River Sentry: Predictive Vs Reactive (Part Four)
River Sentry: Predictive Vs Reactive (Part Four)
We define River Sentry as a reactive flood warning system. What does that mean? Most flood systems attempt to provide predictive
notification from hours to days. They use an extensive suite of water
sensors in various tributaries, rain gauges, weather radar, soil moisture sensors
and other technologies. These combine to create hydrologic forecast models,
create thresholds, then issue alerts.
River Sentry’s reactive system is different. It accepts a
minimized alert window in pursuit of higher performance, resiliency and accuracy.
It positions water sensors well outside moderate flood thresholds existing on the line between
significant and dangerous. Most flooding will not activate it. This is by design
for metrics explained as probability of detection (POD) and False Alarm Ratio
(FAR).
POD is an important parameter as low POD scoring is indicative
of a system with poor detection ability, a dangerous shortcoming. A perfect POD
score is (1); one detection per one event, zero missed events.
False Alarm Ratio (FAR) is the occurrence of alarms compared
to actual flood events. Low FAR is indicative of an accurate and trustworthy system.
High FAR is indicative of a system that “cries wolf”, creates “warning fatigue”
and may become ignored by its exhausted audience.
Trust is the third parameter we are introducing. Do people
trust it? Do they believe in its methodology and react to its warning? A low
trust system is doomed to be blocked and dismissed.
Fire code is the fourth parameter. This is an interesting attribute unique to River Sentry which sets "effect" benchmarks designed to address immediate danger and sleep. All River Sentry devices
and methods must adhere to commercial fire code for the purpose of decibel levels and lighted egress paths.
River Sentry’s defining algorithm and ethos is “POD 1/FAR 0/T100/FIRE”, further defined as a perfect detection ratio, zero false alarms, 100% trust, and fire code adherence. Such a score would make it the most accurate flood detection and warning system in existence.
How do we get there? Join us later for part five as we dig down and explore terms and methodology.
Comments
Post a Comment