Construction Machine Operation Analysis for Collision Risk Extraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional operation record analysis systems for construction machines like excavators struggle to efficiently identify and extract work contents that may lead to a decrease in operation rate due to potential collisions with workers or objects, requiring extensive manual data analysis and often resulting in overlooked risks.

Innovation Solution

An operation record analysis system for construction machines that includes sensors to detect the position and posture of the machine body, operation state, and objects in its vicinity, calculating potential overlaps and integrating this information to quickly identify and highlight potential collision risks in the operation data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If operation data from multiple construction machines is collected and analyzed manually, then operational issues can be identified, but the process requires significant time and labor resources

Engineering Contradiction:
Improveoperational issue detection accuracyVSAvoiddata analysis time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical data analysis with an automated image recognition system using machine learning models. The system automatically processes operation data visualizations to detect operational issues, eliminating the need for manual review while improving detection accuracy and reducing time consumption.

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

Solution Approach 2:

The patent introduces an image generation module that converts raw operation data into visual representations (images). This intermediary step enables automated analysis by transforming complex numerical data into visually interpretable formats that can be processed by image recognition algorithms, thereby reducing manual analysis time while maintaining detection reliability.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If detailed operation data is collected from construction machines, then operational status can be monitored, but data security risks increase

Engineering Contradiction:
Improveoperational status monitoring accuracyVSAvoiddata security risk
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent extracts only the essential operational parameters needed for monitoring and analysis, rather than collecting and storing all raw data. By selecting and analyzing only critical data points, the system maintains monitoring accuracy while minimizing data security risks associated with storing and transmitting large volumes of sensitive operational data.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent creates visual copies (images) of operation data for analysis purposes without storing the original raw data. This allows the system to maintain operational status monitoring accuracy through image-based analysis while reducing data security risks by avoiding storage of sensitive raw operational data in centralized repositories.

Inventive Principle:
Principle #26Copying

3Productivity

If operation data is transmitted to a server for analysis, then centralized processing can be achieved, but communication failures may occur

Engineering Contradiction:
Improvedata processing efficiencyVSAvoiddata analysis reliability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent segments the data analysis function into two parts: centralized image generation at the server level, and distributed image recognition at the terminal device level. This segmentation allows efficient centralized data aggregation while improving reliability by enabling local analysis capability that can operate independently when communication fails.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary data processing and image generation at the server before transmission to terminal devices. This preliminary action ensures that analysis-ready data is prepared in advance, improving processing efficiency while allowing terminal devices to perform local recognition without requiring continuous server communication, thereby enhancing reliability.

Inventive Principle:
Principle #10Preliminary action

4Measurement precision

If complex analysis algorithms are used to process operation data, then analysis accuracy improves, but processing speed decreases

Engineering Contradiction:
Improveoperational issue detection accuracyVSAvoiddata processing speed
Core Design Contradiction:
Measurement precisionVSSpeed

Solution Approach 1:

The patent replaces complex algorithmic processing with image recognition technology. By transforming operation data into visual images and using trained neural network models for analysis, the system achieves high detection accuracy through pattern recognition rather than complex computational algorithms, thereby maintaining processing speed while improving accuracy.

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

Data Source

PatentEP4015715B1Operation recording and analysis system for construction machinery
Publication Date: 2026.04.29 HITACHI CONSTRUCTION MACHINERY CO LTD
  • EP4015715B1 patent drawingFigure 1
  • EP4015715B1 patent drawingFigure 2
  • EP4015715B1 patent drawingFigure 3

AI summary

An operation record analysis system for a construction machine is provided which can efficiently extract work contents that may possibly make a factor of decrease in the operation rate from within operation information recorded at the time of operation of a construction machine. The operation record analysis system for a construction machine includes an object sensor that senses an object existing around a machine body. A controller calculates the position of the object on the basis of the information from the object sensor, decides, on the basis of the information from a machine body position sensor and a machine body posture sensor, and the position of the object, whether or not the machine body and the object are close to each other, and adds a result of the decision to the operation information.