Work Machine Operation Identification via Frequency Pattern Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for identifying the operation of work machines, such as hydraulic excavators, are not accurate due to reliance on expensive sensors and cannot effectively classify all activities performed by these machines, leading to suboptimal performance and productivity.
Innovation Solution
A control system that includes an operator input device generating a command data stream, which is converted into a frequency data stream to identify patterns and classify operations, allowing for the triggering of events and generation of machine application profiles based on these classifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If expensive sensors are used to sense operational characteristics, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces expensive physical sensors with a software-based pattern recognition system that processes existing command data streams. Instead of using additional hardware sensors to detect operational characteristics, the system uses signal processing algorithms to extract patterns from the operator input device data, thereby eliminating the need for complex sensor infrastructure while maintaining or improving detection accuracy.
Solution Approach 2:
The patent creates a virtual copy of operational characteristics by deriving them from command data streams through pattern analysis. Rather than directly measuring physical operational parameters with sensors, the system generates replicated information about machine operations by analyzing the temporal patterns in operator commands, achieving measurement functionality without the physical sensors.
2Device complexity
If partial solution methods are used to analyze operator input devices, then device complexity is reduced, but measurement precision and operational identification accuracy deteriorate
Solution Approach 1:
The patent makes the operator input device analysis system universal by designing a single pattern recognition framework that can identify multiple different operations and activities. The same command data stream processing system handles various operation types (digging, loading, transporting, etc.), eliminating the need for separate specialized analysis systems for each operation while maintaining high classification accuracy across all operation types.
Solution Approach 2:
The patent improves operational identification accuracy by transforming the command data stream into the frequency domain using Fast Fourier Transform (FFT). This parameter change from time domain to frequency domain allows for better pattern recognition and operation classification, enabling the system to accurately distinguish between different operations based on their characteristic frequency patterns.
3Adaptability or versatility
If work machines are used for a wide variety of tasks, then adaptability is improved, but the ability to accurately identify all activities deteriorates
Solution Approach 1:
The patent implements a dynamic pattern recognition system that adapts to different operations and activities through continuous learning and pattern updates. The system maintains a database of operational patterns and uses machine learning algorithms to update its classification capabilities as new operation patterns are encountered, enabling accurate identification across a wide variety of tasks without requiring manual reconfiguration.
Solution Approach 2:
The patent incorporates feedback mechanisms where the system continuously monitors and refines its operation identification accuracy by analyzing the relationship between command patterns and actual machine behavior. This feedback loop allows the pattern recognition algorithm to improve its classification accuracy over time as it processes more data from diverse operations, maintaining high accuracy despite the machine's versatility.
Data Source
AI summary
Systems and methods are disclosed for identifying operations of a machine. The system includes a work tool and an operator input device configured to receive input indicative of a desired movement of the work tool and to generate a command data stream associated with the received input. The system also includes an actuator configured to move the work tool according to the command data stream and a controller in communication with the operator input device and the actuator. The controller is configured to convert the command data stream into a frequency data stream and identify a pattern in the frequency data stream. The controller is also configured to make a classification of a current operation of the machine as one of a plurality of known operations based on the identified pattern. The controller is further configured to trigger an event associated with the current operation of the machine.