Smell as a Computational Resource - A Lesson We Can Learn from the Ant

Some insects are known to use chemicals called pheromones for various communication and coordination tasks. In this paper we follow an ancient advice 1 and investigate the ability of a group of robots that communicate by leaving traces, to perform the task of cleaning the floor of an un-mapped building, or any task that requires the traversal of an unknown network. More specifically, we consider robots which leave chemical odor traces that evaporate with time, and are able to evaluate the strength of smell at every point they reach, with some measurement error. Our abstract model is a decentralized multia(ge)ntadaptive system with a shared memory, moving on a graph whose vertices are the floor-tiles. We describe three methods of cooperatively covering a graph, using smell traces that gradually vanish with time, and show that they all result in eventual task completion, two of them in a time polynomial in the number of tiles. As opposed to existing traversal methods (e.g. DFS), our algorithms are adaptive: they will complete the traversal of the graph even if some of the agents die or the graph changes (edges/vertices added or deleted) during the execution, as long as the graph stays connected. Another advantage of our agent interaction processes is the ability of agents to use noisy information at the cost of longer cover time. A similar smell-oriented mechanism can be used to keep a spanning tree of a dynamic network. 1“Go to the ant, thou sluggard; consider her ways, and be wise” (Proverbs,vi,6).

[1]  Shmuel Gal,et al.  Search in a Maze , 1990 .

[2]  Jean-Louis Deneubourg,et al.  Random behaviour, amplification processes and number of participants: how they contribute to the foraging properties of ants , 1986 .

[3]  F R Adler,et al.  Information Collection and Spread by Networks of Patrolling Ants , 1992, The American Naturalist.

[4]  R. Andrew Russell,et al.  Laying and sensing odor markings as a strategy for assisting mobile robot navigation tasks , 1995, IEEE Robotics Autom. Mag..

[5]  Shimon Even,et al.  Graph Algorithms , 1979 .

[6]  W. Alt.  Biased random walk models for chemotaxis and related diffusion approximations , 1980, Journal of mathematical biology.

[7]  P. Zweifel.  Advanced Mathematical Methods for Scientists and Engineers , 1980 .

[8]  Jing Wang,et al.  Theoretical problems for the realization of distributed robotic systems , 1991, Proceedings. 1991 IEEE International Conference on Robotics and Automation.

[9]  Shlomi Dolev,et al.  Self-Stabilizing Depth-First Search , 1994, Inf. Process. Lett..

[10]  Edsger W. Dijkstra,et al.  Self-stabilizing systems in spite of distributed control , 1974, CACM.

[11]  Robert E. Tarjan,et al.  Depth-First Search and Linear Graph Algorithms , 1972, SIAM J. Comput..

[12]  J. Hopcroft,et al.  Efficient algorithms for graph manipulation , 1971 .

[13]  Uriel Feige,et al.  Short random walks on graphs , 1993, SIAM J. Discret. Math..

[14]  Solomon Lefschetz,et al.  Differential Equations: Geometric Theory , 1958 .

[15]  Richard J. Lipton,et al.  Random walks, universal traversal sequences, and the complexity of maze problems , 1979, 20th Annual Symposium on Foundations of Computer Science (sfcs 1979).

[16]  A. Fraenkel.  Economic Traversal of Labyrinths , 1970 .

[17]  Jayme Luiz Szwarcfiter,et al.  Hamilton Paths in Grid Graphs , 1982, SIAM J. Comput..

[18]  Rodney A. Brooks,et al.  Artificial Life and Real Robots , 1992 .

[19]  B A Huberman,et al.  Cooperative Solution of Constraint Satisfaction Problems , 1991, Science.

[20]  Shmuel Katz,et al.  Cooperative Distributed Algorithms for Dynamic Cycle Prevention , 1987, IEEE Transactions on Software Engineering.

[21]  D. Chialvo,et al.  Pattern Formation and Functionality in Swarm Models , 1995, adap-org/9507003.

[22]  Eli Upfal,et al.  Trading Space for Time in Undirected s-t Connectivity , 1994, SIAM J. Comput..