An ontology for semantic classification of construction areas and tasks

Army Engineers require in-depth site characterization of the deployed environment to fully implement and plan the required (de)construction tasks that will prepare an area of operations to its end goal. With the increasing use of satellite imagery and voxelized LiDAR data, digital elevation and terrain models can be utilized for this site characterization. However, a standard semantic classification schema for the potential obstacles, the associated material properties, and the required (de)construction tasks is currently unavailable to the Army Engineer. Using ontology software such as Protégé, an ontologically grounded semantic classification can be created. The information included in the ontology includes semantic titles of potential obstacles from various military and civilian construction standards, a wide variety of construction equipment available to the Army Engineer, various material properties, the characteristics required for production estimates, and the potential (de)construction tasks that are carried out by each category of equipment. The ontology can be extended past solely military applications and adapted to civilian construction projects. With the advent of machine learning, this ontology can be used as a framework for automatic classification of features based on geometric and volumetric information from LiDAR and voxel data.