Systems and methods for supervision of local analytics

Abstract

Systems and methods for dynamically optimizing models used for sensor data analytics. An action is taken based on an analytics determination by systematically varying parameters of the analytical model using actions taken based on the analytics to determine the relative frequencies of hits, misses, false alarms, and correct rejections for particular model parameters. The model parameters for localanalytics are selected based upon on signal detection theory analysis and the value or cost of each hit, miss, false alarm, or correct rejection.

Description

Background technique [0001] As the "Internet of Things" grows, more and more devices will have sensors and processing power embedded in them. While this can provide intelligent action (advanced criteria for alerts, indicators of wear and tear on parts in equipment, etc.), these embedded analytics are often slow to update, in part due to preprogramming models and basing analysis methods and parameters on precomputed models. These models are often based on first principles or "big data" driven aggregation of historical data; either of these approaches will contain imprecision due to assumptions and, in the case of aggregated data, third variable and directionality issues . These modeling approaches are also limited in their ability to identify errors and self-correct due to lack of experimental controls and use of aggregated passively collected data. Furthermore, errors in these models can become increasingly expensive or risky as they are applied to expensive capital equipmen...

AI Technical Summary

Problems solved by technology

However, the longer responders waited, the longer sites stopped operating, including leaving industrial sites vacant or residents barred from their homes; this was considered a "miss" in signal detection theory, failing to Recognize when action should be taken

Images