[INFR11239 (UG)/INFR11150 (PG)]

School of Informatics, University of Edinburgh

Please contact Professor D K Arvind ([email protected]) with any questions you may have.

Contents

• Introduction
• Downloading the git repositories
• Setting up the data analysis environment
• Setting up the Android Development environment
• Connecting to the sensors

Introduction

Welcome to the PDIoT course! You will experience the different facets of designing and implementing a complex IoT system, from data collection (Coursework 1) and system specification to demonstration of a prototype implementation, over the course of 10 weeks (Coursework 3). The practical work will be complemented by knowledge gained through personal research on foundational topics in Internet of Things to be distilled in one 3000-word essays (Coursework 2).

Each student will be provided the following for Coursework 1 & 3:

• A wearable Respeck sensor
• Respeck accessories: plastic bags and MeFix tape
• An mBed development board (NRF52-DK)
• A Nordic Cube (Thingy) - a compact multi-sensor prototyping platform based around the NRF52 SoC

You will need a smarphone running Android 6.0 or higher to run the apps needed for the course. If no one from your team owns an Android phone we can provide you with one upon a special request.

This year, your task will be to implement a human activity recognition system for a range of common activities listed below by analysing data from a wearable sensor using machine learning techniques and displaying the results in real-time in an Android application.

You will first collect data using two sensors:

• the Respeck sensor, worn on the lower left ribcage, sampling accelerometer and gyroscope data at 25Hz