M1: Artificial intelligence in instrumentation work

The DAQ (Data Acquisition) pole at LAPP develops instrumentation, testing systems, and automation tools for detector electronics used in particle physics experiments. The following two internship topics target Master 1 or engineering students with interests in instrumentation, automation, robotics, or applied AI. The student will choose one topic.

Topic 1 – Laboratory Equipment Detection & Configuration Interface (Python, PyVISA, SCPI)

Context:
Laboratory electronic equipment (oscilloscopes, signal generators, power supplies, multiplexing cards, etc.) increasingly relies on standard communication protocols such as SCPI, controlled via interfaces like USB, GPIB, LAN, or serial links. Designing unified and intelligent control software is crucial for test benches and automated
validation workflows.

Objectives: 
• Develop a Python-based interface for detecting, identifying, and configuring laboratory instruments using PyVISA and SCPI.
• Explore the use of Large Language Models (LLMs) to assist configuration generation and command composition.
• Depending on progress: automate test sequences for production batches of electronic boards (e.g., multiplexing cards) and produce validation reports.

Tasks:
• Familiarization with laboratory equipment and communication standards.
• Development of modular Python APIs for command/control.
• Integration of optional LLM-based assistants for SCPI command reasoning.
• Design of automated test routines.
• Application on real hardware from LAPP production lines.

Keywords :
Robotics, computer vision, VLAM/VLA, visual inspection, automation.

Topic 2 – VLAM for Instrumentation, Control and Visual Inspection (Robotics, Vision, VLAM)

Context:
Visuo-Linguistic Action Models (VLAM/VLMA or VLA) represent the next generation of robotics control: instead of fixed scripts, they enable a robot to see, understand, and act using natural-language prompting combined with visual perception. These models drastically simplify task teaching and enable adaptive manipulation.
At LAPP, such capabilities would benefit several activities, including automated inspection of detector electronics, quality control of production cards, or robotic manipulation for calibration systems.

Objectives:
• Explore the use of VLAM frameworks to teach a robotic arm to perform tasks through prompting.
• Implement visual guidance using cameras and AI-based perception.
• Develop small demonstrators relevant to LAPP’s instrumentation workflows.
Possible Applications
• LAr-Calib projects: manipulation of multiplexer cards, test preparation for upcoming Cabane boards.
• Visual inspection: defect detection on freshly wired boards.
• Documentation scanning: reading QR codes from LADOC / CLAROC for board identification and database insertion.

Tasks:
• Study existing VLAM/VLA frameworks and robot control interfaces.
• Set up a perception pipeline (camera, calibration, basic detection).
• Teach the robot simple manipulation tasks using prompts.
• Prototype demonstrators based on LAPP instrumentation needs.

Keywords :
Robotics, computer vision, VLAM/VLA, visual inspection, automation.

————————————————————————————–

Environment & supervision :
Internships will be hosted at LAPP (Annecy) within the DAQ pole, working closely with engineers and physicists involved in electronics testing, robotic systems, and automation.

Desired profile :
• Interest in instrumentation, robotics, or applied AI.
• Programming skills (Python recommended).
• Curiosity and autonomy in experimental setups.

Duration & location :
• Duration: ~12 weeks.
• Location: LAPP, Annecy.
• Supervision: DAQ pole engineers and physicists

Email:

eric.chabanne@lapp.in2p3.fr

olivier.arnaez@lapp.in2p3.fr