Is your smoke detector working properly? Robust fault tolerance approaches for smoke detectors

Billions of smoke detectors are in use worldwide to provide early warning of fires. Despite this, they frequently fail to operate in an ongoing fire, risking death and property damage.
A significant fraction of faults result from drift, or reduced sensitivity, and other faults in smoke detectors’ phototransistors (PTs). Existing approaches attempt to detect drift from the PT output in normal conditions (without smoke). However, we find that drifted PTs mimic the output of working PTs in normal conditions, but diverge in the presence of smoke, making this approach ineffective.

This paper presents two novel approaches to systematically detect faults and measure and compensate for drift in smoke detectors’ PTs. Our first approach, called FallTime, measures a PT “fingerprint,” a unique electrical characteristic with distinct behavior for working, drifted, and faulty components. FallTime can be added to many existing smoke detector models in software alone, with no/minimal hardware modifications. Our second approach, DriftTestButton, is a mechanical test button that simulates the behavior of smoke when pressed. It offers a robust, straightforward approach to detect faults, and can measure and compensate for drift across the entire smoke detector system. We empirically evaluate both approaches and present extensive experimental results from actual smoke detectors deployed in a commercial building, along with custom-built smoke detectors. By conducting tests with live smoke, we show that both FallTime and DriftTestButton perform more effectively than existing fault tolerance techniques and stand to substantially reduce the risk that a smoke detector fails to alarm in the presence of smoke.