GI_Forum 2016, Volume 4, Issue 1, pp. 41-52, 2016/06/29
Journal for Geographic Information Science
Advances in sensor technology have triggered the development of model-sensor frameworks that integrate real-time sensor data into simulation models. Successful implementations include sensor streams of physical properties of the environment, and traces of human mobility as captured by mobile devices. Implementation of real-time sensor-data streams from tagged animals is not yet in place. However, these are anticipated in the near future. In this research, we explore a model-sensor framework with a novel agent-based flocking model for homing pigeons and an emulated sensor stream that is derived from a previously recorded GPS track. Results point to conceptual shortcomings but also opportunities in transferring conventional models to an integrated model-sensor framework. From these insights, directions for further research towards automatic model-calibration and adaptive rulesets were identified.
Keywords: spatial simulation; real-time sensors; agent-based modelling; movement ecology
