Real-Time Control via Serialized State Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing real-time control systems in server-based environments require persistent memory and processes for each control system, leading to inefficient memory usage, scalability issues, and fragility in cloud environments due to high start-up and shutdown requirements.
Innovation Solution
Implementing a scalable real-time control system that utilizes FIFO queues and serialized controller data structures, allowing control cycles to run on any server without persistent memory, using chained HTTP requests and distributed computing to manage actuator positions across multiple servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If persistent memory and processes are allocated for each control system in a server-based environment, then control reliability is improved, but memory usage increases significantly and scalability deteriorates
Solution Approach 1:
Multiple control systems share a single server instance and pooled memory resources through the use of serialized controller data structures that can be loaded and saved to persistent storage, eliminating the need for dedicated memory per control system while maintaining control reliability through periodic serialization to disk
Solution Approach 2:
Controller state is copied to serialized data structures that can be persisted to storage media, allowing the system to recover control state without requiring continuous in-memory persistence, thus reducing memory usage while maintaining reliability through recoverable state copies
2Stability of the object's composition
If persistent processes are instantiated for each control system, then control stability is improved, but system complexity increases and cloud environment adaptability deteriorates
Solution Approach 1:
The system transitions from static persistent processes to dynamic state serialization, where controller state can be persisted to storage and reloaded as needed, allowing the system to adapt to cloud environment dynamics including server startup/shutdown cycles while maintaining control stability through recoverable state
Solution Approach 2:
A single server instance can host multiple control systems through shared memory and serialized state management, making the system universally deployable across cloud environments with varying resource availability, while maintaining control stability through isolated serialized controller data structures
3Manufacturing precision
If dedicated server processes are used for each control system, then control precision is maintained, but processing efficiency decreases due to wasted server capacity
Solution Approach 1:
Multiple control systems are merged into a single server process with shared memory resources, eliminating wasted server capacity while maintaining control precision through isolated serialized controller data structures that prevent interference between control systems
Data Source
AI summary
Systems and methods for real time control are provided. One embodiment includes executing a first control cycle that includes retrieving an input message, where the input message includes an identifier of a network device with an input and an actuator and retrieving a context object that includes a serialized controller data structure from a previous control sample. The first control cycle may additionally include retrieving data related to the input for the network device and bundle the data related to the input with the serialized controller data structure into a run message, de-serializing the serialized controller data structure, and using the controller data structure in conjunction with the data related to the input to determine a new actuator position of the actuator. In some embodiments, the first control cycle includes updating and serializing the controller data structure and causing the actuator to be set to the new actuator position.