16 | ENDEAVOUR ENGINEERING AND IT EXHIBITION - Program Guide 2017

ELECTRICAL AND ELECTRONIC ENGINEERING

Electrical and electronic engineers play a key role in the design, implementation and management of electrical systems to solve practical problems, such as systems for automation, surveillance, energy conversion, power distribution, telecommunications and information processing.

Embedded wireless accelerometers for sporting applications

Laura Brennan, Jessica Bartels, Lewis O’BrienIn this project, integrated low-power accelerometer and wireless communications technologies will be combined in the design of a family of battery-powered wireless accelerometers suitable for embedding in mechanical systems. Our accelerometer is designed for an AFL boot, to count the kicks a player makes in a game or training session.

Managing a healthy and safe workplace

Mufei Pan, Yuanpu li, Shengyu ZhangRecent research has shown a direct relationship between a healthy and safe workplace with increased productivity. Chronic health conditions can develop over an extended period of time, especially for employees of highly strenuous working environments in industries such as construction, mining, factory and farming. In the short- term, fatigue and reduced alertness can potentially increase safety risks. The aim of this project is to better manage health and safety in such environments using a real-time monitoring solution comprising a body sensor network connected to a software management hub. The body sensor network comprises multiple sensors that collect motion data from various location of the body. The collected data will be wirelessly transmitted to the software management hub. Data analytics will be used to evaluate a worker’s level of fatigue and tiredness. Further, using predictive analytics, future events of fatigue and tiredness can be predicted to reduce safety risks.

Designing a wearable Photoplethysmograph (PPG) device for monitoring

Da Zou, Sebastian Gawin, Agam RajThe aim of this project is to design, develop and build a wearable Photoplethysmograph (PPG) device for the monitoring of important human physiological signals such as PPG, SPO2 and respiration. These physiological signals are vital in detecting and diagnosing numerous diseases and monitoring the general health of the patient. While there are many devices used in hospitals and clinics that already perform this function, our device will allow patients to monitor their health at home due to its portability, low cost, wireless capabilities and integration with the Android platform. Patients will be able to access this information through an easy to navigate user interface.

Adversarial radio-location Angus White, Patrick Samon, Gareth SeddonTraditional computer-based radio scanning systems can cost in the order of thousands of dollars. A recent innovation in this area has been the repurposing of low-cost TV tuner dongles to produce extremely cheap alternatives to these systems. In this project, we investigate the applications of using these dongles to construct a radio-location system, and propose a game-theoretic strategy for tracking a transmitting adversary.

Fine pointing control of nanosatellites

Daqi Liu, Qigang Wu, Jieru NiThese satellites will require a high degree of pointing accuracy to communicate with future intra-solar system satellite missions as well as pointing scientific instruments. The aim of this project will be to develop the necessary software and hardware to control the orientation of a nanosatellite capable of meeting the pointing requirements while subject to the stringent mass, volume, power and thermal constraints of a nanosatellite. In particular, this project has designed an Arduino-based detumbling flywheel systems, Raspberry Pi-based star tracker and relevant control algorithms implemented by Simulink and hardware.

Power systems voltage stability analysis

Yu Liu, Hong Chen, Fanghui ZhangThe aim of this project is to study voltage stability issues related to power systems, and to design adequate compensation schemes. This involves derivation of mathematical models of a power system, which will then be used to analyse and control design.

Electronic voting response system

Haoge Deng, Yao Peng, Yuyang WuElectronic Voting Response Systems (EVRS) are used in large classes as a means of promoting student engagement through participation in quizzes and surveys and allowing students to provide instant feedback during lectures. These types of systems are typically costly and therefore difficult to roll out to a large subject with many students. This project serves to develop the hardware, including small, relatively low-cost push button wireless controllers, a receiver base station that can receive many transmitters’ responses in a short period of time and the software to record the student response data and potentially display it on the lecture screen in real time.

Nanosatellite electronic power system

Herbert Trent Moore, David Nguyen, Alexander BarnardThe PowerKey™ EPS (Electronic Power System) for a nano satellite supports a wide range of subsystems power requirements and includes a battery management architecture. Voltage and current sensors are used to accurately model power consumption for predictive purposes. An on-board CPU and FRAM ICs are used to estimate and update subsystem power consumption as well as available power. The PowerKey™ communicates with the CDH (Command and Data Handling System). It is capable of directly transmitting and receiving data through the communications module. This enables basic system functionality if the CDH fails. Our system is innovative for its robustness, efficiency, and is less expensive than any other off the shelf EPS for nano-satellites currently available.

