Third Year Project (BSc)


The physical environment within games and simulations is its foundation, as it is where all interactions between entities take place. Traditionally these are manually developed, often limiting the ability for emergence to appear, as everything has to be scripted in; events such as forest fires and floods do not emerge unless the developers script them in.

Cellular automata can be used to model a variety of things; physical phenomena such as gas and fluid
dynamics, predator-prey interactions and patterns found in nature. Though difficult to predict, the aggregation of locally applied rules within automata can give rise to complex, emergent and sometimes unexpected but desired behaviour.

The final product uses a cellular automaton to model physical substances such as water, sand and tar,
in order to imitate a physical environment. With a small number of rulesets for each substance, physical
properties such as flow, density and viscosity are imitated by the cellular automaton without them being
explicitly defined.

Water has rules for downward movement (imitating gravity) and sideways movement (imitating flow).

Similarly, sand also has rules for downward and sideways movement. Although not a liquid, sand has the ability to form piles which gives it the illusion of flow.