SBIR-STTR Award

Athenium Analytics’ enterprise, decision-support software leverages deep Machine Learning (ML) and Computer Vision (CV) to measure, mitigate, and manage risk through imagery analytics. By adapting its commercially-successful facility damage assessment tools (i.e. confirming desired effects on target), Athenium Analytics’ capabilities suite will be able to provide automated Battle Damage Assessments (BDA) to Air Force ISR elements. The company’s technology can also be leveraged to accelerate and optimize weapon and/or munitions selection. Athenium Analytics R/R&D with an Air Force end-user will lead to the production of BDA reports algorithmically, using high-resolution, panchromatic, multi-spectral, and/or electro-optical aerial imagery, significantly shortening the BDA delivery cycle to less than 24 hours after imagery is received from the Air Force, with the ability to assess large volumes of target locations simultaneously. This revolutionary application and adaptation of productized ML and CV assessment technology to the Advanced Battle Management System (ABMS) program will represent a significant increase to warfighter safety, reduce latency during multiple stages of a mission, inform weapon selection automation in future operations, and translate into significant cost savings. The adaptation of the non-defense commercial solution of IRIS through R/R&D for the Air Force will require modifications to the COTS solution’s current product workflows, technical advancements in ML and CV algorithms by way of supervised and unsupervised training, and the development of new predictive structural damage insights. The required developments would also allow for the integration of Air Force-provided imagery into IRIS’ image annotation

Athenium Analytics will employ semi-supervised learning to reduce the amount of image annotations needed during R&D to more quickly develop new AI image predictors. The annotated attributes will also be used to build on research conducted at Ellsworth AFB on the deterioration of paved surfaces by a combination of liquids that drip or leak from aircraft and extreme temperatures as had been observed and studied by the installation. Moreover, the state-of-the-art deep convolutional networks (DCNs) and semi-supervised AI training processes to be employed as part of this R&D program will allow for the automated development of lists with the types of aircraft types that can land at each location based on site-specific conditions and input from Air Force SMEs. Athenium Analytics IRIS and Gauge solutions, which will be used as part of this program, are assessed at a defense technical readiness level of seven (7). As part of the proposed Phase II R&D program, Athenium Analytics will address Air Force-specific use-cases by training existing and new artificial intelligence-driven algorithms (“bots”) to identify surface or infrastructure features leveraging high-resolution imagery and IRIS’ tested R&D workflows. Athenium Analytics’ Phase I feasibility discussions with Malmstrom and Ellsworth AFBs were generated from Athenium Analytics’ work program with AFIMSC, whereby the parties explored additional AI/ML capabilities and use cases similar to such work that could benefit new Air Force stakeholders and end users. As part of ongoing work with AFIMSC and AFCEC, Athenium Analytics delivered structural condition, damageability & natural hazard risk assessments for approximately 80,000 buildings across 100 Air Force installations within five (5) months, using artificial intelligence applied to aerial imagery & without on-site visits. This effort, which focused on roof condition, involved the segmentation of multi-section roofs as inventoried by the Air Force, with assessments provided at the section-level. The success and advances achieved under the roof assessment program prompted feasibility discussions for a pilot that would adapt Athenium Analytics capabilities for the automated of road, runways and paved surface assessments in and around Malmstrom and Ellsworth AFBs. For the proposed R&D program under an SBIR Phase II, Athenium Analytics has partnered with Malmstrom AFB and with Ellsworth AFB. More specifically, the proposed program will be undertaken directly with Malmstrom AFB’s 341st Civil Engineering Squadron of the 341st Mission Support Group & Ellsworth AFB’s 28th Civil Engineering Squadron of the 28th Mission S