Modelling of irrigation channels with culverts

Jingzheng Zhang, Peiwen Xie, Zhiyi ChenDecentralised distant downstream control is the preferred strategy to control open irrigation channel network due to its inherent advantages. While in distant downstream control mode, a gate controls the distant downstream water level in its pool. The majority of the open irrigation channel network of Goulburn Murray Water, Southern Rural Water, and Coleambally Irrigation are operated in a distant downstream control configuration. Culverts and bridge structures associated with road crossings obstruct the water flow and cause issues for distant downstream control. As a result, a useful solution is to design a model that incorporates the dynamics of the culvert.

Optimising the scheduling of operations in an underground mine

Junzhu Jiang, Jianing Lyu, Elliot FitzgeraldThe problem of designing and scheduling an underground mine in the best way possible is of great practical importance to the mining industry. This project implements recent breakthrough work on scheduling construction of access and extraction of the ore in order to maximise the net present value of the mine over its lifetime. This project developed an optimisation algorithm in MATLAB with an appropriate user interface. The algorithm was tested to ensure the output sequence is the most optimal. The project also extended the modelling to a more realistic approach of drilling tunnels in two rock faces per month (as opposed to just one) to further utilise the drilling equipment. A suggested drilling schedule is given as a second output.

Crowdsourcing of mobility hazards

Lingyu Song, Qiong Wu, Changyang LiuNon-protruding hazards, such as potholes, descending curbs and stairs, are genuine fall risk factors for those using ambulatory aids such as wheelchairs and walking frames. This project uses crowdsourcing to build a comprehensive hazard database with a user-friendly mobile application. Uploaded hazard locations for users of ambulatory aids will be stored in a central database in a hosted environment. A user will be alerted to a known hazard through their mobile phone when they approach a hazard. Collectively, the mobile app and its database of known hazards is expected to positively impact the independence and significantly reduce the fall risk of a growing group of people reliant on ambulatory aids.

RFID reader/writer for hearing aid identification

Maynega Krishnananthan, Evelyn Sian, Bao ShengnanIn collaboration with Blamey Saunders, the aim of this project is to design and build a handheld radio-frequency identification (RFID) device capable of reading and writing information to RFID tags embedded in hearing aids for identification purposes. This involves designing a custom printed circuit board (PCB) that integrates a touch screen, RFID chip and SD card. To further enhance the handheld capability of the device, it will be powered by a rechargeable battery and enclosed in a 3D printed case. The device will allow Blamey Saunders to streamline their inventory processes as well as simplify client interaction across their facilities remotely.

Development of fully autonomous drones for disaster management

Hao Tang, Isaac Hook, Lachlan DowlingIn collaboration with the Metropolitan Fire Brigade, this project aims to build an autonomous drone prototype to image and map an enclosed area during and after an emergency. Through the design and implementation of different sensors, algorithms and interfaces, the drone will be able to detect and avoid obstacles and map its surroundings, while also providing crucial information back to a base station. This ultimately reduces exposure of humans to hazards.

Gesture driven user interfaces for Internet of Things

Hongyou Zhao, Yunqi Liu, Lidi XiaGesture is widely considered as the new user interface for many systems and devices that can be controlled and managed. We have designed a gesture based user interface in our project. This interface consists of two parts: Microsoft Kinect as the hardware platform and a Windows Presentation Foundation program as the software. Using machine learning methods, this interface uses six gestures that can be applied to control a set of different Internet of Things (IoT) devices. In the demonstration, there will be two IoT lights and one IoT speaker. Each device is associated with several different functions which can be performed by a certain gesture.

Smart trains Zhanhuai Huang, Chenxu Gong, Xin JiThe project seeks to design advanced solutions that consist of sensors that collect data about deteriorating conditions for passengers in trains, such as increased congestion, strong vibration and temperature. Once conditions are calculated to be unbearable, warnings would be sent, helping passengers navigate to less congested carriages in real time. The collected data is also valuable to operators, allowing for better scheduling and understanding of passenger conditions. Further sensors can be tuned to target sections of tracks that require maintenance.

