The shakedown: developing an indoor-localization system for quantifying toilet usage in offices

• Published in: Architectural science review 2020-07, Vol.63 (3-4), p.325-338
• Main Authors: Doherty, B., Gardner, N., Ray, A., Higgs, B., Varshney, I.
• Format: Article
• Language: English
• Subjects: Data science; evidence-based design (EBD); indoor-localization; machine learning; post-occupancy evaluation (POE); sensors; toilet usage
• ISSN: 0003-8628; 1758-9622
• DOI: 10.1080/00038628.2020.1748869

The design of sanitary facilities in Australia is subject to regulations prescribing minimum provision. In commercial office buildings, this is tied to male and female employee numbers. These requirements are derived from mathematical models, using queuing theory. Evidence of inadequate sanitary provision in numerous contexts points to the necessity to refresh the data, and thinking, that underpins these regulations. Collecting empirical data on human occupancy in sanitary facilities using data science methods is a new way to achieve this and support a shift towards the evidence-based design of sanitary spaces. Accordingly, this article outlines the development and implementation of a novel, privacy-preserving, indoor localization system (ILS) that combines sensors and machine learning to collect and analyse toilet usage data in an office. By evaluating the system's capacity to identify occupancy patterns this research contributes to scholarship on ILS methods as well as a valuable data-set on Australian toilet usage..