Drug discovery describes the discipline of identifying new drugs to fight diseases through a series of preclinical and clinical trial studies. Drug candidates are being evaluated after each study phase, and promising drugs are being tested in subsequent rounds. Due to modern laboratory standards and advances such as high-throughput screening, a large number of potential drugs may be tested in initial study phases. This calls for (partially) automated evaluation of test results. We have experience in building automated evaluation pipelines, which significantly simplify the candidate selection, especially during early study phases.

Scheduling problems occur in different forms and shapes: optimizing load balancing in servers, planning optimal delivery routes of freight, or stocking warehouses to ensure flawless production chains. We have experience in translating business requirements into mathematical optimization problems, and to solve them using powerful software frameworks and metaheuristic optimization methods. Furthermore, large scale scenarios are challenging due to the NP-hardness of the underlying problem. We use our machine learning expertise to reduce the problem complexity, allowing for the development of more efficient solvers.

We developed the open-source software OpenTidalFarm for simulating and optimising tidal turbine farms. The positioning of the turbines in a tidal farm is a crucial decision. Simulations show that the optimal positioning can increase the power generation of the farm by up to 50% and can therefore determine the viability of a project. However, finding the optimal layout is a difficult process due to the complex flow interactions. OpenTidalFarm solves this problem by applying an efficient optimization algorithm onto an accurate flow prediction model.