The inflow of wastewater into pumping stations, and from there, into wastewater treatment plants, follows a bi-modal curve that reveals the daily demand placed upon the system by its habitual users. The entire network, including the throughput capacity of the wastewater treatment plant, must cope with the peak demands, which means that the network is underutilized for around 80% of the day.
In this talk, we present results from a pilot study where we tested a theory that we can change the shape of the bi-modal curve by utilising an optimisation and data-driven approach for the way we use the pumping stations. We show that we can achieve our goal even though we only indirectly control the flow from the pumping stations, and though we provide a static solution for a stochastic problem.
Principal Data Product Owner, Water Corporation
Dr Burt obtained a PhD in Optimisation at Curtin University in 2008, for which she was awarded a Chancellor’s commendation for excellent thesis. She researched network modelling and algorithms in the Department of Mathematics and Statistics at the University of Melbourne, before heading to Vienna to complete a postdoc in Transportation algorithms at the Austrian Institute of Technology. She returned to the Computer Science department at the University of Melbourne to develop decomposition algorithms for scheduling problems. In 2015 she was awarded the President’s Mid-career Plenary Award for MODSIM.
From 2016, she has worked in consulting, developing optimisation and artificial intelligence solutions. She has published 15 peer-reviewed articles and co-authored a book on these topics. She currently works in the Analytics Centre of Excellence at the Water Corporation, where she creates high benefit data products for stakeholders within the organisation.