Enhanced Readiness and Deployability through Machine Learning and Intelligent Assessment

Orchestra is a digital data collection tool that harnesses the analytical power of computing for assessment results at scale. The Air Force continuously assesses its personnel throughout their service careers, across a spectrum of requirements. This assessment data typically sits passively in various isolated databases - Air Force personnel currently has 200 applications running on 111 different legacy systems ONLY logging data. The absence of a flexible but integrated digital collection interface means that opportunities for analytical insight are being squandered at both the point of performance, and at the level of strategic oversight. Orchestra can mitigate this data dilemma. The biggest challenge the Air Force faces is not having the most lethal weaponry, but the readiness, resiliency and performance of its personnel: “people, not planes,” in the words of former United States Air Force Chief of Staff Mark Welsh III. The U.S. Department of Defense has some 10,000 different types of human performance assessments (fitness, skills, psychological, etc.), of which ~40% are associated with the Air Force. This means some 10 million data points per annum that a common digital measurement platform can use to identify emergent performance trends and test and train prediction models. Orchestra’s commercially available digital scoring solution has already been used to support hundreds of different assessments and their associated scoring metrics. With no specialized knowledge or training, an end user can customize or “configure” the software for any kind of performance assessment, and sort, analyze and export the inputted results. In concert with our Air Force end user, we propose to demonstrate Orchestra’s utilization within the operational environment, and to develop enhanced analytical capabilities that support Air Force objectives. Digitized data collection is a first step. As the data set scales up, we can apply machine learning algorithms to process and track deeper patterns, generating prediction models with potential applications to all aspects of unit performance and individual Airmen’s careers.