Aug 2026· Applied Sciences· 0 citations· 22 references
TL;DR
ChemAutoAgent is presented, a multi-agent system that converts text-based instrument manuals into tested and reusable laboratory instrument drivers through a staged, evidence-traceable pipeline and demonstrates the feasibility of reusing published drivers in a multi-device workflow.
Abstract
Instrument control remains a practical bottleneck in laboratory automation and self-driving laboratories. Although large language models (LLMs) have shown strong potential in document understanding, code generation, and scientific workflow automation, most existing systems assume that ready-to-use instrument interfaces are already available. However, converting heterogeneous device manuals and communication protocols into tested, reusable software drivers therefore still requires substantial manual effort and iterative hardware-level debugging. In this work, we present ChemAutoAgent, a multi-agent system that converts text-based instrument manuals into tested and reusable laboratory instrument drivers through a staged, evidence-traceable pipeline. We evaluate ChemAutoAgent tested and reusable drivers on three representative instruments—a magnetic stirrer, a peristaltic pump, and a Raman spectrometer—spanning three communication protocols: ASCII/NAMUR, MODBUS RTU, and a custom binary-frame protocol. Following iterative testing and repair, all evaluated driver functions passed the predefined tests for connection establishment, parameter configuration, command execution, and data acquisition. The evaluated drivers required between one and four repair cycles, with autonomous operation ratios ranging from 73% to 81%. A cross-device invocation experiment further demonstrates the feasibility of reusing published drivers in a multi-device workflow.
Programmable Logic Controllers (PLCs) are central to industrial automation, yet their programming remains largely manual and time-consuming. While Large Language Models (LLMs) excel at general-purpose code generation, their use in industrial control is limited by tight coupling with physical hardware, constrained I/O signals, and strict safety requirements. We propose a novel approach to PLC code generation that takes as input structured lists of sensors and actuators from electrical CAD tools, together with the required control functionality. The task is formulated as generating machine-specific control functions grounded in available signals, ensuring physical implementability. The system is based on a single LLM adapted through multi-stage fine-tuning and enhanced with Retrieval-Augmented Generation (RAG). Experiments on error detection, code fixing, and code generation show that fine-tuning significantly improves performance. RAG provides strong gains in error detection and code generation, while offering limited benefit for code fixing, highlighting its task-dependent effectiveness.
Teresa Cacciapaglia, D. Ruffino, D. Naso et al.· International Conference on...· 0 citations
Autonomous wet-lab experimentation requires more than plausible protocol text: biological intent, quantitative procedures, device constraints and experimental feedback must remain aligned from protocol and SOP design to code and physical execution. We developed ProtoPilot, a self-evolving multi-agent system, together with an expert-grounded benchmark and evaluation framework for testing this conversion as an experimental automation problem. The framework spans 294 synthetic-biology and molecular-biology tasks derived from 98 gold-standard protocols, wet-lab expert rubrics, device-level validity gates and real experimental tests. ProtoPilot incorporates layer-wise verifiability, multi-agent orchestration and a runtime-updated skill library to generate protocols, expand SOPs, synthesize SDK-compliant code and revise workflows from wet-lab feedback. It achieved a Top@3 expert-preference rate of 90.2%, an overall protocol-to-code gate pass rate of 89.5% and an Opentrons pass rate of 88.24%, compared with 32.35% for OpenTrons-AI. Wet-lab validation produced interpretable readouts, Sanger-confirmed products and feedback-corrected PCA-assembled DNA targets, establishing a verifiable route to autonomous experimentation. Together, these results show that the evaluation framework captures execution-relevant requirements for autonomous wet-lab automation, and that ProtoPilot can meet them by converting protocol and code generation into validated execution and feedback-guided revision.
Yankai Jiang, Wei Tang, Haoran Sun et al.· arXiv.org· 0 citations
Scalable Modular Instrumentation for Laboratory Experiments (SMILE) is a lightweight framework for the rapid development and networking of laboratory instrumentation demonstrated with a low-cost Arduino micro-controller. The approach extends the fast-prototyping paradigm of Arduino by enabling a seamless transition from standalone devices to distributed, network-accessible systems without requiring complex control infrastructures. The system architecture follows a simplistic and intuitive development workflow: devices are first implemented and defined through a human-readable serial interface, which is then reused without modification by a Python-based driver. The driver can be directly accessed or enabled as a network service via ZeroRPC, allowing transparent remote access to instrument’s functionality. A key feature of SMILE is the automatic mapping of serial commands to Python functions, which facilitates immediate integration of newly defined device commands into higher-level control and automation workflows. SMILE design allows heterogeneous system integration with both custom-built instruments and laboratory equipment with standard interfaces (e.g., GPIB, RS232/485, USB, Ethernet) to be incorporated into a unified distributed system through lightweight software layers. Presented demonstrator examples and test results show that SMILE provides a lightweight and accessible approach for physics laboratory automation, conceptually inspired by distributed control systems such as TANGO and EPICS, while remaining focused on small-scale experiments and rapid prototyping.
