One of the biggest challenges in modelling dynamic biological systems is being able to capture the physical-spatial aspects of entity and system behaviour. Equation-based approaches are generally unable to cope with spatio-temporal dynamics when the system is highly complex (i.e. contains many interacting components). Simplification (e.g. discretisation or approximation) can overcome this, but may result in both over-computation and excessive loss of accuracy.

The agent-based framework is concerned with behaviour at a highly abstract level and hence does not itself assume anything about the representation of space (agents can even be non-physical, making spatial representation irrelevant). However, in biological systems modelling, agent behaviour can be expressed in physical-spatial terms, making the simulation of spatio-temporal dynamics more tractable.