Self-Aware Autonomous City: From Sensing to Planning

Abstract

This article presents a knowledge mining model, where a city can plan its development based on existing knowledge during city expansion, for example, telecommunication resource allocation and crowd forecast in a new region. Unlike most works that focused on Internet-of-Things (IoT) sensing, this study is aimed at urban planning by using harvested data, from the perspective of city architects. For large-scale metropolitan areas, a massive amount of data is generated every day, either from static surveys or dynamic IoT sensing. For urban planners, data collection is not their prior concerns. How to transfer harvested knowledge from exiting parts of the city to suburban/rural/ untapped areas is a new challenge. This is because those areas still lack sufficient statistics, and the density of IoT deployment is low. Therefore, development is risky and uncertain. To exploit new regions requires knowledge inference. Such a transition needs data interpretation from historical city dynamics, involving sensor deployment, human activities, and resource allocation in the vicinity. With the proposed model of this article, a city can estimate the requirement for resources when the peripheral areas on the outskirts of a city develops. The same model can be applied to enterprise sides for resource deployment, and applications are not merely limited to governments.