Agility Prime (X20.D) : Virtual Training Assistant and Assessment Application using Adaptable Knowledge Graph Repository
This Small Business Technology Transfer (STTR) project will develop a Virtual Training Assistant (VTA) using a closed-domain Conversational AI (CAI) platform and adaptable Knowledge Graph. The VTA will provide hands-free support for aviation technicians during training, repair activities, and final inspection.
The VTA platform is based on three basic elements:
- Data Ingestor and Domain-specific Knowledge Graph (DI-DKG);
- CAI; and
- User Interface and Domain and Application specific Dialogue Models.
Example applications can include Scenario-Based Training (SBT), Hands-free verbal access to technical manuals and information, and step-by-step inspection procedures with form-filling and annotation support. While we will demonstrate a functional system that includes all three components, each element will be able to operate independently and communicate with other components using RESTful APIs. In particular, the DI-DKG could be used independently to support other CAI systems as well as Augmented Reality devices that need ready access to up-to-date and readily accessible technical information.