Implementation of Deep Learning for Smart City Applications: Lessons Learned

Abstract

The new urban digital industry 4.0 has greatly brought modern technologies to fulfill in the last decade. The smart city gains the attention of researchers and governance to advance the living quality by conducting several studies and testing deployment. In this paper, we aim to demonstrate the concept of systems on the Internet of Things (IoT) devices that consider several components of efficient performances in smart city perspectives. During the implementation between sensing devices in the infrastructure layer and cloud/edge computing in the data processing layer, the data communication technologies (e.g., ZigBee, Bluetooth, Wi-Fi, etc.) and IoT devices (sensor, actuator, and transmitter) are highly essential to secure the data transferability. Deep learning (DL) is to tackle optimally facilitating several smart city applications (e.g., smart healthcare, smart public safety, etc.). Furthermore, DL assisted in delivering advanced services to clinicians and improving the reliable performance of disease classification/detection discussed. We highlight the need for IoT and computing resource management in massive smart city applications for further enhancement with intelligent orchestration.