Bridging research and industry through AI innovation

This project delivers an accessible toolkit for contextual performance profiling and time-series anomaly detection across industrial machinery.

This project develops data-efficient methods for selecting experiments and optimizing process parameters when trials, simulations or raw materials are costly.

This project validates simulation-based workflows for reinforcement learning-enhanced industrial process control, including fine-tuning RL agents on existing control systems to improve safety and performance.

This project delivers an efficient on-device speech recognition solution optimized for Dutch and Flemish speech, combining strong recognition accuracy with low latency and reduced computational requirements.

This project develops a mobile robot that can navigate residential care environments and assist with hydration by handling bottles and pouring water.

This project combines high-speed sensor data and AI models to detect potential quality issues and improve the stability of laser-based polymer 3D printing.

This project supports targeted agricultural treatment through real-time monitoring and AI, while providing explanations and feedback to farmers to increase trust and technology adoption.

This project provides a toolbox that enables domain experts to inspect and correct AI models through their local explanations, demonstrated for automated seizure detection.

This project enhances the Pythia congestion-forecasting platform with an operational dashboard and geographic visualizations that help grid operators explore and interpret predictions.

This project validates data-efficient optimization methods, including surrogate modelling and multi-fidelity approaches, to reduce the number of costly simulations required for complex engineering design.

This project combines AI with physics-based modelling to create accurate, data-efficient and physically consistent models for faster design and optimization of complex engineering systems.

This project validates a privacy-preserving computer vision approach that masks sensitive visual information while retaining the data needed for accurate industrial AI applications.

This project demonstrates how large language models can be personalized with sensitive data on devices in the home, without transferring that data to the cloud.

This project combines onboard odometry with inter-robot distance measurements to help robots and drones determine their relative positions and collaborate without external positioning infrastructure.

This project combines visual processing, large language models and information retrieval to answer questions and extract information from complex documents.

This project combines visual processing, large language models and information retrieval to answer questions and extract information from complex documents.

This project integrates predictive clustering into the Pythia platform to improve and scale congestion forecasting for low-voltage electricity distribution grids.

This project develops and validates a toolbox that combines machine ontologies, knowledge graphs and data-driven reasoning to provide transparent and interpretable failure diagnosis across heterogeneous industrial machinery.

This project identifies and validates key technological requirements for robust robotic systems that can support labor-intensive orchard operations such as harvesting, pruning, thinning and crop management.