Small disaster relief robots with swarm intelligence routing

Abstract

This paper presents details of small disaster-relief robots that move inside a building to locate a specific object or person and find the shortest route from the object to outside using swarm intelligence routing. Multiple small robots are used for redundancy and ease of access to remote areas. Each robot has different sensors (Camera, ultrasonic sensor, temperature sensor etc) to locate the object of interest in various lighting conditions. These robots communicate with each other, dynamically isolate the failed robots and form a group of nodes (robots). In this paper we discuss also ant colony optimization (ACO) techniques to find the shortest route between the object of interest and the outside. The ACO algorithm used is described briefly. We use a practical test bed consisting of TI Evalbot small robots. Real time task scheduling on these robots are done using uC/OS. Description of the object identification image processing algorithm is also given.