Development of AI toolbox for Calibration Services

• Unité de recherche: METAS
• Numéro de projet: F-1040.30126

Mots-clé (Anglais)

Confidence, Condition Monitoring, Interdisciplinary Framework, Sensor Network Calibration

Description succincte (Anglais)

Digitalization triggers new requirements for metrology and, in return, enables the realization of novel measurement capabilities. METAS is building new competencies to prepare for the upcoming change. The present project is the first step of the digital transformation. The aim is the development of a toolbox based on artificial intelligence (AI) techniques, which should enable the creation of new services and the adaptation of old services for the digital market across all technical fields. The AI tools will initially support further development of specific in-house technologies. The toolbox will focus on confidence evaluation of simulations and models used during calibration, condition monitoring (CM) of measurement setups and AI-aided validation and verification (V&V) for calibration procedures. Finally, the present project will define the fundamentals for the AI supported evaluation of sensor networks, towards subsequent projects of the digital transformation

Objectifs du projet (Anglais)

The aim of this project is to set the bases for metrological services in the digital age:

1. To establish the framework for interdisciplinary tasks (FFIT) as the common institutional ground of calibration services for the digital market, to develop the shared strategy for the digital transformation of METAS services. A first step for this framework will be to develop and implement a portable Digital Calibration Certificate (DCC).
2. To develop an AI toolbox for measurement and calibration services, focusing on METAS technologies. Based on data-driven innovations, confidence evaluation tools for simulations (e.g., for CT) will allow better measurements services. In addition, model selection tools (e.g., BRDF-model for Appearance applications) and CM tools will be developed to optimize measurement and calibration setups.
3. To define the fundamentals to formalize and propose standards for calibration services of sensor networks. As a first step towards calibration of measuring networks, calibration services for sensor networks will be evaluated and simple standards will be proposed (e.g. for AI-aided calibration of low cost air quality sensor networks).

Mise en oeuvre et application (Anglais)

The digital transformation affects all areas of METAS. The project is intended to lay the foundations for adapting the services of individual laboratories to the changing needs of customers in a timely manner. The project is designed so that its results can be applied early in the project in individual laboratories. The project should help to exploit the potential of new digital technologies in the work of laboratories and, on the other hand, to meet changing customer needs on time. With this the project, the individual laboratories will handle the change in a coordinated manner, and important development steps may be better identified.