Steven Paligo, a.i. solutions; Sara Fields, a.i. solutions
Keywords: training, simulation, breakup, collision, observations
Abstract:
Military space operations have transitioned from a benign environment to a congested, contested, and competitive one. Maintaining space superiority in this changing domain requires space operators to have advanced test and training capabilities. Those are necessary to develop the tactics, techniques, and procedures for adequate threat response. To be effective, these capabilities must be sufficiently realistic and allow for repeated execution. Additionally, they often need to be sharable with allied forces to enable coalition space operations.Space surveillance sensor observations are the primary initial indicators of hostile adversary space activities. Therefore, they are a critical component to SDA test and training activities, and the organizers of those activities must be able to source them.Often, organizers rely on data sets of real, historical sensor observations. Although very realistic, these data sets are difficult to obtain and often too highly classified to be shared with allied and commercial space operators. Additionally, this use of predefined observations prevents operators from attempting courses of action that deviate from the historical record.To overcome these challenges, these types of test and training activities often devolve into tabletop exercises where the organizers use “white card” inputs to cue operator responses to adversary actions. Thus, the initial indications of adversary events normally present, are not ingested into the mission systems on which the space operators are using.In another approach to these activities, organizers manually build data sets of fictitious sensor observations. This has the benefit of being fully customizable to meet requirements. However, that requires great skill, is time intensive, and is error prone. While these data sets are customizable, they are pre-scripted, so space operators are still limited to the specific courses of action determined ahead of time by the organizers.What is needed is a way to generate space surveillance sensor observations that are realistic, easy to obtain, mimic adversary events, and respond in real time to operator actions. To bridge this capability gap and determine best practices to do so, our team developed a simulation environment called ObsSIM. It allows test and training organizers to generate realistic space events (launches, break-ups, RPOs, etc.) in minutes. It also models the ground and space-based sensors of the military and commercial space surveillance sensors to produce simulated physics-based sensor observations of those events. The result is observations representative of adversary satellites performing potentially hostile on-orbit operations. Because they are simulated and unclassified, these observations can be shared with allied forces.ObsSIM has been successfully used by the Joint Task Force – Space Defense’s (JTF-SD) Commercial Operations (JCO) cell as its primary simulation system for every Sprint Advanced Concept Training (SACT) exercise since 2020. During that time, it has been used to generate simulated observations representative of on-orbit breakups, maneuvers, RPOs, conjunctions, and launches in a simulation-over-live constructive environment, for a group of international participants. We have generated hundreds of simulation scenarios and provided over 155 million simulated space surveillance observations. This data simulation capability supports the development and testing of new advanced algorithms and analysis tools to provide on-demand awareness of adversary actions in space.Our solution allows scenario changes to be assessed and implemented in minutes as compared to the weeks or months it previously took to generate a single scenario. Without a system like ObsSIM, training exercises would have to consist solely of white card inputs and scripted responses as compared to having representative digital data that can be processed and displayed in real-time.
Date of Conference: September 19-22, 2023
Track: Space Domain Awareness