School of Infocomm (SOI)

IoT data ingestion, cleaning, processing (including ML) and visualisation

This project, in partnership with ColdCane, leverages data science and machine learning to recommend optimal locations for ColdCane's vending machines to maximise sales and customer satisfaction:

• Footfall & Movement Analysis: A custom script analyses foot traffic patterns and demographics using ColdCane's data to pinpoint high-demand areas.
• Earth Observation Integration: Python algorithms analyse Earth observation data (likely weather or sunlight) alongside foot traffic patterns.
• Raspberry Pi Footfall Sensor: Captures real-time foot traffic data for vending machines which can be stored locally or transferred to AWS for further analysis.
• Data Visualisation Platform: Layers of data are visualised on a map using APIs, providing clear insights for strategic decision-making.
   

Team Members:
Dilton Goh Ying Ren, Stephanie Chan Xian Yi, Katapalli Brunda Varshini, Santhanaraj Jesica

Supervisor:
Mr Kelvin Tan

Creating a Smarter Store for Tomorrow

Stock-taking with a manual system is a long and tedious process. Risks, such as security breach by unauthorised personnel accessing the hardcopy information, might be encountered. 

Automated System Solution: 

By automating the processes with an inventory management system using available and emerging technologies such as Internet of Things (IoT), risks can be mitigated. Labour-intensive and mundane work can be also reduced while excess manpower can be effectively reallocated for more productive tasks. 

Requirements: 

• A new system to ease the process of registering, loaning and returning of equipment
• Automation of the stock-taking process for accuracy and authenticity

Solution:

Using Bluetooth Low Energy (BLE) IoT technology, a cloud-based system is created to simplify the steps to register, loan and return equipment to a designated area such as a store.

This technology enables the movement of equipment to be sensed and tracked within a defined perimeter. Such a system can thus be further expanded to automate massive stock-staking processes.

Technologies: 

• Bluetooth Low Energy (BLE) IoT
• Cloud technologies
   

Team Members:
Fang Yan, Tan Wan Harn, Muhammad Syafiq Bin Wahinudin, Tan Yong Da

Supervisor:
Mr Joseph Lim

IoT – Smart Home

Developing a mobile application for users to be able to remotely control their Philips Hue Systems through their smartphones. It is designed to be user-friendly towards everyone. 

Technologies: 

• Philips Hue Kit
• Hue API
• JavaScript
• React Native

Team Members:
Firmanain, Iman, Naga, Juraimi

Supervisor:
Mr Koh Thong Hwee

IoT – Smart Home

Many home users tend to forget to switch off the lights before leaving the house, leading to wastage of electricity, as there is no easy way for them to switch off the lights once they are out of the house.

Requirements: 

Using the Internet of Things (IoT) technologies, users can remotely control their smart devices using their smartphones. 

IoT-based Solution: 

User needs to have a Philips Hue Starter Kit (Hue Bridge and Philips LED Bulbs) as well as an Android application to control the light bulbs. Through the internet, the Hue Bridge and the Android application can communicate anytime, anywhere.

As a result, users are able to remotely control the smart light bulbs through a mobile application at their convenience. 

Technologies:  

• Android Studio
• Java
• Google Firebase
   

Team Members:
Tham Zhi Yang, Henry Maung Teng Han, Er Yi Chen, Tey Jia Jia

Supervisor:
Mr Koh Thong Hwee

Custodia

The current situation has seen a slight increase in "Crimes Against Persons". Recent events, such as the Orchard Towers incident on 2 July 2019, where a man was killed after a fight broke out, and the incident at Bugis Cube on 22 June 2019, where a man was beaten up by four other drunken men, have highlighted the need for our solution.

Raspberry Pi with Sensors Solution:

This project utilises Raspberry Pi equipped with a camera module and a sound sensor. The camera module streams video, takes pictures at intervals, and sends them to the Microsoft Cognitive Services Face API for training in recognising people and reading their emotions.

If the anger level and sound level reach a certain threshold, it is identified as an anomalous activity. An alert is then sent to the security guards on duty via a mobile app to investigate.

Technologies:

• Raspberry Pi and related sensors 
• RPi Camera Module 
• Flutter Framework 
• Azure Web Services 
• Visio Studios 2017 
• Azure SQL Database 
• Azure Cognitive Services
   

Team Members:
Ilyas Yao, Sean Liew, Martin Lee

Supervisor:
Mr Desmond Lee

 

CamPlus for Smart Classroom

CamPlus is a smart classroom solution to monitor students' condition using IoT and AI. The solution will take snapshots of students' activity and analyse their emotional profiles.

Parents of the student can view these profiles using mobile phones. The solution will also seek to perform automatic attendance taking. 

Key Technologies: 

• ASP.NET Core
• Google Flutter Framework
• Windows IoT (Raspberry Pi)
• Azure Cognitive Services
• Azure Cloud Services
   

Team Members:
Gu YaoChen, Ng Jin Wei, Xiong Wenxi

Supervisor:
Mr Frankie Cha