Kameshwar Poolla

Bio: Kameshwar Poolla is an academic researcher from University of California, Berkeley. The author has contributed to research in topics: Deep learning & Support vector machine. The author has an hindex of 2, co-authored 2 publications receiving 19 citations. Previous affiliations of Kameshwar Poolla include California Institute of Technology.

A One-Class Support Vector Machine Calibration Method for Time Series Change Point Detection

Abstract: Identifying the change point of a system’s health status is important. Indeed, a change point usually signifies an incipient fault under development. The One-Class Support Vector Machine (OC-SVM) is a popular machine learning model for anomaly detection that could be used for identifying change points; however, it is sometimes difficult to obtain a good OC-SVM model that can be used on sensor measurement time series to identify the change points in system health status. In this paper, we propose a novel approach for calibrating OC-SVM models. Our approach uses a heuristic search method to find a good set of input data and hyperparameters that yield a well-performing model. Our results on the C-MAPSS dataset demonstrate that OC-SVM can achieve satisfactory accuracy in detecting change point in time series with fewer training data, compared to state-of-the-art deep learning approaches. In our case study, the OC-SVM calibrated by the proposed model is shown to be useful especially in scenarios with limited amount of training data.

A One-Class Support Vector Machine Calibration Method for Time Series Change Point Detection

Abstract: It is important to identify the change point of a system's health status, which usually signifies an incipient fault under development. The One-Class Support Vector Machine (OC-SVM) is a popular machine learning model for anomaly detection and hence could be used for identifying change points; however, it is sometimes difficult to obtain a good OC-SVM model that can be used on sensor measurement time series to identify the change points in system health status. In this paper, we propose a novel approach for calibrating OC-SVM models. The approach uses a heuristic search method to find a good set of input data and hyperparameters that yield a well-performing model. Our results on the C-MAPSS dataset demonstrate that OC-SVM can also achieve satisfactory accuracy in detecting change point in time series with fewer training data, compared to state-of-the-art deep learning approaches. In our case study, the OC-SVM calibrated by the proposed model is shown to be useful especially in scenarios with limited amount of training data.

Predictive Maintenance on the Machining Process and Machine Tool

Abstract: This paper presents the process required to implement a data driven Predictive Maintenance (PdM) not only in the machine decision making, but also in data acquisition and processing. A short review of the different approaches and techniques in maintenance is given. The main contribution of this paper is a solution for the predictive maintenance problem in a real machining process. Several steps are needed to reach the solution, which are carefully explained. The obtained results show that the Preventive Maintenance (PM), which was carried out in a real machining process, could be changed into a PdM approach. A decision making application was developed to provide a visual analysis of the Remaining Useful Life (RUL) of the machining tool. This work is a proof of concept of the methodology presented in one process, but replicable for most of the process for serial productions of pieces.

An Encoder-Decoder Based Approach for Anomaly Detection with Application in Additive Manufacturing

Abstract: We present a novel unsupervised deep learning approach that utilizes an encoder-decoder architecture for detecting anomalies in sequential sensor data collected during industrial manufacturing. Our approach is designed to not only detect whether there exists an anomaly at a given time step, but also to predict what will happen next in the (sequential) process. We demonstrate our approach on a dataset collected from a real-world Additive Manufacturing (AM) testbed. The dataset contains infrared (IR) images collected under both normal conditions and synthetic anomalies. We show that our encoder-decoder model is able to identify the injected anomalies in a modern AM manufacturing process in an unsupervised fashion. In addition, our approach also gives hints about the temperature non-uniformity of the testbed during manufacturing, which was not previously known prior to the experiment.

Change Point Enhanced Anomaly Detection for IoT Time Series Data

Abstract: Due to the exponential growth of the Internet of Things networks and the massive amount of time series data collected from these networks, it is essential to apply efficient methods for Big Data analysis in order to extract meaningful information and statistics. Anomaly detection is an important part of time series analysis, improving the quality of further analysis, such as prediction and forecasting. Thus, detecting sudden change points with normal behavior and using them to discriminate between abnormal behavior, i.e., outliers, is a crucial step used to minimize the false positive rate and to build accurate machine learning models for prediction and forecasting. In this paper, we propose a rule-based decision system that enhances anomaly detection in multivariate time series using change point detection. Our architecture uses a pipeline that automatically manages to detect real anomalies and remove the false positives introduced by change points. We employ both traditional and deep learning unsupervised algorithms, in total, five anomaly detection and five change point detection algorithms. Additionally, we propose a new confidence metric based on the support for a time series point to be an anomaly and the support for the same point to be a change point. In our experiments, we use a large real-world dataset containing multivariate time series about water consumption collected from smart meters. As an evaluation metric, we use Mean Absolute Error (MAE). The low MAE values show that the algorithms accurately determine anomalies and change points. The experimental results strengthen our assumption that anomaly detection can be improved by determining and removing change points as well as validates the correctness of our proposed rules in real-world scenarios. Furthermore, the proposed rule-based decision support systems enable users to make informed decisions regarding the status of the water distribution network and perform effectively predictive and proactive maintenance.

An Applicable Predictive Maintenance Framework for the Absence of Run-to-Failure Data

Abstract: As technology advances, the equipment becomes more complicated, and the importance of the Prognostics and Health Management (PHM) to monitor the condition of the equipment has risen. In recent years, various methodologies have emerged. With the development of computing technology, methodologies using machine learning and deep learning are gaining attention, in particular. As these algorithms become more advanced, the performance of detecting anomalies and predicting failures has improved dramatically. However, most of the studies are cases that depend on simulation data or assumed abnormal conditions. In addition, regardless of the existence of run-to-failure data, the methodologies are difficult to apply to the industrial site directly. To solve this problem, we propose a Predictive Maintenance (PdM) framework based on unsupervised learning in this paper, which can be applied directly in the industrial field regardless of run-to-failure data. The proposed framework consists of data acquisition, preprocessing data, constructing a Health Index, and predicting the remaining useful life. We propose a framework that can create and monitor models even when there are no accumulated run-to-failure data. The proposed framework was conducted in two different real-life cases, and the usefulness and applicability of the proposed methodology were verified.

CNN and LSTM based Encoder-Decoder for Anomaly Detection in Multivariate Time Series

Abstract: The purpose of anomaly detection is to detect data that deviates from the expected, and is widely used in intrusion detection, data preprocessing and so on.For data anomaly detection, we propose a data anomaly detection algorithm based on convolutional neural network and Encoder-Decoder architecture CNN-LSTMED (Convolutional Neural Networks Long Short-Term Encoder-Decoder).First,we use the convolutional neural network to encode the time series data to obtain the encoded sequence,and use the features extracted from the sequence as the input of the nonlinear model long short-term memory network LSTM (Long Short-Term Memory) to decode and output the decoded sequence. Finally, the reconstruction error is calculated and the threshold is set to determine the abnormal point. Through experimental comparison with GRUED (Gated Recurrent Neural Encoder-Decoder), LSTMED (Long Short-Term Memory Encoder-Decoder) ,and other algorithms on the KDD99 data set and credit card fraud data set,it turns out that our algorithm has strong robustness and accuracy .