Research Projects

1. Video Analysis

Slaven Marusic, Aravinda S. Rao, Fu-Chun Hsu and Marimuthu Palaniswami
Partners: ARUP, Melbourne Cricket Ground, SenSen, Australian Research Council

The prevalence of camera networks for surveillance, together with the decreasing cost of infrastructure, has produced a significant demand for robust monitoring systems. Current systems offer limited functionality, particularly in their reliance on centralised processing of gathered information. This project addresses end-to-end system challenges of wireless visual sensor networks. Integrating developments across the spatial, spatio-temporal and decision domains, the project will incorporate distributed sensor network technology with intelligent fusion of information, to deliver unique long-term behaviour analysis capabilities for efficient planning in highly crowded environments.

2. Internet of Things for Urban Noise Monitoring

Slaven Marusic, Aravinda Rao, Yee Wei Law and Marimuthu Palaniswami
Partner: Melbourne City Council, ARUP, Australian Research Council

By 2050, 70% of the world’s population and over 6 billion people are expected to live in cities and surrounding regions. So, cities will need to be smart, if only to survive as platforms that enable economic, social and environmental well-being. Smartness of a city is technologically enabled by the emerging Internet of Things, which can be seamlessly integrated into the urban infrastructure (transport, health, environment, etc.) and thus forms a digital skin over the city. The city needs smart networking solutions in order to fully utilize the data collected and further define a common operating picture of the city, which is useful in policy decisions. The aims are to create a Smart City capability through seamless urban environment monitoring via large scale sensing, data analytics and information representation. Interconnection of sensing and actuating devices as Internet of Things (IoT) addresses the ability to share information across platforms through a unified framework, developing a common operating picture for city management. The interpretation of events and visualisation of information for end users will ensure sustainability and higher quality of life in the urban environment. Under the Smart City heading, the ISSNIP group is participating in several ongoing IoT projects, described and grouped according to the technical themes below.

The aims include development of low-cost noise monitoring hardware, deployment of energy-efficient noise monitoring sensor network, and development of protocols for efficient collection of high-sample rate noise data from sensors deployed over large geographical regions, and developing methodologies for analysis and visualisation of the noise pollution profile. This will offer a city council understand dynamic noise pollution profile, assess its impact on health and wellbeing, and better plan for noise reduction and desirable urban sound-scape.

3. Smart devices for ubiquitous health monitoring

Marimuthu Palaniswami and Bernard Yan, Terence O'Brien
Partners: Royal Melbourne Hospital, NeuroGlide Pty. Ltd.
Provision of healthcare services remains a critical challenge across all levels of society, with the cost burden and constrained resources limiting the accessibility to healthcare. Advances in hardware development have made available efficient, low-cost, low-power miniature devices for use in remote sensing applications. This iconic project will drive the development of smart devices for low-cost monitoring, analysis and treatment of a number of medical conditions addressing continuous monitoring, aged care, and rehabilitation.

4. Autonomous Live Animal Response Monitor (ALARM)

Aravinda S. Rao, Steven Marshall, Marimuthu Palaniswami
Partners: CAPIM, Victorian Government
With increasing human population and consequently, because of the radical urbanization, our precious water resources are being polluted at alarming rates. In Victoria, 80% of the waterways are in poor to moderate condition. A biological early-warning system can aid in forestalling the impact of the immediate pollution. The system comprises of a real-time electronic sensor platform measuring water quality with sensors alongside behavioural ecotoxicology with machine vision algorithms.

Currently catchment managers collect water samples periodically to monitor the water quality, which is expensive and time-consuming. The high-cost routine influences the measurement methodology to limited samples and thereby prevents the efforts of containing pollution. Simultaneous, continuous real-time quality measurement can help isolate the pollution by regular sampling and reduced measurement cycles in real-time at a moderate increase in cost.

The project focuses to identify the possible pollutant by monitoring physical properties in water streams, improve the health of water-dependent aquatic and terrestrial lives, and provide environmental benefits.

Design of low-power, high-accuracy sensor network poses significant challenges in terms of sampling sensor data and data communication. Continuous monitoring causes sensors to provide deviated results because of lack of calibration. For behavioural analysis, computer vision algorithms demands significant processing power for mathematical computations. Furthermore, the murky water environments deter the efforts of computer vision algorithms.