River Sentry: The pursuit of perfection, POD 1 (Part Five)

Probability of Detection (POD) ratio is the major metric of the Flood Early Warning Systems (FEWS) arena. It is the measure of events divided by warnings plus missed events. The perfect ratio is 1 flood warning per flood event, no missed events. A 1/1 ratio can simply be expressed as 1…perfection. 

It sounds simple, it is anything but. The European Flood Awareness System (EFAS) is estimated to have cost over $30 million and hovers around POD .70. That ratio falls in the more mountainous and remote areas. How is some unknown new tech company in central Texas going to beat that kind of score? It’s a great question, let’s explore it.

Before we compare apples, let’s define the arena. EFAS is a regional system attempting to provide flood warning over a large geographic area utilizing a myriad of sensors and software solutions to provide predictive warnings prior to an actual flood event.  River Sentry builds localized area protection with additional downstream advanced warning using networked devices. It is a creek by creek advance aimed at providing highly resilient and instantaneous flood warning in any close protected structure employing a collection of devices. It is purely reactive and hyper focused on ENSURING it provides a crucial 5-minute warning to save your life, not ATTEMPTING to give you five hours to hook up your RV and leave town.

Five hours is better than five minutes, why build a reactive system? Poor performance and death are why. Extreme convective wet microburst activity (rain bombs) coupled with mountainous terrain drastically reduce predictive system value. In July 2021, EFAS was significantly tested for the first time. The results were horrific. 190 dead, many drowned in their homes caught asleep by the rapidly advancing water. Sleep, rising water, and proximity meet again. The flood death triangle is complete.

It turns out EFAS lacked focus on smaller tributaries and mother nature chose to drop 8 eight inches of rain in an area without. Even if they had, many of the local sirens were knocked out due to power outages. Such is the fate of complex fragile systems fraught with single point failures. Trust them at your own risk.

The River Sentry approach to POD perfection is direct and clear. Instead of disbursing detection resources across a large geo area, our towers are nestled close to people with simple tactics. Draw the line between water and people and guard it. Multiple towers protect each area; each tower has multiple water detection switches. This line is further reinforced by upstream early warning towers and where applicable, low water crossing variants. Redundancy is prime, resiliency is law. 

Our equipment is placed in the flood path forcing the conflict. Low tech resilient maritime industry water detection switches protected by metal armor imbedded in concrete await. Instead of complex data producing sensors, we opted for battle-hardened ‘Water Witches” proven in the caustic saltwater bilges of Alaskan fishing fleets. Standard equipment for boat manufactures, these switches are good at only one thing; detecting water where it shouldn’t be. River Sentry approves but we want two and add a testing circuit so we can constantly test it.

The six tower River Sentry installation at the Vistas Youth Camps on the Guadalupe in Ingram, Texas incorporate 12 such switches. Each tower can set off the other 5 wirelessly. The entire collection can be set off by upstream towers and in turn set off downstream collections. Instead of 1 computer processing data, River Sentry will use 100 customized computers mounted 8 feet high and armed with three spectrum wireless antennae. Each device will be positioned to detect two or more other towers. Single point lines of communication are prohibited.

For increased performance, we use long range lower frequency two-way encrypted Long Range Wide Area Network (LoRaWAN) communications instead of higher frequency RF which are more prone to electromagnetic disturbance associated with lighting strikes. Regular device “challenge/response” communication protocols will ensure all devices are reporting on an hourly basis. Multiply this formula and spread it across miles of river front. As each proximate facility adds its own sensors, the chance of a missed event is smashed into micro probability oblivion.

POD 1 is achievable. This is the way!

Join us soon for part 6: FAR 0 and the Wolf Ratio!