Exploiting Time Series Data for Task Prediction and Diagnosis in an Intelligent Guidance System

Exploiting Time Series Data for Task Prediction and Diagnosis in an Intelligent Guidance System
Abstract

Time series data has been exploited for use with Case Based Reasoning (CBR) in many applications. We present a novel application of CBR that combines intelligent tutoring using Augmented Reality (AR) and prediction. The MonitAR system, presented in this paper, is intended for use as an intelligent guidance system for astronauts conducting complex procedures during periods of a communication time delay or blackout from Earth. Our approach takes advantage of the relational nature of time-series data to detect a task that the user is completing and diagnose the issue when the user is about to make a mistake.

Authors
Haley Borck Steven Johnston Mary Southern Mark Boddy, Ph.D.
Year of Publication
2016
Source
International Conference on Case-Based Reasoning (ICCBR) 2016