On-Sensor Anomaly Detection Using IMU PCA and Rolling Variance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional anomaly detection systems suffer from latency and resource constraints, making them unsuitable for real-time applications, especially in devices with limited memory, processing power, and computational capacity, such as vehicular technologies and industrial environments.
Innovation Solution
An on-sensor anomaly detection system using an internal measurement unit (IMU) circuit that accumulates sensor data, calculates rolling variance and principal components, and detects anomalies based on distance metrics to cluster centroids, signaling alerts in real-time without relying on cloud-based analytics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If cloud-based analytics are used for anomaly detection, then detection capability is improved, but latency and response time increase
Solution Approach 1:
The patent merges the anomaly detection functionality directly into the sensor device by integrating a processor with the sensor array. This combination allows the device to perform local anomaly detection using principal component analysis and chi-square tests, eliminating the need to transmit data to cloud-based analytics systems and thereby reducing latency and response time while maintaining detection capability.
2Measurement precision
If cloud-based analytics are used for anomaly detection, then detection capability is improved, but bandwidth consumption increases
Solution Approach 1:
The patent merges the anomaly detection functionality directly into the sensor device by integrating a processor with the sensor array. This combination allows the device to perform local anomaly detection using principal component analysis and chi-square tests, eliminating the need to transmit data to cloud-based analytics systems and thereby reducing bandwidth consumption while maintaining detection capability.
3Measurement precision
If traditional anomaly detection systems are used, then detection functionality is provided, but resource consumption exceeds available capacity in constrained devices
Solution Approach 1:
The patent changes the computational parameters by implementing simplified statistical tests (principal component analysis with chi-square tests) that are specifically designed to run on resource-constrained devices. The system monitors a limited number of parameters (principal components) rather than analyzing complete sensor datasets, thereby reducing memory, processing power, and computational capacity requirements while maintaining anomaly detection functionality.
Data Source
AI summary
According to an embodiment, a method to detect anomalies in a device is proposed. The method includes accumulating q samples of sensor data within a rolling window, each of the q samples of sensor data corresponding to a temporal characteristic of the device during its normal operation; calculating a rolling variance for the q samples of sensor data; extracting a first principal component of the q samples of sensor data; calculating a minimum distance and a mean distance to cluster centroids based on a previously collected first principal component during a training phase; detecting an anomaly within the device based on the rolling variance, the minimum distance to the cluster centroids, and the mean distance to the cluster centroids; and signaling an alert signal in response to detecting the anomaly.