Industrial SEP Engine for Stateful Stream Pattern Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional pattern recognition programs for condition-based maintenance and condition monitoring in industrial systems are cumbersome and time-consuming to develop, requiring intensive collaboration between field engineers and data analytics specialists, and often result in suboptimal performance due to the lack of knowledge about industrial assets and unsuitability of data analytics tools for stream processing.
Innovation Solution
A signal and event processing engine (SEP engine) is deployed on target devices within industrial systems, comprising subscriber, processing, and publisher nodes that perform signal and event processing language operations, including stateful and stateless functional operations, to simplify pattern matching and enhance performance by processing signal and event data streams directly on the industrial system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional data analytics tools and programming languages (Python, Scala, Java, SQL) are used for pattern recognition, then analysis capability is improved, but device complexity and programming difficulty increase significantly
Solution Approach 1:
The patent introduces an intermediary layer that translates complex data analytics requirements into simplified patterns. This intermediary mechanism allows field engineers to work with intuitive pattern definitions rather than complex programming languages, while still achieving accurate pattern recognition through the underlying analytics engine.
Solution Approach 2:
The patent creates simplified copies or representations of complex data patterns that can be easily defined and manipulated. Instead of requiring engineers to work with raw data streams using complex languages, the system uses pattern templates that capture essential characteristics in a simplified form that is easier to understand and implement.
2Measurement precision
If conventional data analytics tools are used for stream processing, then analysis capability is improved, but processing performance and resource efficiency deteriorate
Solution Approach 1:
The patent extracts only the essential pattern-matching functionality from complex data analytics tools and implements it as a lightweight, dedicated stream processing engine. By taking out only the necessary components for pattern recognition and removing unnecessary overhead, the system achieves both accurate analysis and high performance on resource-constrained devices.
Solution Approach 2:
The patent changes the operational parameters of the processing system by optimizing for stream processing specifically. The engine is designed with parameters tuned for real-time data flow processing rather than batch processing, enabling efficient handling of continuous data streams with low latency and high throughput on embedded devices.
3Adaptability or versatility
If pattern recognition programs are developed using conventional tools, then functionality is improved, but development time and collaboration complexity increase
Solution Approach 1:
The patent enables field engineers to independently define and implement pattern recognition logic using intuitive pattern definitions rather than requiring collaboration with data analytics specialists. The system provides self-service capabilities where domain experts can directly create, test, and deploy pattern matching programs without needing to learn complex programming languages or work extensively with analytics teams.
4Adaptability or versatility
If generic event processing operations are used, then processing coverage is improved, but processing efficiency and relevance to industrial assets deteriorate
Solution Approach 1:
The patent applies local quality by providing generic event processing operations that can be customized with domain-specific knowledge. The system maintains versatile event processing capabilities while incorporating asset-specific patterns and rules that are locally optimized for industrial applications. This allows the same processing framework to handle both general events and domain-specific patterns efficiently.
Data Source
AI summary
Provided is a signal and event processing, SEP, engine deployed on a target device of an industrial system, the SEP engine including subscriber nodes adapted to receive at least one of signal and event, SE, data streams from sources of the industrial system; processing nodes adapted to perform signal and event processing language, SEPL, functional operations on the received SE data streams according to an SEPL script of the SEP engine, wherein the SEPL script includes predefined stateful event or data pattern matching operations to generate result SE data streams; and publisher nodes adapted to forward the generated result SE data streams to sinks of the industrial system.