Name: Aengus McCullough
Institution: Newcastle University
Research: Scalable Road Traffic Monitoring using Grid Computing
Recent technological advancements in sensing systems have caused an explosion in the volume of data relating to real world phenomena. For example systems monitoring weather, road traffic and natural disasters such as floods, earthquakes, tsunamis and volcanoes all generate a large amount of data. Typically this data is geographically referenced and is most useful in the short term.
Accessing such data over the internet has been greatly facilitated by the introduction of Sensor Web Enablement (SWE), a specification framework defined by the Open Geospatial Consortium that provides a standard web service based interface through which to task heterogeneous sensor networks and retrieve their observations. However, processing the large volume of data generated by geographical sensors in a short time period presents an enormous computational challenge. This is where grid computing can help.
Aengus’s research focused on processing near real-time geographic information delivered through SWE interfaces using Grid computing infrastructure. His research is helping to drive technology forward by integrating emerging software standards in Sensor Web and Grid Computing. Furthermore, his research attempts to classify near real-time processing operations on geographically referenced data with respect to Grid computing architectures, thus outlining the types of application for which this architecture is suitable. Aengus used traffic monitoring and routing as a real-world application case study through which the strengths and weaknesses of current specifications can be evaluated and different types of processing operations can be considered.
One geoprocessing operation Aengus investigated was termed map-matching and involves matching GPS tracks from a fleet of vehicles to the road network. Aengus used the GridSAM job submission service to run open ended compute jobs on the NGS which retrieved GPS tracks in near real-time from a SWE service, combined this with base map data of the road network and publish the matched position to another SWE service.
The map-matching operation is not particularly computationally intensive when data from a single vehicle over a short time period is considered. However, the benefit of executing this computation on the NGS is clear when it is necessary to perform the operation for several hours a day for a large fleet of vehicles. For example when managing a fleet of city council vehicles, map-matching needs to be carried out as a pre-cursor to network analysis functions such as allocating vehicles in response to an incident, or routing vehicles between two location. Aengus successfully tested the system with up to 250 vehicles.
Aengus explained that “The NGS has made endless CPU hours available to me through a standards based interface; it has been an invaluable resource for my research”.
Project funding - EPSRC / School of Civil Engineering & Geosciences, Newcastle University
Grant no. - EO/P503612/1
Project PI - Philip James, Dr Stuart Barr
Download a summary slide of this case study.
