We are proposing a data processing pipeline that uses a modern Evidential Occupancy Grid as its back-end to fuse multi-modal data for the purpose of locating concealed explosive hazards. Our goal is to provide real-time, human-readable data on the location of buried ordinance. Previous work in de-mining using Evidential Occupancy Grid methods have focused on batch processing using a standard Dempster-Schafer Theory framework. We expand on this work by using a modern update to Dempster-Schafer Theory known as Dezert-Smarandache Theory (DSmT), in particular Proportional Conflict Resolution rule #6. To complete this work, we will take pre-recorded, location-tagged EMI and GPR data and pass that into the front-end of our pipeline before passing it into our back-end. Our pipeline front-end will designed to be flexible enough to either, A) Take raw data and run signal processing algorithms that have been used by SSCI in the past to produce usable data. Or B) Pass through high level, ed data that can be immediately consumed by the back-end. Phase 2 of this program will see our code implemented on a robotic platform outfitted with a sensor suite for detecting explosive hazards. Information about the location of the hazards will need to be relayed back to the platform operator.