K. Kamenov, Viktor Angelov, L. Karlov et al.· Automation· 0 citations
Software engineering research spans many activities: surveying literature, building software, running experiments, writing papers-that typically require different tools and expertise at each stage. We present a ScholForge - hierarchical multi-agent LLM system that tackles this end-to-end. Manager agent delegates to five specialists (Librarian, Proposer, MLE Solver, Writer, Reviewer) through a Code-Act execution loop, with external capabilities exposed via three Model Context Protocol (MCP) servers. Given a natural-language research goal, the system produces working code, experimental results, and a compiled manuscript without human intervention. We present one representative case study: Prompt2Edge, an autonomously developed agentic ML pipeline for edge devices that achieved 86.71% test accuracy on Speech Commands with a 17.8k-parameter model and 0.9638 Macro-F1 on UCI HAR, consuming approximately 5 M tokens and completing the full pipeline in under 5 hours.
A. Stepin, Boris Tolstokulakov, Vladislav Kulikov et al.· SIGSOFT FSE Companion· 0 citations
The rapid development of photonic integrated circuits (PICs) is shifting the design flow from traditional graphical user interface (GUI)-based methods to script-based methods for higher flexibility, portability, and maintainability. However, script-based design introduces new challenges, requiring designers to possess additional proficiency in tool application programming interfaces (APIs) and programming. It also demands greater effort and time because it is inherently less intuitive and more complex than GUI-based methods. As PICs grow in scale and complexity, the productivity gap between design needs and manual scripting capabilities continues to widen. To address this gap, we introduce PICopilot, the first large language model (LLM)-based agentic framework that assists in PIC design via automated design script generation from natural language instructions. PICopilot leverages a multi-agent architecture with a feedback mechanism and a specifically designed retrieval-augmented generation (RAG) pipeline, achieving a high success rate and reliability. Experimental results on a benchmark of diverse PIC scripting tasks demonstrate that PICopilot successfully completes all 48 tasks and outperforms other LLM-based approaches without incurring substantial extra latency or cost, even solving 21 more tasks than the advanced GPT-5 model with a general RAG pipeline.
Xiaohan Jiang, Zeyu Li, Wei Zhang et al.· 0 citations
Industrial device commissioning requires engineers to manually extract hundreds of protocol-specific parameters from heterogeneous PDF manuals and transcribe them into supervisory control systems, a time-intensive, error-prone workflow. This paper presents SysName, a production-oriented pipeline that automates device configuration end-to-end for Modbus RTU, OPC-UA, Profibus DP, and CANopen. It builds a hybrid dense-sparse retrieval index augmented by an ontology graph derived from ECLASS, AAS, and SOSA/SSN, using a BGE-M3 encoder with a cross-encoder reranker to surface relevant manual passages. A local LLM (T=0.1) generates ontology-aligned JSON-LD configurations via protocol-specific prompts and a four-step repair pipeline. A two-stage abstention gate, combining a reranker-score threshold and an IRI resolution ratio, blocks unsafe LLM invocations and filters low-coverage configurations before SHACL validation. On a gold set of 28 field-level queries, the hybrid retriever reaches 0.96 HitRate@10, and the reranker raises MRR@10 from 0.56 to 0.63 with perfect score separation for abstention. The generator attains field-level F1=0.87 with exact match on 9 of 12 runs. End-to-end runs on an H100 GPU complete in 2.6-6.6s per device with zero unsafe writes and zero silent failures on a five-device benchmark; every unsuccessful run is flagged by abstention or deployment verification. Component-wise evaluation localises the single systematic failure to OPC-UA generation, invisible to end-to-end metrics alone. A case study commissions a physics-simulated Universal Robots UR5e robot from unmodified vendor documentation (254-page manual, 8-page register list, 496 chunks), reaching field-level F1=1.0 over three runs with read-back and joint-consistency verification. An ablation study and comparison with five industrial-LLM systems complete the analysis.
A. Ganie, Saad Ezzini, Naveed Farooz Marazi· 0 citations
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduAug 17, 2026
A USAF cadet and a Lincoln Laboratory researcher found AI chatbots can help nontechnical service members produce viable software applications for their unique problems.
