US 11,810,349 B2
Ensuring security on the fueling forecourt
Henry Fieglein, Leander, TX (US); Kalpit Singh, Jaipur (IN); Rohith Chinnaswamy, RoundRock, TX (US); and Hob Hairston, Cedar Park, TX (US)
Assigned to Wayne Fueling Systems LLC, Austin, TX (US)
Filed by Wayne Fueling Systems LLC, Austin, TX (US)
Filed on May 28, 2020, as Appl. No. 16/885,946.
Prior Publication US 2021/0374413 A1, Dec. 2, 2021
Int. Cl. G06V 20/40 (2022.01); A62C 3/00 (2006.01); G06V 20/10 (2022.01); G06N 20/00 (2019.01)
CPC G06V 20/41 (2022.01) [A62C 3/00 (2013.01); G06V 20/176 (2022.01); G06N 20/00 (2019.01)] 16 Claims
OG exemplary drawing
 
1. A method comprising:
receiving data characterizing a video feed acquired by a camera continuously and automatically gathering images, the camera being oriented toward and including a field of view of a forecourt of a fueling station;
continuously monitoring the video feed for hazards, the monitoring including performing automatic hazard detection on the video feed using at least one predictive model that predicts a presence of a hazard within the forecourt of the fueling station;
transmitting a command to a forecourt controller, the forecourt controller being configured to manage operation of the fueling station, the command causing the forecourt controller to deactivate at least a portion of the fueling station, and the command identifying the at least the portion of the fueling station to be deactivated;
causing a graphical prompt to be shown on an end user device, the graphical prompt being indicative of the predicted presence of the hazard and the graphical prompt including a first response option that designates the predicted presence of the hazard as a false alarm and a second response option for activing a fire suppression system disposed in the forecourt;
receiving, from the end user device, data characterizing a designation, by a user of the end user device interacting with the graphical prompt, of the predicted presence of the hazard as a false alarm; and
training the at least one predictive model by at least providing the data characterizing the designation to the at least one predictive model.