Flexible communications module for existing smart meters

Rui Wang, Mucheng Ren, Shenglin XuCurrent smart meters are very good at measurement, and have basic communication capabilities. Typically meters use a RS-232 port to connect with a modem or communications module to connect to carrier networks or other custom mesh networks. This project will build a new kind of module that incorporates these basic functions, but adds several new important capabilities, such as displaying the real-time electricity consumption on PC or mobile phones and remotely controlling home devices through web application. This project aims to realise a ‘smart home’ by implementing the concept of IoT (Internet of Things).

Indoor air quality monitoring with pollutant source classification and real-time data visualisation

Jack Hurley, Jack Mackintosh, Sam ShackletonPeople spend upwards of 90% of their time indoors. The increase of chemical pollutants in products and more stringent building energy efficiency requirements are conducive to poor Indoor Air Quality (IAQ). Exposure to such environments can pose a severe health risk. This project aims to (1) develop a smart system to monitor IAQ, (2) analyse the data for possible health impacts, (3) implement an Artificial Neural Network to classify the source of pollutants and (4) design a responsive web application for real-time visualisation of data and status notifications.

Emotional analytics of social debates based on visual and verbal features

Fangyin Wu, Xiaojun Mu, Yueyi JiaVerbal messages often lead to ambiguity and may not clearly express the speaker’s attitude. Body language and non-verbal cues are very important in communication. This project combines verbal messages, tone messages, facial expressions and gestures by allowing users to express their sentiments via voice and video message, to indicate that the attitudes of users in the social debate: strongly agree, agree, disagree or strongly disagree. Our main methods are feature extraction and machine learning.

Golf practice animation Wei Gao, Le Xu, Sijie LiuA system is produced to measure the golf shot information when leisure golf players are practicing at their own place, for example in the backyard. The acquired information includes ball launching speed, spin rate of golf ball, predicted trajectory, which can be used to correct a golfer’s swing.

Well-being monitoring for elderly citizens

Wenqi Lei, Yujie Zhan, Qingyuan YangThis project aimed to design a smart hardware and a paired smartphone application to detect users’ falling motion. When the device detects that the user has fallen down, it sends an alarm to the mobile phone application and notifies the user’s family of the accident. By this detect-alarm system we will be able to detect accidents in a timely manner and take measurement effectively.

A campus recognition system of the University of Melbourne

Xin Qi, Kenan Yang, Liqing ZhouAugmented reality (AR) is a live direct or indirect view of a physical, real-world environment with elements that are augmented (or supplemented) by computer-generated sensory input such as sound, video, graphics or GPS data. In practical application, travellers use AR to access real time informational displays regarding a location, its features and comments or content provided by previous visitors. Advanced AR applications include simulations of historical events, places and objects rendered into the landscape. This industry-based project involves machine learning, big data and image recognition.

MUR-E low voltage systems Hermawan Mulyono, Andrew Elsegood, Nick Hunn, Arjun BatraLow Voltage Systems is broadly responsible for electrical non-tractive systems on the electric car. This project specifically involves the design, manufacture and implementation of safety and data acquisition systems critical to high vehicle performance during race.

MUR-E accumulator Foad Munir, Williem Kartasasmita, Christian RatnapalasariThe accumulator is primarily responsible for cell selection and implementation in the electric car. This involves designing the battery, appropriating suitable charging and isolation conditions for it, as well as suitable breaking at high voltages and currents - ensuring the safety of all electrical systems is kept top priority at all times.

MUR-E battery management systems

Edwin Austine, Tony Zhang, Lucas ScodellaroThis project developed an effective and reliable battery management system for the electric car. Broadly, this comprises of the development and implementation of the battery control system to ensure the battery’s longevity, efficiency and safety. In addition, the team also handles monitoring, balancing and estimation of the battery pack.

MUR-E electric powertrain Louise Lawson, Mark Robinson, Jesse ScaramuzzinoThe electric powertrain is primarily responsible for the analysis and control of the alternating current (AC) motor and inverter, taking power from the accumulator, transforming it into mechanical energy and transmitting it to the sprocket. This involves linear and non-linear modelling of the permanent magnet synchronous salient pole motor and inverter implementing space vector modulation, system identification, mapping of operational characteristics, stability analysis, controller design and implementation.

