EMIA – Epi-terrestrial Multi-modal Input Assimilation

overview

Abstract

Extreme weather is an escalating urban issue while hundreds to thousands of screens showing traffic congestion with streams of pedestrians sprawling across. Can a smart city system flexibly react to every unique population and deal with impending urban flash floods?

The missing keys are (i) knowing what happens on every road between the screens, (ii) how to make all these data relevant to human urban activity, habitat and climate change, and (iii) synchronize the data as intrinsically as hundreds of thousands of traffic coordinators coordinating in unison on every road, under rain sunshine or even flooding.

The EMIA project is designed to do that. It is an extension from an AI forecasting urban flooding residing within a revolutionary superconstruct (AIMU). AIMU assimilates digital twin and makes it legible to AI, while staying geo-topologically accurate and relevant.

Unlike conventional machine learning algorithms in traffic CCTV, EMIA proposes a refocus of implementing road safety standards by extracting human centric parameters from existing traffic CCTV, as well as extending the reach of information by incorporating and analyzing video footage transmitted from enrolled dashcams among frequent road users. The source of the footage will be anonymized, and any unsuitable footage will be automatically deleted prior to transmission. The participants will be encouraged by an innovative reward system during the development period.

In addition, EMIA proposes AI-powered conversion of existing LIDAR points cloud data of Singapore and Virtual Singapore into a High-Definition structure that can be interactive with humans and AI within the AIMU. This new approach enables faster conversion and easier automatic assimilation and computation of change in urban infrastructure, coastal erosion, traffic diversion and evacuation in a flooding disaster especially in cities susceptible to climate change.

Finally, EMIA will mesh the traffic Big Data in the AIMU superconstruct for synchronization and data interaction so that this new form of city AI will be useful for semi-supervised computation for optimized traffic. These efforts fulfil 6 of 17 United Nation Sustainable Development Goals.

The proposed project is 2.5 years and its end product can be immediately added to our product offering to existing list of clients, or an independent entity commercializable by its unique universality and sustainability. The proposed project is a joint effort under the collaboration of Citymatrix Pte Ltd (Singapore) and Interactive Coventry Ltd (UK) with research and development conducted by both parties, mutual technology transfer and capacity building.