Exploration of Approaches for Robustness and Safety in a Low Code Open Environment for Factory Automation
Gustavo Quiros A., Yi Peng Zhu, Tao Cui, Shaokai Lin, Marten Lohstroh, Edward A. Lee
TL;DR
The paper investigates robustness and safety in low-code, open environments for factory automation by leveraging Lingua Franca (LF) as a deterministic, time-aware coordination language. It situates LF relative to IEC 61499 and outlines how LF's reactor-based, time-tagged model can enable safe, synchronized, distributed automation across heterogeneous OIE modules, with discussion of safety, timing, and potential dataflow extensions to Behavior Trees. Key contributions include a LF-based architecture for networked safety, exploration of multicore and realtime scheduling strategies (including PretVM for hard real-time guarantees), and application examples such as a vision-guided robotic assembly use case and the Princeton Future of Automation Lab concept. The work highlights open questions in regression testing and deterministic pub/sub, proposing pathways to bring software-defined automation closer to industrial certification requirements while preserving reliability, predictability, and timing accuracy in distributed settings.
Abstract
This report is a compilation of technical knowledge and concepts that were produced by the authors and additional contributors in the context of the collaboration projects "Abstraction Requirements for Language of Choice in Industrial Automation" (FY21-22) and "Approaches for Robust and Safe Low-Code" (FY23-24) from Siemens Technology and the University of California, Berkeley. The primary objective of these projects was to assess Siemens Open Industrial Edge (OIE) engineering capabilities by defining a concept that ensures the satisfaction of coordination and safety requirements when using disparate OIE modules. The objective was to use the Lingua Franca (LF) coordination language to demonstrate how to address challenges in: 1. engineering modular, distributed, and flexible automation solutions that ensure, by design, robust and safe operation1; 2. the use of IEC 61499, the event driven execution model for specifying the execution order of OIE modules (defined as function blocks); 3. support large-scale distributed OIE automation solutions, and eventually 4. define optimal solutions with synchronization and time-optimal mechanisms.