UTS team works to deliver real-time flood sensing with 5G
A team at the University of Technology Sydney is working on bringing some of the integrated sensing capabilities that are expected to be part of 6G to Australia’s 5G networks, as part of an effort to help protect people from the risk of flash flooding.
A team led by the NSW Connectivity Innovation Network’s Technical Director, Distinguished Professor and Director of the UTS Global Big Data Technologies Centre, Jay Guo, is working with the NSW State Emergency Service on the Flood and Storm Intelligence project.
Professor Guo, the principal author of the world’s first 6G antenna engineering book, told CommsDay that the project is seeking to tackle two key challenges facing the SES when it comes to flood response.
The first is that the SES relies heavily on data from the Bureau of Meteorology for rainfall forecasting.
“But the Bureau forecasting is not granularised, not localised,” he said. The Bureau’s forecasts are for large areas and not necessarily real time, he noted. “So therefore we cannot precisely predict localised flooding, particular flash flooding, because flash flooding happens in a small area. All of a sudden it comes and before you know it, it might have gone. But in between, people’s lives are impacted.”
The second major challenge is a lack of visualisation and prediction tools. “For example, if the Bureau said ‘That place is going to have rainfall of 100mm’ — there is no tool for them to tell what that means for the community, right? Which house, which street, which road might be under water,” Guo said.
There are some tools that can simulate how the flood may evolve in time, but they involve complex mathematical simulations and for a large area that could take hours, if not days, he said. “That’s kind of impractical for real-time risk management and flood prediction.”
The project emerged from discussions with the SES in 2021, Guo said. Although the researchers didn’t have any ready-made solutions, they had some “big ideas” to challenge the twin challenges faced by the emergency response organisation.
“The first idea is to use cellular network — 5G in particular — as a sensor,” he said, noting the broad coverage 5G is expected to have in NSW meant that it could potentially act as a substitute for billions of dedicated sensors by employing 5G beams.
“The base station, the user terminal, they are all transmitting signals and the signal is received on the receiver side,” Guo said. “If the environment changes, the received signal changes; basically the signature of the environment is embedded in the received signal. By detecting such signatures, we can sense how the flooding is evolving. We can sense the rainfall, infiltration, the level of water in any street, any area, et cetera.”
“Potentially we can have cost-effective large area sensing — both large and localised — because base stations are everywhere, they have beams, they can point their beam in any location, any area required,” Guo said.
The project team is using digital twin technology to address the SES’s second challenge around real-time visualisation and prediction. While digital twins are not new inventions by UTS, they haven’t been used as a solution for flood prediction, Guo said. The team has been working on advanced machine learning algorithms to accelerate the ability of the digital twin to predict and visualise floods, and has recently proved its idea works, he said.
The team is initially aiming at validating the approach of using 5G networks for flood predictions, and if it works it can be expanded across the state or even nationally, he said.
“Right now the international telecommunications community is developing integrated communication sensing for 6G, [which is] roughly seven years away,” Guo said. “But at UTS our research is quite advanced. We are the front-runners in this field. We believe we could accelerate this technology’s development using 5G, instead of waiting for 6G.”
The team has had some early discussions with mobile network operators about the technology. Deployment would involve software updates for base stations. For 5G, the approach will involve minimum hardware changes, while once 6G is deployed, the necessary hardware and software for integrated sensing and communication will be ready from the get-go.
“So if we do all the technology development now, then we’ll be ready not only to make an impact in the next few years” but also when 6G arrives and if the necessary infrastructure is available globally the technology could potentially be commercialised across the globe, he said.
Over the next two years the team aims to demonstrate the feasibility of the key technology it is developing and then expand the project to more areas in the state.
