US 12,438,947 B1
Event detection system
Saleh ElHattab, San Francisco, CA (US); Justin Joel Delegard, West Chester, OH (US); Bodecker John DellaMaria, San Francisco, CA (US); Brian Tuan, Cupertino, CA (US); Jennifer Winnie Leung, Berkeley, CA (US); Sylvie Lee, Pittsburgh, PA (US); Jesse Michael Chen, San Francisco, CA (US); Sean Kyungmok Bae, San Francisco, CA (US); Angel Manalastas Lim, Chino Hills, CA (US); and Timothy John Passaro, Vancouver, WA (US)
Assigned to Samsara Inc., San Francisco, CA (US)
Filed by Samsara Inc., San Francisco, CA (US)
Filed on Jan. 9, 2024, as Appl. No. 18/408,292.
Application 18/408,292 is a continuation of application No. 18/169,031, filed on Feb. 14, 2023, granted, now 12,137,143.
Application 18/169,031 is a continuation of application No. 17/301,658, filed on Apr. 9, 2021, granted, now 11,611,621.
Application 17/301,658 is a continuation of application No. 16/395,960, filed on Apr. 26, 2019, granted, now 10,999,374.
This patent is subject to a terminal disclaimer.
Int. Cl. H04L 67/12 (2022.01); G06V 20/40 (2022.01); G07C 5/00 (2006.01); H04L 41/069 (2022.01)
CPC H04L 67/12 (2013.01) [G06V 20/40 (2022.01); G07C 5/008 (2013.01); H04L 41/069 (2013.01); G06V 20/44 (2022.01)] 17 Claims
OG exemplary drawing
 
1. A system comprising:
at least one sensor device mounted at a vehicle to generate sensor data;
a memory; and
one or more processors configured to perform operations comprising:
applying a machine learned model to video data generated by the sensor device;
detecting a feature within the video data based on the machine learned model, the feature indicating an expected behavior of a driver of the vehicle and corresponding to an event type and a timestamp, the event type including a deviation action of the driver related to a traffic proceeding;
accessing a procedure associated with the event type, the procedure including an identification of a client device based on the event type; and
presenting a notification at the client device based on the procedure, the notification including a display of at least a portion of the video data that includes a scrub bar that comprises an event indicator that corresponds with the event type at a position along the scrub bar based on the timestamp indicating an occurrence of the deviation action from the expected behavior of the driver of the vehicle.