Build and extend a smart meter Yitao Deng, Wenwen Liu, Ziwei LanThis project is to build and extend a smart meter, which can measure power consumption in a short period of time and report it automatically. This project contains both hardware and software tasks. Particularly, it aims to enhance and extend previous models, achieving a less than 1% error. An analogue front end circuit is implemented and improved to achieve more accurate measurement. A Raspberry Pi controller and customer-friendly interface is connected to the device, which enables the device to display the measurement to customers and transfer it to the energy company.

Automatic camera positioning Chao Chen, Siqi Wu, Yang LiuThe aim of this project is to build an automatic camera positioning system that tracks a user-selected moving target. The project involves a combination of image processing algorithm development, hardware development and system implementation on the selected platform. Our goal is to build a tracking system that is general, compact, affordable and easy to use. The system could easily be adapted for applications in video surveillance, video conference, lecture capturing, and video blogging with minimum modification. The highly modular nature of the system enables effortless porting between different hardware platforms so that a balance between performance and budget could be easily achieved for most applications.

MUR-E integration Max Ueda, Brennan Lam, Shiddij ShresthaIntegration is primarily responsible for creating and implementing high level concepts for the electric car. This responsibility is enforced through management of subsystem milestones leading up to overall timeline. To supplement this, integration also facilitates procurement of components for the car and manages faculty relations to organise appropriate facilities and training for the team. Most importantly, this team ensures the safety of the overall project has been kept in check.

Electronic chess board Bryce Widdicombe, Nicholas JonesChess is a strategy board game played by two players and usually played under a specific time constraint. Duration of the game varies from five minutes to seven hours, depending on whether it is a rapid or classic game. The time is controlled using a specialised game clock with two displays, one for each player. Customarily, after each player makes a move, the player has to manually press the timer button of the clock to start counting the opponent’s time. We have developed an electronic chess board with Bluetooth to communicate to a digital game clock that counts time automatically after every move. This enables players to fully concentrate on the chess game and not worry about pressing the clock timer.

Multi-room music system Daniel Rebellato, Ethan Clarke, Sam PritchardMost households now have a form of centralised data storage such as a computer, server or network attached storage system that can send music via Ethernet or wirelessly to other devices around the house. However, the amplifiers and speakers required to play this music out loud are bulky items and therefore tend to be located in a single room. Implementing a multi-room music system would usually require the addition of extra speakers and cabling to span the additional distance to the other rooms. This project involved the design of a small receiver module that can act as a music sink and stream music from a network storage location, permitting connection directly to small speakers or external amplifiers. Multiple receiver modules can be synchronised and placed throughout a house, thus creating a seamless multi-room music system, using simple amplifiers that do not need to be network-connected.

XBee locating Fangzhou Liu, Junyi Zhai, Jiayi LiuThis project is aimed at locating using XBee modules. To identify a movement point in a specific range place, the methodology is to judge the signal strength in different locations. To find a more accurate and more energy efficient method, two different methods are chosen by changing the role of receiver and transmitter. Wireless network is used to transmit the data of the movement point to the register, and the analysis result will be displayed on the computer.

Exploring molecular communications

Anthony Godfrey, Tim Schiesser, Eric BendtsenSending messages using electromagnetism is school an outdated process. Inspired by biological processes such as synapses firing in the brain, we designed a communication system that sends messages using molecules instead. Our system uses an alcohol spray for transmission, and a resistor-based alcohol sensor on the receiving end, in order to communicate across an open air environment and through a closed channel. Through the magic of bin coding, our molecular communication system achieves throughput in the single bit per second range, and helps provide insight into potential nanoscale communication options where electromagnetic communication is not feasible.

Lightweight conductor temperature sensor

Syed Sameer Hasan Mustafa, Muhammad Zia Ul Qamar, Chen WuOur aim is to use the Internet of Things methodology and develop a wireless temperature sensor to be used inside overheating switchboards. The sensor and the transmitter are to be self-powered, that is, energy has be to be harvested from the environment in order to get the sensor working. All this is to be done with multiple such sensors working simultaneously at the same time and give us an effective real time monitoring system.

Real-time wireless patient monitoring in intensive care units

Gurbir Singh Pannu, Chiyu Xie, Ao LeiThe Wireless Patient Monitoring System is developed in collaboration with the Royal Melbourne Hospital, to reduce the network of cables from patient-sensor interfaces that limit patient mobility in Intensive Care Units.

