These pages contain links and notes on Artificial Life and our research activities towards understanding it. Artificial Life (ALife) is a sub-branch of Artificial Intelligence with connections into Complex Systems and other areas of Science.
The work of our group is focussed on understanding quantitatively how artificial life systems explore (and expand) their phase space. We are investigating "animat" approaches to ALife simulation and studying the lifeforms and populations thereof that can be constructed and that arise from simulated evolutionary sequences.
A recent talk to the Institute postgraduate students, entitled "Artificial Life Thumbing its Nose at Thermodynamics" is available as PDF slides. The talk covers some of the main ideas behind our research programme.
A recent article discusses the meaning of artificial life in a not-too-serious manner.
Some of our technical notes related to this work are listed in the CSTN series. See for example:
|
|
A short MPEG movie (2.5 MB) shows the emergent spiral patterns of red predators, blue prey against a background of green corpses.
A list of links to other ALife research activities is also available.
Our AgentWorld is a series of simulation systems for studying digital organisms and their behaviours. AgentWorld is a series of highly optimised simulation programs - mostly run in "batch mode" rather than graphically, aimed at sampling the phase space of ALife-forms.
Of special interest to us is the physics of digital life forms and their environments, and the geometry and topology of different possible artificial life environments.
A closely related field to ALife is that of software agents, which can be treated as a methodology for implementing artificial life forms, but can also be used as a powerful approach to building complex software control systems. We have an ongoing project investigating how agents based software architectures can be used to control distributed robots and other mechatronics "life forms".
Contact us for more information.