Data Ontology Changes Maintenance Paradigm on a Global Scale

Within the DOD, each Service, code, shop, and department may be using different terms to refer to the same part or process. This lack of ontology alignment between the various maintenance environments causes bottlenecks, delays in maintenance activity, supply chain inefficiencies, and a lack of weapons systems readiness.  

Within a machine learning (ML) and artificial intelligence (AI) eco-system, nonuniform data become problematic. The Joint Artificial Intelligence Center (JAIC) is the DOD’s AI Center of Excellence providing a critical mass of expertise to help the Department harness the game-changing power of AI. JAIC is tasked with accelerating the delivery and adoption of AI to achieve mission impact at scale. Creating a common ontology is an important first step toward predictive maintenance goals by creating a singular global reference that enables standardization of datasets. This alignment would enable seamless deployment across all Services.  

On February 4, attendees representing all the Services, JAIC, and Condition-Based Maintenance (CBM+) subject matter experts gathered for a presentation on the results of a year-long CTMA project creating an ontology for the Army H-60 air platform. The H-60/aviation AI models were chosen as the demonstration ontology because all the Services use this platform. 

Each Service has developed Interactive Electronic Technical Manuals (IETMs) that include engine, airframe, and other physical structure, for maintenance, and for parts. The ontology has been created by incorporating IETMs from various Army H-60 models to create a universal resource for all models. 

Industry partners, Tamr and LMI, used a virtual platform to illustrate their process for integrating maintenance ontology with generic helicopter language, and finally H-60 variant parts ontology. Key benefits included: 

  • Establishing the future gold standard for maintenance data and process documentation 
  • Repeatable and scalable creation of adaptive digital twin 
  • Force multiplier for data analysis for improved readiness and cost avoidance 
  • Ontology-enabled investigations 

The demonstration was a huge success with transformational potential across the DOD and supports the expansion of the use of AI and ML. JAIC plans to share the results across multiple platforms while continuing the next phase of the project.