[Stardust Colonies]

Stardust Colonies, formerly known as Alien Farm, is a strategy game that I created based on my Undergraduate dissertation 'Inner States in Game Agent Systems'. In the game, the player-controlled units are aware of and affected by both their own memory and preferences and their colony's well-being.

Alien units in the game are able to collect resources and breed. To move, they use an obstacle avoidance algorithm for continuous environments that was based on Bourg's and Seeman's work. The units have an inner state which consists of happiness and activity - two variables which are adjusted for each alien based on what it did previously and whether it liked it or not, as well on how well the entire colony does in terms of the amount of stored food and living space. The objective of the game is to expand the colony to have 30 members and achieve an average happiness of at least 70%.

In Stardust Colonies, each unit is an individual and its behaviour and work performance are affected by its happiness.

Frustrated aliens perform tasks slowly and sometimes disobey the player's orders, but doing their preferred job makes them very efficient. Furthermore, movement of an alien is slowed down if it is tired of frequent travelling of because of its age. Player therefore has to keep a constant eye on all colony members and make sure they stay happy and active. Fuzzy logic is used both for evaluation of a unit's current inner state as well as to adjust its behaviour.

The game was tested by a number of players who found the individualised behaviour interesting and harder to predict. The Kickstarter and Steam campaigns generated some attention, although the marketing and development had to be paused indefinitely due to my other engagements.

Alien Farm in its original state was generally a great success amongst students and academic staff and was presented on the University's open day in May 2009.