Day: July 7, 2018

Recurrent Neural Nets (RNN) detect features in sequential data (e.g. time-series data). Examples of applications which can be made using RNN’s are anomaly detection in time-series data, classification of ECG and EEG data, stock market prediction, speech recogniton, sentiment analysis, etc.

This is done by unrolling the data into N different copies of itself (if the data consists of N time-steps) .
In this way, the input data at the previous time steps  can be used when the data at timestep  is evaluated. If the data at the previous time steps is somehow correlated to the data at the current time step, these correlations are remembered and otherwise they are forgotten.

By unrolling the data, the weights of the Neural Network are shared across all of the time steps, and the RNN can generalize beyond the example seen at the current timestep, and beyond sequences seen in the training set.

This is a very short description of how an RNN works. For people who want to know more, here is some more reading material to get you up to speed. For now, what I would like you to remember is that Recurrent Neural Networks can learn whether there are temporal dependencies in the sequential data, and if there are, which dependencies / features can be used to classify the data. A RNN therefore is ideal for the classification of time-series, signals and text documents.

So, Lets start with implementing RNN’s in Tensorflow and using them to classify signals.

To see the full blog-post click here. https://goo.gl/EPSpwm #DataScience #Cloud

The manufacturing industry is overwhelmed with the advent of machines, technology, and the internet. The fourth industrial revolution is set to make manufacturing more digital, and only intelligent machines can make this happen. With the rise of machines, robotics, more automation, agile manufacturing, and increasing customization, manufacturing operations have become very complex. This demands better performance and reliability, especially for asset-heavy industries, where the reliability of industrial assets is mission-critical. At the same time, industry is experiencing revenue and profitability challenges that call for enhanced operational reliability and lower maintenance costs. However, reliability is a complex function, it involves monitoring, control, and improvements over the entire asset lifecycle – from acquisition, commissioning, operations, and maintenance, through disposal.

There is only one answer and that is asset performance visibility.

Management thinker Peter Drucker once said, “you can’t manage what you can’t measure.” This rings true for maintenance operations. You can’t improve reliability or performance unless you have complete visibility with respect to asset performance. In the past, industry professionals spoke about asset visibility more in terms of tracking the physical location of critical assets. But in the digital age, it is possible to monitor location, status, and performance in real time.

* The business benefits of the asset visibility program are not limited to productivity and safety; they are more about reliability and profitability. The benefits could be as high as a 7% improvement in return on assets (ROA) and 11 % improvement in overall equipment effectiveness (OEE). There are several key benefits you should consider when creating a viable business case for the asset visibility program:
* Improved reliability of maintenance processes based on operational insights
* Reduced overall maintenance cost thought-focused analysis
* Ability to identify root cause, leveraging diagnostic analytics capability
* Accurate predictions and timely prevention of asset failures
* Improved performance through KPI and best-practice benchmarking
* Building a foundation for Industry 4.0 by leveraging IoT-enabled solutions.

Asset visibility is not a new concept. Organizations have been investing in visibility solutions for over a decade. The challenge has always been to gain actionable insights. Thanks to the digital technologies available today, the Internet of Things, sensors, the cloud, analytics platforms, and mobile solutions make asset performance visibility an easy task. IIoT and sensors capture data in real time, analytics platforms with cloud computing analyze massive amounts of data in no time, and mobile solutions make insights available to people instantly. That’s the power of digital – connected operations with real-time collaboration.

Research suggests that only 20% asset-intensive companies have invested in asset visibility programs.  Gartner predicts that there will be 20.4 billion connected “things” worldwide by 2020. This would certainly mean more assets, equipment, machines, and devices to monitor and more data to analyze and visualize. Organizations will fail if they focus only on technology solution (tools) to perform analytics. The key to a successful maintenance visibility program is a structured approach to:

* Identify key performance indicators (KPIs)
* Define KPI calculation logic
* Identify correct data elements
* Streamline asset hierarchy
* Design dashboards and reporting structure
* Devise role-based access and continuous performance monitoring and improvement.

We can summarize with a quote from Elon Musk: “The future will be dominated by autonomous maintenance, machine capable of fixing itself.”  But will this be possible without asset performance visibility?

By:

Amit Supe (APICS CSCP, SMRP CMRP),

Abhisek Nanda (APICS BSCM, SMRP CMRP)

The original article was published here. https://goo.gl/zrfkb5 #DataScience #Cloud