The Minerva framework


Our unique approach to building simple and complex scientifc models can be used for both predictive simulation from models, experimental design, and inference from observations


  • Build complex scientific models in a modular way
  • Modelling and analysis is done in a single generic way for the whole experiment
  • Complete transparency and control over model assumptions
  • Tracking and storage of evolving scientific models
  • Simulation and inference/analysis using the same model
  • Control over influence of uncertainties in data as well as models
  • Advanced inference methods: linear and nonlinear optimization, genetic algorithms, Gibbs sampling and a number of different Markov Chain Monte Carlo schemes
  • Automatic model linearisation for fast approximate models around arbitrary control points
  • Data fusion – optimally combine observations from different heterogeneous sensors
  • Experimental design – easily change parameters and aspects of instrumentation, predict future observations and analyse, during planning phase for a new experiment
  • Cloud computing – external computationally heavy programs can run on a local or external cloud through standard service architecture
  • Scientific models can be easily communicated between researchers
  • Runs on all major platforms, including Windows, Linux, and Mac OS.