Back to feed
News Story
量子位
1 sources

Bringing PLC Simulation to the Browser: Small Parameter Models Achieve High Engineering Pass Rates

Shanghai Jiao Tong University and the Sema research team have launched SemaPLC, which connects AI agents with the PLC engineering toolchain, enabling a complete workflow from natural language requirements to compilable, verifiable, and simulatable control programs in the browser. The system uses a web IDE and tool suite to generate, verify, and simulate PLC programs, aiming to improve engineering pass rates.

SynthePulse Insight · AI deep readingMembers

SemaPLC: Taking AI-Generated PLC Code from 'Can Generate' to 'Ready to Deliver'

Version 1 · 1 source

Shanghai Jiao Tong University and the Sema team have launched SemaPLC, which closes the loop between AI agents and the PLC engineering toolchain, enabling a complete workflow from natural language requirements to compilable, verifiable, and simulatable control programs. Experiments show that across 7 large models, it ranks first in function-level verification pass rate, averaging 72.6%, with a dynamic behavior score of 52.2, far exceeding baselines.

  • SemaPLC decouples and encapsulates AI agents with PLC compilation, execution, verification, and simulation engineering capabilities, providing a Web IDE and a reusable tool suite.
  • On the function-level benchmark (117 tasks), SemaPLC ranks first in verification pass rate across all 7 models, averaging 72.6%, 8.8 percentage points higher than the strongest baseline.
  • On the project-level benchmark (65 real industrial projects), it achieves a dynamic behavior score of 52.2, 20.8 points higher than the strongest baseline, approximately 1.7 times that baseline.
Open section navigationFrom 'Code Generation' to 'Engineering Closed Loop'

From 'Code Generation' to 'Engineering Closed Loop'

AI-generated PLC code is not new, but industrial sites require programs that are compilable, run correctly, are exception-safe, and are understandable by personnel. SemaPLC's core philosophy is that AI writing code is not the end goal; the engineering closed loop is key.

SemaPLC, developed by Shanghai Jiao Tong University in collaboration with the Sema research team, decouples and encapsulates AI agents with PLC compilation, execution, variable read/write, behavior verification, and process simulation engineering capabilities. It is provided as a tool suite, usable in an IDE or integrated into other systems via command line and MCP services.

The Web IDE integrates requirement input, code viewing, ladder diagram reading, variable monitoring, verification results, and process simulation into a single browser window, reducing tool switching and speeding up debugging feedback.

Free for now

Read the full analysis

3 more sections of analysis, plus the full takeaway

Loading

Credibility boundary

The information in this article primarily comes from a contributed article published by QbitAI, which is a secondary source. All experimental data, model names, benchmark sizes, etc., are from that article and have not been independently verified. The paper is publicly available on arXiv and the code is open-sourced on GitHub, but this article has not verified its contents.

Sources for this version

  1. 把PLC仿真搬进浏览器，小参数模型也能实现高工程通过率量子位

Primary report

量子位

Primary source