readers, 3D scanners, and imaging devices

Methodology Applied

Scientific EffectRadio frequency identification: Electromagnetic Induction

Implementation Method 2

3D scanners

Methodology Applied

Scientific EffectLight reflection: Reflection

Data Source

PatentUS10133276B1Object avoidance with object detection and classification

Publication Date: 2018.11.20 AMAZON TECH INC

AI summary

A robot equipped with an object detection system and an object identification system is used to move inventory holders throughout a warehouse or other environment. The robot can detect objects in its path using the object detection system, which can include one or more sensors for this purpose. The robot can then classify the object using the object identification system to determine an appropriate course of action. The robot can classify the object as an inventory item, warehouse equipment, or a person, among other things. The robot can take action based on the object classification. The robot can reroute around inventory items and warehouse equipment. When encountering people or objects that cannot be classified, the robot can stop and await further instructions. In some cases, the robot may wait for a manual reset before continuing along its path.