Digital Cities project researching smarter use of sensor data

News / A2I2 Media / October 24, 2022

Applied Artificial Intelligence Institute (A2I2research fellow Dr. Anj Simmons has had a paper accepted at the 2022 iteration of the prestigious International Semantic Web Conference.

Dr. Simmons’ research project, conducted under a two year grant from the Office of National Intelligence, examines how diverse types of data from cameras, motion sensors, and even social media, may be integrated into a single network to detect and escalate suspicious events more reliably.

Despite a growing number of sensors installed in buildings and cities for automatically monitoring and controlling everything from building security to traffic control, machines don’t currently understand the ‘big picture’ of what is going on. 

They don’t piece together the clues from different information sources in the same way as a human would, causing them to fail to detect certain kinds of activity (false negatives) or misinterpret non-suspicious activities as suspicious (false positives).

 

Video: Simulation of a digital city, with prediction of what is occurring shown in bottom-right of video.

 

Dr. Simmons aims to support machines to more intelligently integrate and reason about the underlying cause of sensor observations, thus making the best possible use of the information available, and ultimately, making cities safer for citizens without the need for an excessive number of sensors.

By creating a platform to integrate and reason about sensor sources, we can visualise what is happening in real-time and hopefully detect and respond to attacks sooner,” Dr. Simmons said. 

The International Semantic Web Conference is the premier international forum for the Semantic Web and Knowledge Graph Community. This year’s conference will bring together researchers, practitioners and industry specialists to discuss, advance, and shape the future of semantic technologies.

To read Dr. Simmons’ paper, SignalKG: Towards Reasoning about the Underlying Causes of Sensor Observations, please visit: https://arxiv.org/abs/2208.05627.

For further details or collaborations, please reach out to Dr. Anj Simmons at a.simmons@deakin.edu.au