The ongoing safety of built structures such as buildings and bridges relies on the early identification of any damage. But most approaches to maintenance are time-based, scheduling inspections only at predetermined intervals of time. Dr Ali Anaissi’s research aims to convert this procedural convention to a condition-based approach, whereby continuous monitoring of a structure by networked sensors would detect early signs of damage and trigger a responsive inspection.

“The condition-based approach that my research explores is known as structural health monitoring (SHM). It is a continuous automated process that aims to detect damage in a structure using data gathered from several networked sensors attached to it. In a structure such as a building or bridge, such early detection of damage is critical to avoid further risks to life, safety and economic loss.

“The main idea of this data-driven approach is for machine learning algorithms to ‘learn’ a model from the data sensed, in order to construct a baseline. This learned model is then applied to the new continuous real-time measurements being taken by the sensors, in order to generate real-time health scores for components of the structure by comparing the new measured responses to the established baseline.

“In this way my team and I have equipped the Sydney Harbour Bridge with a large number of networked sensors to provide information about the safety of the structure.

“We have also equipped a vehicle with sensors to collect vibration data during its journey on the roads, and then applied machine learning algorithms to assess the condition of those roads and detect any damage.

“I started working in the field of machine learning and data mining when I began my PhD in 2009, and later brought my research to the University of Sydney.”

DrAli Anaissi

BIO

SCHOOL

FIELDS OF RESEARCH (2020)