Sangjun Eom MS Thesis Final.pdf (2.46 MB)

TupperwareEarth: Knowledge-Based Ontological Semantics for the "Internet of Kitchen Things"

Download (2.46 MB)
thesis
posted on 01.01.2021, 00:33 by Sangjun Eom
The term “IoT” has evolved to encompass a wide range of diffuse concepts, but the common thread among the myriad definitions has been the convergence of technology to bring advanced conveniences to our every day, but complicated, lives. A long-term focus of the Collaborative Robotics Lab, and a particular focus of many with interests in consumer assistance, has been the kitchen, which acts as the “nerve center” of the home in many cultures. However, despite the grand vision of revolutionizing the kitchen and improving our lifestyles with technology, what today’s IoT-integrated appliances and kitchen-focused conveniences offer is mainly limited to a remote control. While remote control is certainly convenient, it still requires human planning in both cognitive and physical loads in performing cooking activities. The goal of this thesis is to build a framework of the network of IoT-enabled kitchen appliances, TupperwareEarth for the “Internet of Kitchen Things” integrated with an inference engine that utilizes ontology as a knowledge database. From simple clustering of sensor data to recommender systems that employ crowd-sourced preference data, the cognitive burden is reduced with proactive suggestions to high-level queries based upon the current kitchen state. Through the progression of the studies in the “Internet of Kitchen Things,” TupperwareEarth aims to reduce human planning that involves both cognitive and physical loads of burden by inferring solutions to the activities of daily kitchen living using ontological semantics.

History

Degree Type

Master of Science

Department

Engineering Technology

Campus location

West Lafayette

Advisor/Supervisor/Committee Chair

Richard M. Voyles

Additional Committee Member 2

Robert A. Nawrocki

Additional Committee Member 3

Brittany Newell

Additional Committee Member 4

Walter Daniel Leon-Salas