This is a novel and innovative system consisting of wearable wireless body sensor network and a software management hub to collect and display a set of patient’s health information.

The project demonstrates a portable prototype of Electrocardiograph (ECG), Electroencephalogram (EEG) and Photoplethysmogram (PPG) devices working together to display the physiological data.

Live Emergency Personnel Tracking (LEPT)

Hanson Wong, Conor Langan, Dhanuka NanayakkaraEmergency services personnel often enter unknown areas and structures to perform their day-to-day roles, which poses a considerable risk to their safety. For firefighters, commanders need a ‘big picture’ view to ensure their units are placed in optimal positions, and perhaps more importantly, to monitor and proactively protect them from ‘danger spots’. Working with the Metropolitan Fire Brigade, Live Emergency Personnel Tracking (LEPT) leverages the increasing use of Remotely Piloted Aircraft Systems (RPAS) by emergency services organisations and aims to develop a ‘centimetre-accurate’ real-time tracking system that can be deployed quickly on arrival at a scene using innovative ultra-wideband communications technology.

Machine learning for image recognition purposes

James Hutcheon, Rosa LunaEveryday a huge amount of information is communicated via handwritten digits. Common examples include postcodes written on envelopes and mobile phone numbers written on forms. To convert this analogue information to a digital form with minimal human labour, machine learning attempts to classify images of digits with near-human accuracy. Our project aims to explain, investigate and train a number of artificial neural networks and evaluate their performance on the classic machine learning problem of digit recognition. We then consolidated our understanding of the problem by building our own program to convert images of handwritten digits to digital information.

Ultracapacitor charge management system

Matthew David, Michael Warton, Christina LawsonWith the increasing prominence of electric vehicles and renewable energy solutions, intelligent use of power storage devices is essential. Ultracapacitors can be used to emulate battery-like behaviour, and offer a distinct advantage over batteries in the speed with which they can supply power. However, unlike batteries, the voltage supplied by an ultracapacitor is strongly dependent on its state of charge. This project aims to design and build an ultracapacitor based power supply to mitigate this problem, and provide a variety of output voltages using digital feedback control.

Renewable powered microgrids Zheng Li, Yanshan Liu, Jianming WuThis project involves a microgrid rig comprising photovoltaic (PV) solar and wind turbine supplying variable loads. Dynamic models for each component of the rig will be established and its validity through actual measurements will be ascertained. Then issues such as voltage and frequency control, power quality and reliability will be investigated through computer simulation carried out on the derived and verified model and through measurements. An external DC/DC converter and DC/AC inverter will be designed and constructed, and appropriate control scheme will be developed.

Laser range finder Zhenyu Wang, Chenghao Jiang, Zhongchen GuoNowadays laser rangefinders are widely used in many areas, such as golf, hunting and engineering. In this project, our group developed a fundamental laser rangefinder that could roughly measure the distance between five and 600 metres.

We used a 903 nanometre wavelength to detect the distance, which is almost safe for humans. We tried to make the rangefinder work in order and reduce error to less than one metre for every measurement. Further improvement is needed to make the measurement more precise, also other features could be added to increase the function of the laser rangefinder.

Gesture recognition Junchi Liu, Xiaojing Ma, Baohui ZhongGesture recognition (GR) is an important tool that enables customers to interact freely with technology. Utilising radio frequency signals (RF) signals enables GR to be detected around corners, through walls and over greater distances, without requiring additional framework.

In this project we aim to design a sensor that turns the human body into a physical remote control. The challenge is to investigate RF signals by isolating and mapping signal disturbances caused by nearby gestures. A compact system is created to achieve GR in real time and capable of relaying signals to local media devices, that is, to change a song via a gesture.

Wearable multi-sensor optical heart rate monitory device

Desmond Ho Yi Zhen, Xiaoyi Cheng, Satyam KumarOptical heart rate monitor devices have become popular, and are especially used during workouts to track the intensity of physical exercises and how the body responds to a particular exercise. These watches measure heart rate with light sensors that detect changes in the amount of blood pumped by the heart. However, this technology is limited by its accuracy due to motion artifacts problems caused by the regular muscle movements during workouts, location of sensors and optical noise. This project uses an accelerometer and filters to remove the error caused by motion artifacts.