Using an ingenious approach involving virtual robots that possess evolvable genomes, researchers have identified key factors that may play important roles in determining the manner in which communication arises during the evolution of social organisms.
The findings, reported by a group including Dario Floreano of Ecole Polytechnique Fédérale de Lausanne, in Switzerland, and Laurent Keller of the University of Lausanne, appeared online in the journal Current Biology on February 22nd.
Robots communicate to identify “food.”
Credit: Laboratory of Intelligent Design Systems, Ecole Polytechnique Fédérale de Lausanne
Communication is critical for social organisms to ensure their ecological success, but the evolution of communication is very challenging to study because of the difficulty of performing experimental evolution with social animals, and the lack of fossil evidence for changes in communication skills over time.
In the new work, the researchers studied the changing behavior of 100 "colonies" of ten virtual robots over 500 generations, during which their virtual genomes were subjected to mutation and recombination, mimicking the genetic variation introduced by sexual reproduction. Within this experimental system, the robots could forage in a virtual environment containing food and poison sources that could only be discriminated at close range.
Theoretically, the efficiency of food foraging could be increased if robots transmitted information to one another about food and poison locations—however, under some conditions, such communication could be costly to the individual, who could lose out by advertising the location of a valuable resource. This situation reflects the evolutionary pressures facing social animals in real-world conditions, where communication may be costly or harmful to the individual, but beneficial to the group.
Such pressures set up potentially complex evolutionary dynamics that the researchers were able to investigate in their study: For example, they examined how kinship influences the evolution of communication by performing evolution simulations on robot colonies with different levels of relatedness within the group. Similarly, the researchers used the simulations to look at the effect on communication of different "levels of selection"—that is, the scales at which altruism and competition occur within a given group.
Dario Floreano of the Laboratory of Intelligent Design Systems, Ecole Polytechnique Fédérale de Lausanne
Among the findings was the observation that communication evolves rapidly when colonies contain genetically similar (related) individuals, or when evolutionary selection pressure works primarily on the "group" level. The only scenario in which communication did not result in higher foraging efficiency was when colonies were composed of robots of low relatedness, and in which selection was strongest at the level of the individual. In some cases, these conditions gave rise to the use of deceptive communication signals and a concomitant decrease in colony performance. The researchers also found that once a system of communication became established during the evolution experiments, it tended to constrain the development of more efficient communication systems.
At the conclusion of the study, the researchers showed that they could implement "evolved" robot genomes in real robots, and that these robots did indeed display the communication and foraging behavior observed in the simulations.