Developing the innovative and transformative ecosystem for environmental sustainability in Phnom Penh as a model smart and sustainable capital city

• Main Author: Hoy, Jeremy Zhi Hao
• Other Authors: Heng Kok Hui, John Gerard
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Business::General::Government policies; Humanities::History::Asia::Cambodia; Social sciences::Economic development::Cambodia; Engineering::Computer science and engineering::Software::Software engineering
• Online Access: https://hdl.handle.net/10356/163577
• Institution: Nanyang Technological University

In the pursuit of a sustainable growth model, Cambodia is pushing to develop the digital economy to become the nation’s next growth driver. To do so, Phnom Penh, the capital city has been chosen to be the model city for test bedding of digital solutions and policies, and to become the model “Smart and Sustainable City” for emulation by the rest of the nation. The Smart Cities Network has been chosen as the primary consultant for the smart city project, who has chosen the School of Mechanical and Aerospace Engineering in Nanyang Technological University as one of the academic partners involved in the project, thus giving rise to this industry sponsored final year project (IS-FYP). This report thus outlines the work done throughout this IS-FYP, which is a smart city project in Phnom Penh, Cambodia, with a focus on the topic of Environment. This report takes a holistic approach in its analysis. First, a literature review of smart cities using a stakeholder approach is conducted. Next, an evaluation of the smart city potential of Phnom Penh is done by looking at the realities on the ground, including current socioeconomic factors (education, demography, general economic trends), city infrastructure (waste management, electricity, network infrastructure), policymakers’ intentions, coupled with primary source material from interviews conducted with different stakeholders based on interviews conducted during a work visit to Cambodia in the course of the project. From the findings obtained, and the requirements called for by the Phnom Penh Smart and Sustainable Strategic Roadmap 2020/35 document for a real-time monitoring solution for the Integrated Operations Centre in Phnom Penh, a Smart Monitoring Dashboard was developed. The developed dashboard derives its data from the Singapore Smart Nation Sensor Platform, which also served as the inspiration for the application. This dashboard is also intended to serve as a demonstration for the potential of having a similar network in Cambodia, and by using only publicly available information, illustrates the value of having open data sharing with the public. The second phase of the report explores the development of the dashboard which provides three functions, real-time weather monitoring, weather forecasting, and traffic monitoring. Real-Time Weather Monitoring The dashboard with the primary monitoring function was written in MATLAB and packaged using the MATLAB Application Designer and Application Compiler. To obtain the data, API calls were made to the “data.gov.sg” website to obtain real-time tracked data collected from various government agencies. The data is then processed and presented on the developed application in a user-friendly manner. Weather Forecasting Function Two methods were employed to predict future temperature values. First, a time series forecasting using the Box-Jenkins estimation method was used to select the most suitable model parameters of the Seasonal Autoregressive Integrated Moving Average (SARIMA) model using the MATLAB Econometric Toolbox to forecast future temperature values. Next, an artificial neural network, long short-term memory (LSTM), was trained to perform open and closed loop temperature forecasting. Traffic Monitoring Function Using MATLAB’s Computer Vision Toolbox, live traffic images of the Jalan Bahar Expressway collected over a 24-hour period were manually labelled with MATLAB Image Labeller to generate a groundTruth object. This was subsequently used to train a deep learning object detector (You-Only-Look-Once, YOLOv2) to count vehicle traffic from a still image.