Skip to main content

Spatiotemporal Modelling

We have seen an exponential increase in data flow and availability from large numbers of sensors. The DSS Group at CSIRO works on using this data to create models of the underlying sensed phenomena in order to more effectively monitor these phenomena in the future.

Mobility Mining

Distributed sensing systems, from dedicated devices in sensor networks to social sensing from social networks, are generating huge volumes of data from moving objects, animals, and people that captures their movement patterns. Understanding these patterns is critical for delivering a new generation of public services (for instance in transport, planning, or health), ecosystem management, and operational efficiency in logistics. Our group analyses data from diverse sources to characterise the underlying movement patterns.

The distribution of probability in space that an individual with a movement orbit up to 10km will travel around their centre of gravity at (0,0). This was extracted from geo-tagged Tweets in Australia
The distribution of probability in space that an individual with a movement orbit up to 10km will travel around their centre of gravity at (0,0). This was extracted from geo-tagged Tweets in Australia
Spatial tweet distribution of medium distance movers (10-100km) within Australia's south-east corner.
Spatial tweet distribution of medium distance movers (10-100km) within Australia’s south-east corner.

Spatiotemporal Modeling of Animal Movement for Optimal Sampling of GPS positions

Forge Loop of Animal tracking data
Forge Loop of Animal tracking data

Accurate and energy-efficient location tracking of mobile objects is an important component of context-aware services and applications. While GPS receivers offer high accuracy positioning, energy constraints of battery powered devices necessitate duty-cycling of GPS to prolong the system life- time. Energy harvesting can extend the life of tracking appli- cations to a near perpetual operation. However, if movement patterns and energy resources change frequently, GPS sam- pling needs to adapt in real-time to achieve optimal position- ing performance. This work uses energy- and mobility- aware scheduling framework for sampling of GPS in long- term tracking applications.

Spatiotemporal Modelling by Fusing Sensor Data with Existing Data Sources

Conventional monitoring of physical spaces uses manual sample collection to build coarse-grained spatiotemporal models of monitored phenomena. With the increased availability and affordability of low power sensors, using data from these sensors to validate and evolve existing models presents both a challenge and an opportunity. Conventional methods have undergone long-term testing yet they only provide coarse-grained information, while low power sensor networks are relatively new yet provide high resolution spatial and temporal data. The DSS Group is working on new methods for fusing the different data sources from sensor networks and conventional methods for more reliable and representative spatiotemporal models of the underlying environment. We have already applied this capability to the problem of mine rehabilitation, where mining companies must maintain a lease on the mine site until they can show they have restored the ecosystem to its original state.

Illustration of the complexity involved in estimating vegetation growth. We propose to fuse data from multiple methods to achieve best estimates and assessment of rehabilitation success.
Illustration of the complexity involved in estimating vegetation growth. We propose to fuse data from multiple methods to achieve best estimates and assessment of rehabilitation success.

Related Publications

M. Shahzamal, R. Jurdak, R. Arablouei, M. Kim, K.  Thilakarathna, B. Mans, “Airborne Disease Propagation on Large Scale Social Contact Network,” In proceedings of the Second International Workshop on Social Sensing (SocialSens), as part of CPSWeek, Pittsburg, USA, April, 2017.

M. Kim, R. Jurdak, “Heterogeneous Social Signals Capturing Real-world Diffusion Processes,” In proceedings of the Second International Workshop on Social Sensing (SocialSens), as part of CPSWeek, Pittsburg, USA, April, 2017.

K. Zhang, R. Arablouei, and R. Jurdak, “Predicting Prevalence of Influenza-Like Illness in Australia From Geo-Tagged Tweets,” In proceedings of the 1st International Workshop on Social Computing (IWSC), as part of the 26th International World Wide Web (WWW) Conference, Perth, Australia, April, 2017.

K. Zhao and R. Jurdak, “Understanding the spatiotemporal pattern of grazing cattle movement,” Nature Scientific Reports, 6:31967 EP, August 2016.

Thomas B, Jurdak R, Zhao K, Atkinson I (2016) Diffusion in Colocation Contact Networks: The Impact of Nodal Spatiotemporal Dynamics. PLoS ONE 11(8): e0152624. doi:10.1371/journal. pone.0152624

P. Sommer, J. Liu, K. Zhao, B. Kusy, R. Jurdak, “Information Bang for the Energy Buck: Energy- and Mobility-Aware Tracking,” In proceedings of The International Conference on Embedded Wireless Systems and Networks (EWSN), Graz, Austria, February, 2016.Winner of Best Paper Award

B. Thomas, R. Jurdak, I. Atkinson, “Opportunistic Content Diffusion in Mobile Ad Hoc Networks,” Ad Hoc Networks, Volume 45(15) Pages 34–46, 2016.

I. Purnama, R. Jurdak, K. Zhao and N. Bergmann, “Characterising and Predicting Urban Mobility Dynamics By Mining Bike Sharing System Data,” In proceedings of 12th IEEE International Conference on Ubiquitous Intelligence and Computing (UIC), Bejing, China, August 2015.

R. Jurdak, K. Zhao, J. Liu, M. AbouJaoude, M. Cameron, D. Newth, “Understanding Human Mobility from Twitter,” PLOS ONE, 10(7): e0131469. doi:10.1371/journal.pone.0131469. July, 2015.

J. Liu, K. Zhao, S. Khan, M. Cameron, R. Jurdak, “Multi-scale Population and Mobility Estimation with Geo-tagged Tweets,” In proceedings of 31st IEEE International Conference on Data Engineering (ICDE) Workshop BioBAD 2015, Seoul Korea, April 2015.

M. Trad, R. Jurdak, and R. Rana, “Guiding Ebola Patients to Suitable Health Facilities: An SMS-based Approach,” F1000 Research 4:43, February 2015.

B. Thomas, I. Atkinson, and R. Jurdak, “Content Diffusion in Wireless MANETS: The Impact of Mobility and Demand,” In proceedings of the IEEE Wireless Communication and Mobile Computing (IWCMC) Conference, Nicosia, Cyprus, August 2014.

B. Kusy, R. Rana, P. Valenca, R. Jurdak, and J. Wall, “Experiences with Sensors for Energy Efficiency in Commercial Buildings,” In proceedings of the Fifth Workshop on Real-World Wireless Sensor Networks (RealWSN), Como Lake, Italy, September 2013.

S. Purdon, B. Kusy, R. Jurdak, G. Werner-Challen, “Model-free HVAC Control using Occupant Feedback,” In proceedings of IEEE International Workshop on Global Trends in Smart Cities (GoSmart), co-located with IEEE International Conference on Local Computer Networks (LCN), Sydney Australia, October 2013.

R. Rana, B. Kusy, R. Jurdak, J. Wall, and W. Hu, “Feasibility Analysis of Using Humidex as an Indoor Thermal Comfort Predictor,” Energy and Buildings, Vol. 64, pages 17-25, September 2013.

Corke, T. Wark, R. Jurdak, W. Hu, P. Valencia, and D. Moore. “Environmental Wireless Sensor Networks,” Proceedings of the IEEE, Special Issue on Emerging Sensor Network Applications, Vol. 98, No. 11, pages 1903-1917, November 2010. (Invited paper)

Projects