Better reef monitoring through simulation
Researchers at the QUT Centre for Data Science are working with the Australian Institute of Marine Science (AIMS) to optimise reef monitoring processes and improve asset management.
Associate Professor Paul Corry is developing a data-based simulation that will model reef monitoring operations. Scientists at AIMS will be able to explore how changing their monitoring schedule and adopting new technologies can impact their efficiency on the water — all with the click of a button.
The right model
The model is built on the back of data gleaned from GPS tracking from AIMS’s fleet of monitoring vessels, giving accurate information on the geographical parameters of monitoring operations including the sequences of reef visits and travel times between them.
Corry and his team combined this with data from AIMS scientists, mapping the entire process of human and equipment interactions on a monitoring mission to model the timing and dependence of sequential and parallel actions.
“There are a lot of challenges in simulating an operation like this,” Corry said.
“Monitoring involves visiting multiple reefs and undertaking multiple sampling methods over the course of one or two weeks, so there are a lot of elements at play.
“There are restraints around when crews can work, when they need to take breaks, how limited resources can be shared between teams, and when vessels may not be able to move between locations due to tides, weather or other factors.”
The model will allow the team at AIMS to input different scenarios and model combinations of staffing, technologies, vessels and routes to understand what limits the current methods and what types of approaches they can use to optimise their resources that support the work.
“Say there are two vessels with different capabilities, and a range of reefs that need to be monitored — the team needs to decide how many staff should be on each vessel, and how many reefs each vessel will visit,” Corry explained.
“They can plug that information into the model and it will predict the timing for the whole operation.
“From there, they can understand the limitations of their current methodologies. They can see how investing in new technologies, resources and infrastructure can allow them to scale the monitoring operations to efficiently meet the needs of a changing environment.”
The model also draws on historical weather data, which will make the simulation able to quantify uncertainty in the face of potential adverse weather events.
“You never know what mother nature is going to throw at us, and that can hamper accuracy and cause random disruptions in operations,” Corry said.
“Building that weather data into our model gives a more accurate prediction of the possible range in mission timing to 95% level of confidence.”
QUT has a long-standing partnership with AIMS, providing tailored solutions to their specific industry issues.
This six-month project builds on initial work that AIMS staff produced, bringing the technical expertise needed to boost their simulation and bring the project to fruition.
“The fact that AIMS took the initiative and started developing the simulation themselves gave us a great foundation to build on,” Corry said.
“The team at AIMS understand the complexities of the model and the types of data needed to support it, so they were well prepared to see the project through to development.”
Efficient, interactive and immediate
The project will improve operations for AIMS and the reef in multiple ways.
“We’ll save a lot of time in the planning process, as we’re replacing a labour-intensive, tedious number-crunching task with a streamlined computer program,” Corry said.
“The model also allows AIMS to try out more scenarios and alternative monitoring plans in simulation, which they can then refine and optimise to improve efficiency and make better use of their available resources.”
The model is adaptable, so the team can easily simulate new processes or technologies to improve aspects of the operation, and assess the impact before making the financial investment.
“This ability to identify the best places to invest in people, resources and technology is a game changer in ensuring that AIMS has the capability it needs to meet future demands,” Corry said.
“For example, the team might look at replacing a human diver with an autonomous system — while this may save money in the short term, it may be far less efficient than a skilled human, costing more time over longer periods.”
Initially the model will simulate single missions, but with the right data can expand for a 12-month view, which will show the big impact small changes can have.
Researchers are also interested in using computational optimisation methods to further improve the model’s capability.
“By developing clever algorithms, we can program the model to run these what-if scenarios without human input, and try thousands of scenarios autonomously to find the optimal combination of resources and time,” Corry said.
Improving processes for reef monitoring will have a lasting impact on reef health in the long term.
“The Great Barrier Reef alone covers an area the size of Italy, which is a lot bigger than people realise, and there’s a lot of territory that the team at AIMS need to cover for their monitoring and science programs,” Corry said.
“It’s impossible to cover the whole reef, but we can help them monitor as much as their assets and resources will allow.”
More information
Explore more research at QUT’s Science and Engineering Faculty
Find out more about the QUT Centre for Data Science
