Package Sorting Conveyor with Neural Network Vision

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conveyors in the package-handling industry struggle to efficiently separate individual packages from bulk flows, leading to jams and reduced throughput, particularly with oversized boxes and flexible polybags, requiring manual intervention.

Innovation Solution

A conveyor system equipped with a package-unit detector, imaging system, and neural network classifier that uses light curtains, cameras, and multi-core processors to identify and sort packages based on their footprints and classifications, enabling automated separation and diversion to designated destinations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual intervention is used to remove oversized boxes and polybags, then jams are prevented, but throughput is reduced

Engineering Contradiction:
Improvejam preventionVSAvoidthroughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent replaces manual mechanical intervention with an automated vision-based detection and sorting system. The system uses cameras to detect oversized boxes and polybags, identifies their positions, and automatically diverts them to appropriate destinations, eliminating the need for manual removal while maintaining throughput

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

Solution Approach 2:

The system enables the conveyor to self-monitor and self-correct by automatically detecting problematic packages (oversized boxes, polybags) and diverting them without external intervention. The vision system continuously monitors the conveyor and autonomously manages package routing

Inventive Principle:
Principle #25Self-service

2Reliability

If all packages are removed from the sorter to prevent jams, then reliability improves, but overall throughput decreases

Engineering Contradiction:
Improveconveyor reliabilityVSAvoidoverall throughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies different handling strategies to different package types based on their specific characteristics. Instead of removing all packages, the system selectively identifies and diverts only problematic packages (oversized boxes, polybags) while allowing normal packages to flow through the sorter uninterrupted

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system segments the package flow into different categories (normal packages vs. problematic packages) and routes them through different paths. This segmentation allows the majority of packages to maintain high-speed flow while problematic ones are handled separately

Inventive Principle:
Principle #1Segmentation

3Extent of automation

If automated detection systems are implemented, then manual labor is reduced, but system complexity increases

Engineering Contradiction:
Improveautomation levelVSAvoidsystem complexity
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The vision system serves multiple functions: detecting oversized boxes, identifying polybags, determining package positions, and providing data for sorting decisions. This multi-functionality reduces the need for separate specialized sensors and systems, managing complexity while maintaining high automation

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

Solution Approach 2:

The patent introduces a vision system as an intermediary between the physical packages and the control system. This intermediary captures visual data, processes it to identify package characteristics, and translates it into actionable sorting information, simplifying the overall control architecture

Inventive Principle:
Principle #24Intermediary (Mediator)

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

The system effectively separates and sorts packages with high throughput, reducing manual intervention and improving conveyor efficiency by accurately identifying and routing packages to their correct destinations.

Implementation Method 1

A package 13 passing through the light curtain blocks the transmitted light beams it intercepts from reaching the receiver array. As the package 13 passes the package-unit detector 16, its footprint, i.e., its projection on the infeed conveyor segment, can be determined from the pattern of blocked receivers.

Methodology Applied
Scientific EffectLight blocking: Absorption (EM radiation)

Data Source

PatentEP3953067B1Package detection and intelligent sorting
Publication Date: 2026.02.25 LAITRAM LLC
  • EP3953067B1 patent drawingFigure 1
  • EP3953067B1 patent drawingFigure 2
  • EP3953067B1 patent drawingFigure 3~4

AI summary

A sorting conveyor and a method for sorting packages to selected destinations. A neural network executed by a multi-core classification processor uses captured images of package units conveyed on an infeed conveyor to classify each package unit. A control processor controls a sorting conveyor to direct each package unit to a destination depending on the package unit's classification.