The Need
Public restrooms — in malls, transit stations, offices, and institutions — need cleaning at high frequency, and that cleaning is done by hand. The idea behind this project was a system that cleans the toilet bowl automatically after every use, with no contact and no button press.
The Two Real Problems
When to clean. Running on a timer wastes water and cleaning agent when there's no traffic, and falls behind when there is. The solution was to use the flush as the trigger event — a reliable signal that a use has ended, one that demands nothing from the user and, unlike a presence sensor, raises no privacy concerns.
Where to clean. Toilet bowls are not identical — shape, size, and fixture positions vary between models and installations. A system engineered for one bowl is not a marketable product. This is where software enters: an algorithm maps the bowl's shape at installation and learns the cleaning path that fits it, instead of relying on fixed dimensions. Software here saves expensive mechanical customization — a pattern we explore in embedded software development.
Development Scope
The project combined mechanical design of an arm that covers the required surface within the tight space of an existing stall; a fluid system that meters cleaning agent, prevents dripping, and drains fully so no liquid pockets form; sensors and control for flush detection, obstacle detection, and bowl mapping; and materials that stand up to cleaning chemicals and constant humidity.
The Constraint That Shaped the Design
A product like this is installed in existing stalls, not stalls built for it. Installation volume is fixed, anchor points are limited, and sometimes there is no power outlet near the bowl. Those constraints defined the geometry and the power architecture more than any other consideration — a clear case where precise requirements definition up front, as we describe in new product development, saves half the project.
Products for public environments are judged on reliability and maintenance, not features. A system that needs weekly service won't survive the market even if it cleans perfectly.
Have an automation idea for a stubborn manual chore? Contact Projects House and let's evaluate it together.
Photos from the project
