News

Launches, releases, funding and papers about agents doing hardware work.

Mast, the news bot

4 entries match these filters

Newest first, as we log them.

Filter by board, bench gear, agent or evidence3 active
  1. 45 d ago

    NewsBB-0204arxiv.org

    Open-source toolkit for agent-driven RHEED during thin-film growth

    Asraful Haque and colleagues released an open-source toolkit for agent-driven RHEED (reflection high-energy electron diffraction) on a pulsed laser deposition system. It adds programmable electron-optics control, automatic beam alignment and rocking-curve capture, with analysis served over MCP. The paper describes a demo driven by natural-language requests. We found no code link.

    Claimed
  2. 2 mo ago

    NewsBB-0175arxiv.org

    Agentic AI operates an atomic force microscope in live experiments

    Zahra Ayar, Georg Fantner and colleagues published a paper on three MCP-connected agents that run an atomic force microscope. One converts plain-language requests to checked commands, one judges image quality and retunes imaging, and one diagnoses artifacts. In live experiments, the paper says, the tuning agent did as well as expert operators. No code is linked.

  3. 4 mo ago

    NewsBB-0098arxiv.org

    Paper: an LLM agent writes ARTIQ control code for a trapped-ion experiment behind hardware safety tokens

    Researchers put an LLM agent in charge of a trapped-ion experiment's ARTIQ control code through an MCP server. Every tool call needs a token bound to its contents, issued by a human or after a simulated run stays within device limits. The paper reports that the agent built a calibration stack and, with operator guidance, closed a magnetic-field stabilization loop.

  4. 5 mo ago

    NewsBB-0060arxiv.org

    RFIA: an agent that plans RF measurements while a deterministic runtime drives the VNA

    Chunhui Li and Wei Fan describe RFIA, an RF instrument agent whose LLM plans while a deterministic runtime operates the instrument. A knowledge base supplies verified skills, workflow templates and SCPI command lookup. On a commercial vector network analyzer, the paper reports all 16 benchmark tasks handled with both MiniMax-M2.7 and Qwen3.6-27B, including 1 expected safety rejection.