The goal of this project is to create an edge-native AR game which is demonstrable in tech conferences and to carry out data collection and analysis to reveal the advantages of edge computing. An edge-native application should fully exploit the benefits of edge computing which may help resolve technical issues involving computation, bandwidth, and latency. In this project, we will create an AR arrow shooting game with augmented armors by using AI-powered real time pose tracking. The computational load of pose tracking is so heavy that it is beyond the capability of personal electronic devices and needs to be offloaded to an edge server. We will build our work upon two existing AR projects, Disco Arena and OpenRTiST and work closely with our clients from The Human-Computer Interaction Institute at Carnegie Mellon University (CMU-HCII) and Interdigital. We will also carry out intensive play tests and emulations to make comparison between edge computing, cloud computing, as well as local computing on personal electronic devices. The outcome of the project will demonstrate the performance of AdvantEDGE and produce data analysis for quantitative characterization of edge computing applications.

Project Instructors: Carl Rosendahl, Chris Klug

Team Members: Yuting Jing, Chih-Hsuan Kuo, Tianyi Zhang, Yingran Zhang, Yu Zhu

    This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.




    700 Technology Dr. Pittsburgh, PA. 15219