tech track papers

Categories: 2016, Government Science and Communication

Efficient Object Maneuver Characterization to Support Space Situational Awareness

As the number of objects in space increase exponentially, the need for Space Situational Awareness (SSA) to protect assets from intentional and unintentional threats increases as well. SSA is becoming a “big data” problem due to the prevalence of low-cost small satellites, the proliferation of debris and increased presence of symmetric enemy space objects. Additionally, there are multiple, uncoordinated, space surveillance systems collecting data at varying cadences to create datasets that grow in proportion to the number of telescopes and other sensors. While these datasets are large, they are not persistent or conditioned and are frequently noisy, which makes it challenging to maintain a satellite’s chain of custody and detect out-of-class maneuvers in a timely manner. Although many SSA operations remain a manual process; automated analysis tools have emerged to assist in timely exploitation of large streams of multi-source data.  Working with our partners at the Air Force Research Labs, Aptima, Inc. has developed automated satellite maneuver prediction algorithms that can fill in the gaps of persistence, gain credible custody of uncorrelated targets (UCTs), and characterize potential threats. The objective is to not only detect maneuvers quicker than an analyst but also to infer the “intent” of a satellite’s movement pattern to predict future maneuvers based on observed behavior. To achieve this, we adapted computationally efficient machine learning algorithms that we originally developed for automated Activity Recognition in the land, sea and air domains. We have demonstrated the high accuracy of probabilistically detecting and predicting maneuvers of the Galaxy-15 and Anik F1R satellites on noisy, intermittent synthetic and operational datasets. Early results indicate the timeliness of maneuver detections and accurate prediction of future maneuvers from a short time- history of past observations.


Author: Charlotte Shabarekh, Gene Keselman
Topic: Government, Science and Communication

  • Efficient Object Maneuver Characterization to Support Space Situational Awareness

    Download File