Escape algorithm based on a combination of basic actions for small unmanned ground vehicle

Abstract

Since robot has been introduced in 1960s, it has been used in various areas for dealing with the tasks that human was not able to handle. On account of the feature mentioned above, needs for the military robots which are able to manage the military missions without human loss have been concluded as developing multifarious concept of the military robots. From many different concept of the robots for military-use, SUGV(Small Unmanned Ground Vehicle) has been the most practical solution because of the demand of military reconnaissance and explosive detection during the ground engagement. This type of robot is composed of a mainbody equipped with caterpillar tread and two flippers for overcoming obstacles. its performance was proven in Iraq and Afghanistan. However, the limitation of its locomotion type leads to the driving limitation when facing the extreme environments on the field like uneven grounds and ditches. In this reason, there have been conventional approaches studied for overcoming the circumstances. One approach is stabilizing the CoG(Center of Gravity) of the main frame by moving Flippers. the other is avoiding obstacles by sensing the ground environment. In spite of the contribution of these approaches for ensuring its travel-ability, it is evitable for SUGV to face its limitation when the robot is stuck or overturned by sudden changes of environment. As a consequence, this paper presents the escape algorithm with candidate sequences for various extreme environment of SUGV by using of practical sensors. Firstly, the definition of required escape state is to be delivered. In addition to that, this paper proposes the escape algorithm corresponding to the defined state with the sensors mounted on the robot. In connection of with this issue, I wish to address the dynamic simulation for the feasibility of the algorithm. Finally, the algorithm is implemented on the test-bed for validation of the algorithm with various experiment environments.