A Writing Activities Monitoring System for Preschoolers Using a Layered Computing Infrastructure

Abstract

Growing advancements in sensor based smart devices allow monitoring the activities of preschoolers in classroom. The existing systems for monitoring the writing activity of preschoolers use the local processing and Cloud based solutions which require high processing time and seamless connectivity, respectively. Such systems are not feasible for low cost and low power smart devices. This work presents a writing activity monitoring system that monitors the gripping, holding, and orientation of the pencil and transfers the results to their parents and teachers. The system uses a layered computing infrastructure - Edge, Fog and Cloud layers, that allows energy consumption reduction and provides the solution closer to the user. Specially, the system selects the appropriate layers for processing and recognition of writing activities which provide the desired level of accuracy and prolongs the connectivity of the system. A prototype of the system is developed and examined on real user data to validate the accuracy and energy consumption of the system.