US 12,450,905 B2
System and method for analyzing periodic task
Cagkan Ekici, Istanbul (TR); Yusuf Akgul, Kocaeli (TR); and Ozan Kerem Devamli, Istanbul (TR)
Assigned to Khenda, Inc.
Filed by Khenda, Inc., Lewes, DE (US)
Filed on Sep. 30, 2022, as Appl. No. 17/956,910.
Prior Publication US 2024/0112466 A1, Apr. 4, 2024
Int. Cl. G06V 20/40 (2022.01); G06V 10/82 (2022.01)
CPC G06V 20/44 (2022.01) [G06V 10/82 (2022.01); G06V 20/46 (2022.01)] 9 Claims
OG exemplary drawing
 
1. A periodic task analysis system, wherein the periodic task analysis system analyzes a task conducted periodically by an actor, comprising:
a data collecting subsystem, wherein the data collecting subsystem comprises at least one of an imaging device configured to image a state where a plurality of fundamental work operations are performed periodically, and at least one sensing means, wherein the at least one sensing means collects data from the state;
a data receiving subsystem, wherein the data receiving subsystem is operatively coupled to the data collecting subsystem, and the data receiving subsystem is configured to receive at least one of a video file and sensor data, wherein the video file is recorded by the imaging device, and the sensor data is recorded by the at least one sensing means;
a task learning subsystem, wherein the task learning subsystem is operatively coupled to the data receiving subsystem, and the task learning subsystem is configured to identify at least one sub-task from the at least one of the video file and the sensor data, wherein the at least one sub- task follows each other periodically in a loop; wherein the task learning subsystem is configured to generate labeled data by using at least one of frames of the video file and the sensor data, wherein the labeled data indicates a completion percentage of a related sub-task,
a task analyzing subsystem, wherein the task analyzing subsystem is operatively coupled to the data collecting subsystem and the task learning subsystem, and the task analyzing subsystem is configured to
receive a predetermined number of data list, wherein the predetermined number of data list is at least one of a predetermined number of frames of a live video streaming obtained by the imaging device and a predetermined number of sensor data of live streaming sensor data obtained by the at least one sensing means, to evaluate real time operations,
and the task analyzing subsystem is configured to generate a progress position, wherein the progress position indicates the related sub-task, wherein the progress position is generated as a regression output and the related sub-task is being performed by the actor in real time, and wherein the progress position indicates, during performance of the related sub-task, further the completion percentage of the related sub-task based on a continuous evaluation of the predetermined number of data list and the labeled data.