Towards a Platform for Smart City-Scale Cognitive Assistance Applications

Abstract

This position paper describes CognitiveAR, a system that seamlessly interfaces AR devices with smart city environments. Edge computing nodes distributed throughout the city enable multi-user cognitive assistance applications that require (1) real-time sensor data from the environment, such as approaching cars, and (2) computing resources for low-latency video processing. We discuss three such applications to elicit requirements for a platform to support them, and present our system design.