FA864924P0454

Small Business Innovation Research Program (SBIR) Phase II

Reliable and Scalable Dynamic Environment Perception with AI

In order to fill the SECAF s Operational Imperatives of Resilient Basing and Readiness to Deploy and Fight, the Air Force needs to adopt AI solutions to bolster our ability to constitute and maintain new and existing infrastructure. By employing Haxion s solution, organizations will be able to rapidly model 3D point clouds, the solution can be used to more efficiently perform engineering, training, mission planning and tactical operations, with an overall target of achieving 30% increase in battlefield awareness. Haxion is developing an AI system powered by 3D-LLMs, which we call the Haxion Shape Language, designed to give disconnected and distributed teams access to actionable intelligence of large facilities with real-time 3D modeling and ROS-based scan-to-BIM representations, along with advanced queryable analytics at the tactical edge. The Air Force has identified an urgent need to transform base installation management and equip Civil Engineering (CE) personnel with tools that are a force multiplier. It is essential to mission readiness and our force lethality that the bases that serve Airmen are utilizing tools that are advanced as the weapon systems they operate. In the past several years, we have witnessed a wide variety of technologies that enable rich semantic understanding of point clouds. These techniques include explicit geometry based machine perception approaches that translate 3D point cloud features into Building Information Models used to understand building shape, facility infrastructure, and material characterization of an environment.

The goal of phase II is to continue the R&D efforts initiated in Phase I. Funding is based on the results achieved in Phase I and the scientific and technical merit and commercial potential of the project proposed in Phase II.

AFX244-DPCSO1

F2D-10816

X24.4

Meghan Scott