Simplified artificial intelligence fire recognition algorithm (SAIFRA) can be embedded in any microprocessor based alarm or detector having additional 1.5 KB memory. The final product does not require PCB or plastic housing changes. Agency listing is also minimal since the product is not physically modified.
SAIFRA is Smart
Distinguishes between smoke box, burger broil, smoldering, and fast flaming polyurethane fires resulting to a wider UL production range
Differentiates between UL fast flaming polyurethane and other fast fires e.g. wood crib, paper, etc.
Distinguishes between UL smoldering fire and broil with a single ionization chamber
Multi-parameter alarm thresholds (amplitude, velocity, and acceleration) allows product to quickly and accurately go into alarm
Incorporates advanced noise filtering , drift compensation, and auto-calibration
SAIFRA is Single Sensor Capable
Adding another sensor to your design significantly increases cost. Using SAIFRA, a single photoelectric or a single ionization chamber is enough to pass all UL fire tests. Our algorithm identifies the type of fire and automatically adjusts alarm thresholds. If needed, additional sensors may be used to increase resistance to nuisance alarms.
SAIFRA is Less Prone to Nuisance
With our algorithm, distinction between UL burger broil and smoldering is possible. Differentiation between fast flaming polyurethane and stove-top cooking is simple. On top of this, your alarm or detector will have a wide production range allowing for insensitive alarm thresholds. These characteristics enable your product to be very robust when exposed to nuisance fires.