Distributed Intelligent Agents for Industrial Process Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing process automation control systems lack efficient autonomous and semi-autonomous control capabilities, particularly in industrial plants, where coordination across multiple controllers is challenging due to the shift from control engineering to network-oriented systems.
Innovation Solution
A system utilizing short-range, low-power communication protocols for deploying and configuring network nodes with intelligent agents or avatars, enabling in-situ control and management of assets through a communication fog that facilitates master-slave relationships and self-configuring networks for optimized process control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If traditional control engineering approaches are used with centralized coordination, then system stability is maintained, but automation capability and real-time decision-making are limited
Solution Approach 1:
The patent segments the centralized control system into distributed intelligent agents deployed at network nodes throughout the industrial plant. Each agent independently manages local control functions, eliminating the need for complex centralized coordination while maintaining system stability through decentralized decision-making.
Solution Approach 2:
Intelligent agents are equipped with autonomous capabilities to self-manage control loops, self-diagnose issues, and self-coordinate with other agents through the communication fog. This self-service approach enables high automation levels without proportionally increasing coordination complexity, as agents independently handle their own control tasks.
2Adaptability or versatility
If network-oriented systems are implemented to improve automation, then control flexibility increases, but coordination efficiency deteriorates
Solution Approach 1:
The communication fog acts as an intermediary layer between intelligent agents and the network infrastructure. It provides standardized communication protocols and data exchange mechanisms that enable flexible coordination without time losses, as the fog handles message routing and synchronization efficiently.
Solution Approach 2:
The system pre-configures communication pathways and data exchange formats through the communication fog before coordination events occur. This preliminary setup allows intelligent agents to coordinate flexibly in real-time without incurring coordination delays, as the communication infrastructure is already optimized for their interactions.
3Speed
If distributed intelligent agents are deployed for autonomous control, then real-time decision-making improves, but system complexity increases
Solution Approach 1:
The patent implements universal intelligent agent templates that can be deployed across multiple network nodes with consistent functionality. Each agent follows standardized protocols and communication patterns defined by the system architecture, which reduces individual node complexity while enabling fast real-time decisions through distributed intelligence.
4Use of energy by moving object
If short-range low-power communication is used for in-situ control, then energy consumption decreases, but communication range is limited
Solution Approach 1:
The patent compensates for limited communication range by adding spatial dimensionality to the network architecture through the communication fog. Multiple short-range communication links are arranged in a mesh topology, allowing information to reach distant nodes through intermediate relays while each node maintains low power consumption for its local transmissions.
Data Source
AI summary
Systems and methods are described for control and/or asset performance management of a system, in-situ, at an asset. A system application model can be deployed to at least a first network node fixed, in-situ, at the asset using a short-range, low-power communication network. The first network node can execute a master avatar using the deployed system application model and generate, at a second network node, a master-slave avatar using the master avatar executed at the first network node. The master avatar and/or master-slave avatar can be configured to perform control of the system, in-situ, at the first network node and the second network node.