US 12,254,017 B2
Real-time techniques for identifying a truth telling data population in a data stream
Rabindra Chakraborty, San Jose, CA (US)
Assigned to SENSLYTICS CORPORATION, Peachtree Corners, GA (US)
Filed by SENSLYTICS CORPORATION, Peachtree Corners, GA (US)
Filed on Apr. 25, 2023, as Appl. No. 18/139,165.
Prior Publication US 2024/0362240 A1, Oct. 31, 2024
Int. Cl. G06F 16/00 (2019.01); G06F 16/2458 (2019.01); G06F 16/248 (2019.01)
CPC G06F 16/2474 (2019.01) [G06F 16/248 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method of analyzing a trend in a data stream, comprising:
determining, based on a situational map of the data stream, an expected range of values for incoming data points in the data stream as the incoming data points are received, the expected range of values corresponding to an expectation radius from a trend line of the trend in the data stream;
receiving, in the incoming data points during a first analysis of the trend, a data point that is outside the expected range of values, the data point being stored for a second analysis of the trend that reevaluates the data point at a later time, the second analysis being performed after receiving more of the incoming data points;
determining whether the data point that is outside the expected range of values is an error originating from at least one of a sensor malfunction or a mechanical failure, the data point being excluded from the second analysis of the trend when the data point is determined to originate from the at least one of the sensor malfunction or the mechanical failure; and
executing the second analysis of the trend to reevaluate, at the later time, the data point that is outside the expected range of values, the second analysis indicating whether a fitting of the trend line associated with the first analysis of the trend is to be changed.