SWISDATA was proud to present its innovative work on AI-supported decontamination technologies, highlighting the LightAIClean project and the resulting intelligent lab assistant, LabIntel, at the Conference on Applied Surface Chemistry (COAST26) in Vienna. The conference provided a key platform for discussing advancements in applied surface chemistry, with the event hosted by TU Wien.
Impressions
Pioneering Light-Based Decontamination
The Light-AIClean project is an ongoing R&D project which was funded by the FFG, aimed at developing AI-supported, light-based decontamination technologies. This critical technology addresses the need for recycling plastic and glass materials while ensuring they meet stringent food-contact regulations. SWISDATA participated a s a partner in this effort alongside TU Wien as coordinator, MONCON GmbH, FOTEC Forschungs- und Technologietransfer GmbH and Redeem Solar Technologies GmbH.
Introducing LabIntel: The Intelligent Lab Assistant
The cornerstone of our presentation was LabIntel, an advanced AI tool developed within the Light-AIClean project. LabIntel functions as an intelligent lab assistant designed to optimize complex experimental spaces and drive scientific discovery using artificial intelligence (AI).
Key capabilities demonstrated include:
- Advanced Optimization: Utilizing Bayesian Optimization, LabIntel's predictive model estimates outcomes and associated uncertainty, supporting single- and multi-objective optimization.
- Knowledge Integration: It incorporates prior knowledge into the process, enabling faster convergence toward optimal results.
- Explainable AI (XAI): By showing Feature Influence on predictions, LabIntel enhances transparency, allowing researchers to understand why the model reaches a specific conclusion.
- Interactive Chatbot: An LLM-driven chatbot allows users to interact with data and results, helping them to refine optimization goals, setting constraints and to understand complex model behaviors.
LabIntel’s core benefits lie in its ability to efficiently explore complex experimental spaces, combine data-driven learning with expert knowledge, detect potential data issues early, and provide transparent recommendations directly usable by researchers.
LabIntel’s development is co-funded by the Austrian government via the Light-AIClean project (see text and image reference below for the project funded by the FFG) .
The FFG is the central national funding organization and strengthens Austria's innovative power. This project is funded by the FFG. www.ffg.at
More information about the project can be found in the FFG project database: https://projekte.ffg.at/projekt/5134218
Further Information
Learn more about the COAST conference at https://coe-mecs.at/coast/ and explore our concept paper for LabIntel here: https://labintel.org
