All Projects

NOW - Projects - Rich Semantic Track

Rich Semantic Track

Since the tragedy of September 11, 2001, the U.S. government, and by extension the U.S. Department of Defense (DoD), established data sharing policies mandating making data assets discoverable and understandable through published semantic and structural metadata. However, most data models in operational use lack an expression of semantics sufficient to enable automated data interchange. In particular, track data are critical to situational awareness in real-world and synthetic environments, are common throughout, and hold essential interest. Unfortunately, there is no common formalization of track information across DoD today. Rather, there are multiple track data formats in use requiring added cost in development and processing to enable data interchange across systems. Although data and, more broadly, information sharing across organizations and domains is essential to effective warfighting, parochial practices in procurement, management, and engineering continue to produce stove-piped systems that do not readily interoperate through effective exchange of information.

Several administrative and technical trends over the past two decades provide a foundation for addressing this problem. First, information sharing is a recognized need. DoD policies emerged in the early part of the 21st century to mandate data sharing practices across all systems, but unfortunately, over time, these mandates were softened as momentum towards implementation of these policies diminished. Information sharing across DoD remains program-centric, driven by communities of interest consisting of managerial and technical personnel who serve largely on a volunteer basis. At federal and DoD service levels, movement toward enterprise information management and information technology solutions is proceeding, but at a very slow pace. The National Information Exchange Message (NIEM) standard is available for information sharing across U.S. government agencies, with a military operations extension for DoD service use, but adoption and use remains uneven across the department. There simply is no single authority over all organizations and systems in DoD that enforces open data policies that will make data sharing ubiquitous. So, while these enterprises are heading in a positive direction, progress is slow and irregular.

A second trend over the past two decades is emphasis on interoperability based on a foundation of information sharing across systems to enable them to achieve a shared objective. Interoperability across command and control (C2) systems and modeling and simulation (M&S) systems is particularly important, with a growing requirement for interoperability with robotic and autonomous systems (RAS) as well. These requirements include creating training environments that faithfully replicate real-world warfighting by enabling trainees to employ their normal C2 systems (“train as you fight”); supporting test and evaluation of C2 systems under synthetic and controlled conditions; conducting post-exercise evaluation and analysis employing the real-world systems; and use of modeling and simulation with C2 systems to support operations planning, operations rehearsal, situation exploration and analysis during conduct of operations, and post-operation analysis and lessons learned. Researchers have defined levels of interoperability from low-level technical connectivity to high-level sharing of context, meaning, and operations, providing a roadmap for improving interoperability across systems.

A third trend over the past two decades is development and implementation of standards for describing the meaning of information in sufficient formality to enable automated processing by computer systems. This trend is particularly noticeable in the encoding and description of information on the World Wide Web (WWW), where vast repositories of knowledge have emerged. While information on the Web mainly has been connected through hyperlinks at a “page” or “site” level, application of recent techniques using technologies defining the “Semantic Web” is creating a web of linked data, enabling research into and exploitation of semantic relationships across massive knowledge bases.

All of these trends provide a vision of what is possible in dealing with track data. Doctoral research by Curtis Blais, NPS MOVES Institute, examines these trends from the past two decades in detail to establish context for the proposed formal semantic solution. That work focuses on achieving semantic and pragmatic interoperability across C2 systems, M&S systems, and RAS through a shared semantic model of track data (for semantic interoperability) and shared operations on track data (i.e., for pragmatic interoperability).

Dr. Blais's dissertation examined numerous track data models in use across C2 systems, M&S systems, and RAS to synthesize the essential concepts that are communicated. Despite decades of effort, success is elusive and interoperability is lacking: warfighters are unable to “fight as they train” and typically synthesize tracks through human interpretation. The research developed a formal semantic representation of track data to create a basis for unification of track data semantics and pragmatics. The research also examined use of the ontology in several operationally relevant use cases to provide a foundation for community adoption and implementation.

Specific contributions of this research included:

  • the first holistic examination of track data models across commercial and military C2 systems, M&S systems, and RAS to identify common information concepts confirmed by subject-matter experts in the use and modeling of DoD track data;
  • the first formal semantic model of track data synthesized from a variety of track data models across commercial and military C2 systems, M&S systems and RAS;
  • the first implementation and demonstration of a formal semantic model of track data unifying representations across C2 systems, M&S systems, and RAS for achieving semantic interoperability;
  • the first formal semantic representation realizing the vision of prior Rich Semantic Track (RST) model research;
  • the first formal semantic model of track data for unifying operations (for pragmatic interoperability) on track data by providing a common format across multiple C2 systems, M&S systems, and RAS;
  • the first qualitative confirmation of track representation requirements and content of a formal semantic model of track data by subject matter experts in track data modeling and track data fusion;
  • the first representation of track data using the Coalition Battle Management Language (C-BML) international standard for information interchange across coalition C2 systems, M&S systems, and RAS.

A further result of the work was development of the simulation environment enabling generation of multiple track data formats for exploring data interchange issues and for demonstrating transformation to the formal semantic model. This environment provides a foundation for future student and faculty research. The research also produced several transformations from operational track data model formats to the RST formal model that can be used for continued track data research and other purposes.

The RST research firmly establishes the readiness of current semantic technologies to address semantic and pragmatic interoperability of track data fundamental to military operations in real-world and simulated environments. By employing widespread Semantic Web technologies, the RST ontology can be rapidly adopted and applied, creating a foundation for future improvements. For immediate improvement in sharing of this fundamental information, adoption needs to happen now and across the DoD enterprise (at a minimum) to begin to change the way the Department represents and processes track data, as well as how it engineers systems, starting with those that deal with track data and moving to the development of all systems. The phrase “common semantics, common pragmatics, specialized where necessary” needs to become the unifying call for achieving the long-sought information availability and transparency envisioned in the early data sharing and interoperability policies.

Reference

Blais, Curtis L. Rich Semantic Track (RST) Ontology: Unified Semantics and Pragmatics for Track Data Interchange, Ph.D. Dissertation Naval Postgraduate School, Monterey California, 2018.