Agent-based models

Gaining insight into Complex Adaptive Systems through simulation

In recent years, a new approach to modeling and simulations has been gaining traction. While the theoretical foundation and many implementations have been around for a long time, it is recent advances in computational power, which is now unlocking the potential of agent-based models.

Agent-based models addresses challenges in traditional simulation and computation approaches when it comes to complex adaptive systems. By implementing a bottom-up approach with a focus on local, individual behaviour of the actors and having multiple instances of these agents run in a simulation, the properties of the complex system emerges. They are no longer the result of global knowledge and coordination but the result of individual knowledge, behaviour and interactions.

Our work with agent-based models

We are software developers with expertise in implementing efficient algorithms and optimizing data structures. We work with scientists and researchers to optimize their agent-based models, so that they may realize greater benefits from the models. We have worked with ecological models simulating individuals in their shared environment. These models are driven by GIS data to define the habitat and its parameters.

NetLogo to Repast Simphony migration

We recognize that NetLogo is a hugely popular environment for implementing agent-based models. We have also experienced, that while it provides an easily accessible foundation for getting started as well as extensions for many common scenarios, there are limitations in the flexibility offered in model implementation. We have specialized in the Repast Simphony environment, which offers greater flexibility in the model implementation.

Our service offerings includes migration of NetLogo based models to Repast Simphony.

Performance optimization

We are experts in optimizing the performance of Repast Simphony based models. By applying lessons learned in years of implementing and optimizing systems for some of the world's most demanding enterprises, we have managed to greatly enhance the performance of agent-based models.

Using tools to gain insights to the performance bottlenecks of the model, we use this to focus the optimization efforts. This ensures that effort will be spent where it will have the biggst impact on the performance. This often involves optimizing internal data structures and algorithms, while still producing results in accordance with the original model.

With a combination of migrating the model from NetLogo to Repast Simphony and optimizing the model, we have seen results of models executing more than 10 times faster.

Continued model development

When it comes to continued development of the model, we are also your reliable partner. We focus on the software development, while you can focus on scientific research and the theoretical development of the model.

Let's talk about your model

Interested in discussing how we may help with your model? Book a free online meeting or use our contact form to discover how we can help.

  Book an online meeting