DiscoveryLabNL
Dutch Materials Discovery National Laboratory
DiscoveryLabNL is a national research infrastructure driving a revolution in how materials are discovered. AI proposes designs grounded in quantum-physics-informed digital twins, self-driving laboratories execute them, and advanced multimodal characterisation verifies the results, closing the knowledge chain fromatom to function in a single autonomous loop. It shifts materials science from discovery-by-trial to designing what materials do, not only what they are. Hosted nationally with a central hub at TU/e and DIFFER, it gives every Dutch research group an integrated, AI-orchestrated discovery environment, and anchors European strategic autonomy in the materials underpinning the energy, health, and digital transitions.
Infrastructure and User Services
DiscoveryLabNL integrates three mutually reinforcing pillars into a single discovery infrastructure, based on high-end DIFFER and TU/eexpertise (initially) throughout the chain of knowledge in the domain of materials science: Pilot demonstrators target five scientific frontiers: (i) informationmaterials, (ii) designer catalysts and energy-conversion materials, (iii) actuating polymers and functional soft matter, (iv) neuromorphic and information-processing materials, and (v) recyclable thermosets and circular materials.
- Self-Driving Laboratories: robotic synthesis and processing platforms, automated sample handling, and in-line sensing and feedback systems for autonomous,AI-guided experimentation.
- Multimodal Advanced Materials Characterisation: optical, electron, and X-ray microscopy with spectroscopic techniques, leveraging TU/e's nationally fundedleadership. These modalities form a correlative verification layer, feeding machine-readable structural, chemical, and functional data continuously into thediscovery loop.
- AI Core facility: curated models, including quantum-physics-informed digital twins and surrogate models constrained by known governing equations ofmaterials physics, a FAIR data backbone, dedicated AI scientists, and shared compute and orchestration that close the loop across pillars.
Examples/description of the type of research
DiscoveryLabNL facilitates frontier research at the limits of what materials science can currently reach: the autonomous, closed-loop discovery of materials whosefunction is designed, not found. Today the pace of materials discovery is still set by manual iteration, with each cycle of design, synthesis, characterisation, andanalysis taking weeks. DiscoveryLabNL closes this gap by coupling autonomous experimentation with high-throughput multimodal characterisation andmachine learning, so that hypothesis generation, experimental execution, characterisation, and adaptive learning operate as a single, continuously improvingloop.
Initial pilot demonstrators target five scientific frontiers: (i) information materials, (ii) designer catalysts and energy-conversion materials, (iii) actuating polymersand functional soft matter, (iv) neuromorphic and information-processing materials, and (v) recyclable thermosets and circular materials, reflecting theestablished strengths of the TU/e institutes(ICMS, EIRES, Casimir, DIFFER). Research themes include: design and discovery of advanced functional polymers andself-assembling systems; new materials for the energy transition such as battery components, photovoltaic materials, and electrocatalysts; sustainable andcircular material flows including bio-based polymers and design-for-recycling; and quantum and information materials. The infrastructure equally supportsfundamental investigations into structure-property relationships, mechanisms of self-organisation, and the physics and chemistry of emerging material classes.
How to access the infrastructure?
DiscoveryLabNL will offer three modes of access:
- Physical access: researchers work on-site at the TU/e campus hub, using robotic SDL platforms and advanced characterisation instruments with support fromintegration scientists.
- Remote access: national consortium partners operate federated SDL nodes, and/or submit experiments to the central hub via standardised protocols, with theAI Core orchestrating closed-loop workflows across sites.
- Virtual access: the FAIR data backbone, model catalogue, and AI tooling are available digitally to all registered users through shared compute infrastructure.