Mobile Geographic Information Systems (GIS) for Humanitarian Demining

Abstract

The threat of landmines and other explosive remnants of war is a serious concern around the world. While landmines demand attention due to the thousands of civilian casualties they cause each year, perhaps even more shocking is the fear they instill in local populations, inhibiting movement and denying access to thousands of square kilometers of land in more than 80 countries. Humanitarian demining seeks to rid the world of landmines and return local populations to their displaced land. To meet this goal, surveys of hazardous areas, describing their location and contents, are used to produce threat maps for a given location and secure adequate funding from donor organizations for clearance operations. The focus of this study is a mobile GIS system, developed by the Geneva International Centre for Humanitarian Demining (GICHD), which allows rapid, accurate, and completely digital collection of these demining surveys. Using data collected during local evaluations of the demining Survey Tool at the University of Kansas campus and on foreign field deployments in Chile (2004), Albania (2004), Ecuador (2004), and Lebanon (2006), a fit-for-use analysis was performed on each component of the Survey Tool. Experiments were conducted to evaluate the accuracy of its GPS and laser rangefinder mapping devices, and methods for improving that accuracy were investigated. The system was well received by all of its users and gauged to be twice as fast, require half the personnel, and provide higher levels of accuracy than traditional methods for collecting demining surveys. Even though the system was deemed fit for use, suggestions for improving all components of the device resulted from user feedback and observations of the system in the field. The system's GPS receiver was predicted to provide 5 m accuracy 50 % of the time and 10 m accuracy 95 % of the time. If GPS positions were averaged for 1 minute, the 95% accuracy improved to 7.5 m, and if positions were averaged for 4 minutes, the 95% accuracy improved to 5.6 m. The two types of laser rangefinders used by the system were found to have a mean accuracy of 2.7 m when shooting at a location on the horizontal bare earth and a mean accuracy of 1.1 m when shooting at a well defined vertical target. Rangefinder accuracy varied due to level of user experience with rangefinders or other sighting equipment, and thus proved the value of training with these devices. Also, significant errors in bearing measurements with the rangefinders caused by magnetic interference from one user's eye glasses indicated that this issue requires considerable attention by all users of laser rangefinder devices. General themes that were found to be extremely important to the success of the demining system, such as the value of training, the need for system flexibility to match traditional field methods, and the complexities of GIS data collection in the field, should be a focus of any mobile GIS field program.