Deep Learning for Predictive Maintenance Scenario

By for December 19, 2017

Report Abuse
Deep learning is one of the most popular trends in machine learning. Deep learning is used in many fields and applications, including driverless cars, speech and image recognition, robotics, and finance. Deep learning is a set of algorithms that is inspired by the shape of the brain (biological neural networks), and machine learning. Cognitive scientists usually refer to deep learning as artificial neural networks (ANNs). Predictive maintenance is also popular. In predictive maintenance, many different techniques are designed to help determine the condition of equipment, and to predict when maintenance should be performed. Some common uses of predictive maintenance are failure prediction, failure diagnosis (root-cause analysis), failure detection, failure type classification, and recommendation of mitigation or maintenance actions after failure. In predictive maintenance scenarios, data is collected over time to monitor the state of equipment. The goal is to find patterns that can help predict and ultimately prevent failures. Using Long Short Term Memory (LSTM) networks is a deep learning method that is especially appealing in predictive maintenance. LSTM networks are good at learning from sequences. Time series data can be used to look back at longer periods of time to detect failure patterns. In this tutorial, we build an LSTM network to predict the remaining useful life of aircraft engines. The template uses simulated aircraft sensor values to predict when an aircraft engine will fail in the future. Using this prediction, maintenance can be planned in advance, to prevent failure.
The **detailed documentation** for this real world scenario includes a [step-by-step walkthrough on Microsoft Docs][1]. For code samples, click the "**View Project**" icon on the right to visit the [project GitHub repository][2]. Key components needed to run this scenario: - An [Azure account]( (free trials are available). - [Azure Machine Learning Workbench][3]. - Follow the [quick start installation guide][4] to install the program and create a workspace. - For model operationalization: Azure Machine Learning Operationalization, with a local deployment environment set up, and an [Azure Machine Learning Model Management account][5]. We suggest that you examine the data format and complete all three steps of the template before you replace the sample data with your own data. [1]: [2]: [3]: [4]: [5]: