Catalogue of Artificial Intelligence Techniques


Jump to: Top | Entry | References | Comments

View Maths as: Images | MathML

Collective Locomotion (Egress Alogorithm)

Aliases: Egress Algorithm

Keywords: collective robotics, locomotion, movement, navigation

Categories: Robotics

Author(s): John Fraser

Although research into collective locomotion is still ongoing, there have already been successful projects, using implementations of the Egress Algorithm. This algorithm is based on a while loop that is passed a location to get to, and during the attempts to reach said location sets its own sub-goals based on the surroundings. Having classified its surroundings into a pre-constructed category, it can then choose and apply the appropriate rule, and as such set the agent a new sub-destination point and move there, checking for possible collisions and updating the agent's location. As such this is a form of self optimization, as the algorithm is continuously updating the best possible sub-route/s to be taken in order to reach the end objective, taking into account other agents in proximity. As a result, the Egress Algorithm has replicated human crowd behaviour in various implementations with fair success, as crowd phenomena such as panic can be replicated by altering the velocity and collision rules for the agents. By doing so, agents become more aggressive in their route planning, and less inclined to negotiate a mutually beneficial route.



Add Comment

No comments.