AR-driven Model-based Maintenance-training And Decision-support Aid (ARM-MADA)

Maintenance Operations and Training is an integral part of life-cycle management of the Navy Fleet of equipment. A complete life-cycle view requires maintenance inputs of field data and model-based approaches to predict fleet readiness. A model-driven dynamic reasoner can guide the technician in isolating and correcting the root cause(s) of anomalies, and AR (Augmented Reality) can support on the job training, virtual co-location, and just-in-time and just-in-place assistance approach to visualization by providing multiple views and multiple displays. QSI is of the opinion that an agile TrainOps (akin to DevOps) where the maintenance performed on the field with the aid of AR is an integral part of system “Life-cycle”, and where the performance and experience gained on the field is channeled back into the “classroom” training, will improve Navy’s maintenance training effectiveness. Integration of AR into Naval Maintenance Training and Operations requires a holistic approach incorporating Condition Based Maintenance (CBM) driven by intelligent diagnostics and prognostics inferred through monitored sensors and condition indicators. AR with a dynamic intelligent reasoner holds the promise of reducing the technician’s reliance on ground support and paper procedures. Qualtech Systems, Inc. (QSI), AVNIK Defense Solutions, Inc. (AVNIK) and Aptima Inc. (the QSI team) propose a model-based decision support solution for Naval Maintenance Training and Operations, leveraging commercial AR hardware such as Microsoft HoloLens. The solution follows a model-driven agile DevOps-like TrainOps process, leveraging commercial AR technologies and learning sciences concepts, where maintenance is driven by CBM, and where performance and experience gained on the field is channeled into the “classroom” to improve Maintenance Operations and Training. A model-driven dynamic reasoner can guide the technician in isolating and correcting the root cause(s) of anomalies, and AR can support on the job training, virtual co-location, and just-in-time and just-in-place assistance approach to visualization by providing multiple views and multiple displays. The solution aims to automatically guide technician through troubleshooting procedures by integrating a system-level TEAMS® reasoner with AR-based hands-free visual (gestures, eye-gaze) and auditory (speech-driven) interfaces. Additionally, nuances and additional maintenance information generated during the troubleshooting session can be captured and fed back to the depot to augment the underlying model and drive the in-house training activities. QSI’s training methodology uses a model that is the “single source of truth” for both operational and training scenarios. A model-based approach enables creation of comprehensive training scenarios since they are based on inherent system cause-effect relationships and hence more accurately represent actual system behavior and resemble real-world troubleshooting experiences.