With the development of the state-of-the-art mine detectors, remote sensing has been recognized as a viable solution for dealing with the landmine problem in the world. However, the reliability of the remote sensing method largely depends on appropriate manipulation of the detector. A terrain scanning robot is introduced that is essentially a dexterous autonomous mobile manipulator that can scan natural terrain using a typical handheld mine detector. The development of the terrain-scanning robot has been a great step forward towards the feasibility of the remote sensing method. The terrain-scanning robot features real-time map building and path planning. Path planning defines an obstacle free path for the mine detector that scans the terrain from a close distance. Map building involves sensor fusion to generate a reliable terrain map using incomplete and imperfect information obtained from multiple rangefinders. The use of a Kalman filter, in conjunction with other filters in a multi-step filtering process, provides a dynamic modelling framework that updates the terrain map based on local observations in real time. The terrain slope standard deviation plays an important role in the optimality of the output of the filter. The slope standard deviation is determined online using fuzzy terrain typing. The proposed fuzzy terrain typing may be used for characterization of the terrain in applications such as planetary mobile robots. The significance of the proposed map building, path planning, and terrain typing of the terrain-scanning robot was validated experimentally using the MR2.
Landmine detection using an autonomous terrain-scanning robot
Landminensuche mit einem autonomen, das Gelände abtastenden Roboter
The Industrial Robot ; 32 , 3 ; 240-247
2005
8 Seiten, 8 Bilder, 27 Quellen
Aufsatz (Zeitschrift)
Englisch
Marwa: A Rough Terrain Landmine Detection Robot For Low Budgets
Tema Archiv | 2014
|Ardunio controlled landmine detection robot
IEEE | 2017
|Swarm based autonomous landmine detecting robots
IEEE | 2017
|