Home
Robotics and Intelligent Systems
High Integrity Software
Engineering Services Partners
News
Contact
 
     
 
> Introduction
   
  > Intelligent Systems
   
  > Intelligent Agents
   
  > Swarm Systems
   
  > Our New Technology
   
  > Our Services
>

Swarm Systems

Other examples of the new technology will be swarm systems. Instead of being a single, complex, individual agent, these systems will consist of a mass of small, simple agents acting together as a swarm. The inspiration for these types of system comes from social insects, such as ants, termites, wasps and bees. Individually a single swarm member can achieve very little. But by acting together, the swarm as a whole can achieve much more. The behaviour of swarms is emergent - a term given to describe how the swarm can collectively achieve complicated tasks that the individual agents are incapable of on their own. This means that we can build each individual agent as a small, simple - and therefore cheap – machine.

If each swarm agent is cheap, swarm systems may be able to achieve tasks at a much lower price than an equivalent complex agent. We may be able to get better value for money out of the system if it is designed as a swarm than as a single larger machine.

Swarms may also be highly reliable; if the behaviour of the swarm is an emergent feature of the whole swarm rather than of any single member agent, then the swarm will be able to continue performing its task if some of its members break down.

Swarms are highly fault tolerant - a property that is very attractive for applications where safety and reliability are paramount. We can make sure that the swarm can tolerate many such failures, and still do its job, and the cost of adding some replacement agents into the swarm to make up the numbers again is much cheaper than doing the same with complex agents.

 

 

 
Image above: Close up image of ant. Swarm systems will be based upon the behaviour of social insects such as ants.
         
2006 Avian Technologies Ltd.