Elevator Malfunction Detection via Relative Behavior Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting elevator malfunctions, such as blockages, often result in excessive false alarms due to misinterpretation of normal conditions, leading to unnecessary maintenance visits and costs, as they rely solely on data from a single elevator without considering the behavior of other elevators.
Innovation Solution
A method that detects malfunctions by analyzing the relative behavior of an observed elevator in comparison to other elevators, using data from sensors to correlate conditions and reduce false alarms by considering both current and normal relative behaviors, with the option to select similar elevators for data correlation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensor-based malfunction detection is implemented in elevators, then proactive service and early detection of malfunctions become possible, but the number of false alarms increases substantially
Solution Approach 1:
The patent combines data from multiple sensors (door sensors, floor sensors, call sensors) and merges them with data from other elevators to form a comprehensive view of elevator behavior. This integration allows the system to distinguish between actual malfunctions and normal operational variations, reducing false alarms while maintaining detection capability.
Solution Approach 2:
The system continuously monitors elevator behavior and uses feedback from both the observed elevator and comparable elevators to adjust malfunction detection. By comparing current behavior against historical data and peer elevator performance, the system learns to recognize true malfunctions from normal variations, progressively reducing false alarm rate.
2Productivity
If malfunction detection systems are deployed to enable proactive servicing, then maintenance costs and downtime are reduced, but device complexity increases
Solution Approach 1:
The system uses existing elevator sensors for multiple purposes: normal operation monitoring, malfunction detection, and behavior baseline establishment. By making the sensor system multi-functional, the patent avoids adding dedicated hardware for each function, thereby limiting complexity increase while achieving proactive maintenance capabilities.
Solution Approach 2:
The elevator system monitors itself using its own sensors and control data, without requiring external inspection systems. The elevator's control unit processes its own operational data and compares it with peer elevators to detect malfunctions, enabling self-diagnosis and reducing the need for complex external monitoring infrastructure.
3Device complexity
If single-elevator data is used for malfunction detection, then system simplicity is maintained, but detection accuracy deteriorates due to false alarms
Solution Approach 1:
The patent adds a temporal dimension by analyzing behavior over time and a relational dimension by incorporating data from multiple elevators. This dimensional expansion allows the system to distinguish between transient anomalies and true malfunctions, improving detection accuracy without proportionally increasing processing complexity.
Solution Approach 2:
The system changes the parameters used for malfunction detection from absolute thresholds to relative comparisons. Instead of detecting malfunctions based on fixed criteria, the system uses dynamic parameters that compare elevator behavior against its own historical performance and peer elevator data, improving accuracy while adapting to varying operational conditions.
Data Source
Figure 1
Figure 2
AI summary
A method and an elevator controller (31) for detecting a malfunction such as elevator blockage in an observed elevator (3) is proposed. The method comprises: acquiring first data during an application phase, the first data correlating with at least one condition in the observed elevator (3); acquiring further data during the application phase, the further data correlating with the at least one condition in other elevators (5); determining a current relative behaviour of the observed elevator during the application phase based on a comparison of the first data with the further data; and detecting the malfunction in the observed elevator (3) based on an analysis of the current relative behaviour. Preferably, an information about a normal relative behaviour of the observed elevator (3) as learned in a machine learning procedure during a preceding learning phase may be taken into account upon analysing the current relative behaviour of the observed elevator (3). The proposed method may enable automatically detecting malfunctions in an elevator while reducing a probability of false alarms.