Information Flow Dynamics In Foraging Honey Bee Colonies
Andreas Loengarov, Computer Science, University of Paisley, United Kingdom
We consider a bee colony as a dynamical system gathering information from the environment and adjusting its behaviour to it. Intelligent decision-making emerges from the enhancement of the communication level between individuals.
We investigate information flows, and the formation and evolution of information-mapping patterns in the colony during its foraging activity. Similarly to the theory of emergence of self-replicating units (hypercycles), an autocatalytic mode determines the process. That is, when unemployed foragers are recruited, they then become recruiters themselves for a particular source – the information is able to reproduce itself. The mechanism is, thus, defined by positive feedback, or autocatalytic reinforcement of useful information. The dissipation of information occurs when foragers abandon unrewarding food sources. This ensures change in the system and prevents its remaining stuck in local optima. The natural physical limitations of the hive restrict the number of information carriers. The replication of information is thus naturally limited. All these factors lead to the competition of information and, as a result, only the most useful information, i.e. the route to the most profitable nectar source, survives.
The phase transition from disordered to ordered foraging, and the trade-off between exploration and exploitation are analyzed. An analogy to the formation of an employment market, as well as possible applications to design insect-like robot swarms, are considered.