Data-driven discovery methods in clinical and epidemiological sciences are gaining increasing popularity supported by rapid advances in computational sciences, however, their success is contingent on the availability of comprehensive and well-curated datasets. Sleep disorders are highly prevalent in the Australian community and closely interlinked with a range of adverse social, economic and health outcomes. This has attracted increasing policy focus on these issues in recent years. Unfortunately, only a handful of sizable public datasets for the study of sleep and its disorders exist. These are representative of international cohorts and suffer from various limitations in their data quality and make-up.

The Big Sleep ACT is an applied, multi-disciplinary research project that has received public funding to establish a sleep research dataset through collation, curation and organisation of clinical records taken from a large existing database of sleep patients from the ACT and the surrounding region. This data has been offered to researchers by a Canberra-based clinical operator so that its insights and broader public benefit may be explored, maximised and publicised. The volume, technical quality and internal consistency of this data holds promise for the creation of a high-impact national and international research asset. The primary output of this work will be a lasting tangible asset, i.e., an organised databank, which will serve as a basis for future basic and applied research in a variety of health and technical disciplines, both directly and indirectly related to sleep, including but not limited to:

• Sleep medicine
• Health informatics
• Population health
• Advanced signal processing
• Data Science and Machine Learning

Secondarily to the dataset creation activity, the project is also set to engage in physiological modelling and data-driven discovery activities, particularly in relation to data-driven phenotyping of widespread disorder Obstructive Sleep Apnea and its therapy.