Domestic PLC+GPU Dual-Core Edge-Intelligent System with Integrated Sensing and Control
An edge intelligent-control hub built on a domestic, self-sufficient technology stack delivers a secure closed loop from sensing to millisecond-level control, alongside panoramic visualized management of equipment status
Project Overview
Existing cloud-based diagnostic solutions suffer high communication latency and struggle to translate promptly into low-level control commands; moreover, core software and hardware remain heavily dependent on foreign suppliers, making secure and controllable deployment difficult. This project builds a domestic PLC+GPU dual-core edge-intelligent system with integrated sensing and control.
Research Objectives
An edge architecture for lightweight model deployment and real-time closed-loop control
A millisecond-level secure closed loop linking domestic edge devices to PLCs
A CNC-integrated monitoring platform built on a domestic, self-sufficient technology stack
Methodology
Heterogeneous compute acceleration at the edge-computing hub is paired with millisecond-level, low-latency hardwired safety interlocks at the edge intelligent-control center; degradation trajectories are tracked precisely across complex operating conditions, with efficient isolation of multiple fault types.
Technical Approach: How It Works
- 1
Architecture Redesign: Three Tiers to an Edge Closed Loop
The traditional "field sensor → edge gateway → cloud diagnosis → command dispatch → PLC" chain incurs second-scale latency; it is redesigned into an integrated PLC edge closed loop where sensor feature extraction, health assessment with confidence gating, and control response all execute locally, cutting loop latency below 50ms.
- 2
Health Assessment and Confidence Gating
Equipment degradation trajectories are finely tracked at the edge, triggering normal / decelerate / emergency-stop responses by health-grade tier, backed by hardwired safety interlocks.
- 3
Domestic Heterogeneous Compute + Self-Sufficient Tech Stack
The PLC+GPU dual-core edge system is fully domestically developed and controllable end-to-end, with lightweight models deployed on domestic edge hardware.
- 4
CNC-Integrated Monitoring Platform
Panoramic visualized management of equipment status, with efficient isolation of multiple fault types across complex operating conditions.
Figures: Methods & Results
Key Results
Millisecond-level closed loop from sensing to control, replacing high-latency cloud-based solutions
A panoramic, efficient visualized management platform for equipment status has been deployed
Core links are fully domestically developed and controllable
Real-Time Optimization of Production Operations