Container Processing Assembly Fault Prediction via Timed Sensor Profiles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Container treatment systems face challenges in diagnosing and predicting errors across interconnected units, leading to system failures and reduced production efficiency due to the complexity of coordinating multiple components.
Innovation Solution
A method that records and compares reference characteristics from various sensor devices across different treatment devices in a container treatment system, using artificial intelligence and statistical methods to identify potential errors and adjust operations proactively, thereby simulating human experience in error recognition and prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple treatment devices are coordinated to process containers in sequence, then production output is improved, but system reliability deteriorates due to increased interdependence and potential fault propagation
Solution Approach 1:
The patent implements a feedback mechanism where sensor devices continuously monitor reference values from each treatment device and compare them against stored reference profiles. When deviations are detected, the system generates fault information that triggers automated responses, such as adjusting operating parameters or alerting operators, thereby maintaining system reliability despite high coordination complexity
Solution Approach 2:
The system performs preliminary actions by continuously recording and analyzing reference values before actual faults occur. By comparing real-time sensor data against historical reference profiles, the system can predict potential failures and take preventive measures, such as adjusting process parameters or scheduling maintenance, before the fault propagates through the coordinated treatment devices
2Difficulty of detecting and measuring
If reference values are continuously recorded and compared across multiple treatment devices, then fault detection capability is improved, but device complexity increases
Solution Approach 1:
The patent segments the monitoring system into independent modular components: sensor devices at each treatment device, a central control unit with memory for storing reference profiles, and a comparison mechanism. This segmentation allows fault detection to be implemented incrementally across different treatment devices without requiring complete system redesign, thereby managing complexity while maintaining high detection capability
Solution Approach 2:
The system creates simplified copies of reference value patterns by storing historical reference profiles in memory. These reference profiles serve as templates that are copied and compared against real-time sensor data, enabling complex fault detection through simple pattern matching rather than requiring complex analytical algorithms
Data Source
Figure 1~2
Figure 3a~4
Figure 5~6
AI summary
Method for operating a container treatment plant (1), wherein containers are treated in a first predetermined manner by a first treatment device (2) of this container treatment plant (1), are then transported from this first treatment device (2) to a second treatment device (4) of the container treatment plant (1), and are subsequently treated by the second treatment device (4) in a second predetermined manner, wherein a first plurality of first reference values (RK1) characteristic of the treatment of the containers (10) by the first treatment device (2) is recorded by means of first sensor devices (22a, 24a), and a second plurality of second reference values (RK2) characteristic of the treatment of the containers (10) by means of second sensor devices (42a, 44a) is recorded.and wherein these reference parameters (RK1, RK2) are stored in a storage device (16). According to the invention, the reference parameters (RK1, RK2) are recorded with a time value that is characteristic of the temporal occurrence of the respective reference parameter (RK1, RK2), and a plurality of test parameters (PK1, PK2) are recorded, and from a comparison between at least one of these test parameters (PK1, PK2) and at least one reference parameter (RK1, RK2), at least one piece of information (I) is output which is characteristic for determining a present or future fault condition of the system.