Into the Dataverse Hackathon

In 2019, I competed in the Into the Dataverse Hackathon with 3 of my labmates. The purpose of the hackathon was to propose solutions for maintenance data logging issues experienced across several industries including the department of defense. Maintainers have a primary goal of doing the actual maintenance work and a secondary goal of logging all of the actions they took in a series of forms after the procedure. However, this is tedious, places extra burden on the maintainer, and is often done quickly and with errors so they can finish their shift. Nobody likes extra paper work. However, this data has tremendous value in terms of predictive capabilities within maintenance pipelines. Therefore, there is a need to assist maintainers in logging accurate, structured data for their forms, as well as capturing unstructured data not even being considered currently.

Our proposed solution was SCRIBE which stands for Systematic Contextual Recorder with Intelligent Blending of Elements. It is an AI video and audio assistant for maintainers. Below is an example of how the system could work. The user activates the system using keyword “SCRIBE” at which point video and audio begin recording and are processed for logging. Upon ending the procedure, the maintainer would have a form pre-filled for their review.

During the hackathon weekend we pitched our concept and won the competition. We were awarded a $15,000 contract to further pursue the idea which myself and one other teammate continued to do. This involved learning more about the problem from key stakeholders so that we could propose a more detailed and relevant system in a white paper.