A multi-agent model for mine detection - MAMMD

From decades mines have taken thousands of innocent lives and a lot of research is going on mine detection problems. In this paper we have proposed a multi-agent based model for detecting (MAMMD) mines in unknown environment. Mine positions are unknown to the agents and they cannot predict there positions using any probability method. Agents have mine detector devices and they coordinate their actions/movements with each other. MAMMD architecture is implemented using layer based approach to make the system distributed and fault tolerant. We are using an algorithm which is quite similar to depth first search algorithm for movement of agents. Proposed architecture is evaluated on large number of test cases including use of different grids sizes from 10x10 to 100x100. Grids had mines randomly placed, occupying 0% to 30% of the search space. Experiments used 5 to 25 agents for each randomly generated grid with same mine ratio. Experimentally we have observed that MAMMD is effective in both time and solution quality.