Riveting Machine Fault Diagnosis Using State-Time Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial machines, such as riveting systems, face challenges in diagnosing faults due to complex state machines with numerous states and imperceptible signal changes over time, making it difficult for operators to determine the current state and identify faults.
Innovation Solution
A method and system that utilize a state machine to model industrial machine operations, allowing for the identification of current and previous states, determination of fault indicators based on state times, and generation of diagnostic information for user interfaces, including a circular buffer for storing state histories and I/O signal changes to aid in fault diagnosis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a state machine models the functions of the industrial machine, then the machine's operating states can be tracked, but the complexity of diagnosing faults increases due to the large number of possible states and signal changes
Solution Approach 1:
The patent segments the complex diagnostic problem by dividing it into manageable components: identifying the current state, identifying previous states, determining state times, and comparing against expected values. This segmentation allows operators to systematically diagnose faults without being overwhelmed by the overall system complexity.
Solution Approach 2:
The system performs preliminary actions by maintaining a history of states and state times, and by pre-establishing expected time periods for state transitions. When a fault occurs, this pre-collected information is immediately available for comparison, enabling rapid diagnosis without requiring complex real-time analysis of the entire state machine.
2Measurement precision
If the machine monitors all states and signals, then fault detection accuracy improves, but the information processing load increases making it difficult for operators to identify the specific fault state
Solution Approach 1:
The patent extracts only the critical information needed for diagnosis: the current state, relevant previous states, and their corresponding state times. By taking out only this essential data from the complete state machine information, the system provides accurate fault diagnosis without overwhelming operators with unnecessary details.
Solution Approach 2:
The system applies local quality by providing detailed diagnostic information specifically at the point of fault occurrence. Instead of presenting all state machine data uniformly, it focuses information locally on the relevant states and time periods surrounding the fault, making it easier for operators to identify the specific problem area.
3Reliability
If the system tracks incremental signal changes over time, then fault conditions can be detected, but the timing and presence of signals may change imperceptibly making it difficult to determine when a fault occurred
Solution Approach 1:
The system performs preliminary action by continuously tracking and storing state times as the machine operates. This ongoing record-keeping means that when a fault condition is detected through signal monitoring, the timing information is already available, eliminating the need for complex retrospective analysis and enabling immediate fault identification.
4Ease of operation
If the user interface refreshes frequently to show real-time diagnostic information, then operators can monitor machine states more closely, but the interface performance and responsiveness may degrade
Solution Approach 1:
The user interface extracts and displays only the essential diagnostic information: current state, previous states, and state times. By taking out only this critical data for display rather than refreshing the entire interface, the system provides real-time monitoring capability while maintaining interface performance and responsiveness.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for diagnosing faults in an industrial machine. The method comprises identifying a first state associated with the industrial machine, identifying at least one second state associated with the industrial machine based on the first state, for each identified second state, determining a state time associated with the second state, determining a fault indicator condition responsive to determining, based on the at least one second state and the respective state times, that the industrial machine spent a time period different to a predetermined time period in at least one of the at least one second states, generating, responsive to determining the fault indication, diagnostic information comprising an indication of the one of the at least one second state, and outputting the diagnostic information in a user interface of the industrial machine.