IIoT | 5 MIN READ

How Ideas2IT is Driving IIoT Enabled Predictive Maintenance

If you are a manufacturer or someone managing a large number of industrial equipments, machine maintenance is your one of your bigger challenges.

Most manufacturers today take either of the following approaches to solve maintenance issues: Calendar Based Maintenance or routine maintenance wherein equipments are periodically examined and problems are fixed based on a fixed schedule; and Reactive Maintenance wherein problems are fixed once equipments go down.

The latter approach leads to expensive machine downtime, failure to meet sales quota and production deadlines, and skyrocketing costs. For example, in a Fab, if one piece of semiconductor manufacturing machinery goes down for a few hours, wafer fabrication can come to a complete stop.   Mathematically equipment failure takes the “Bathtub Curve”, with frequent failures at the beginning of equipment life being reduced by regular maintenance processes, and ultimately leading to flattening of the curve. However, by the end of life, assets begin to face failure again. Hence, maintenance schedules should be adjusted over time to take into consideration changing failure rates. These schedules cannot be generated manually. This is where our predictive analytics platform comes in.

OVERVIEW

Our Industrial IoT platform, Factory.360, helps accelerate IT and OT integration by converging industrial assets with an infrastructure that supports edge computing, advanced stream analytics, and deep learning capabilities to provide predictive insights. It helps you monitor, assimilate, analyze, and draw inferences on information that is gathered from valuable assets. It also recommends maintenance activities for them. This allows you to take decisions that driver better business results providing benefits such as fewer breakdowns, increased efficiency, reduced maintenance costs, increased visibility of assets, and ability to provide proactive repair of plant equipment.

Our solution will allow you to

  • Predict the possible downtime of a monitored asset providing you an opportunity to fix it and avoid failure
  • Suggest maintenance schedule for an asset when its efficiency seems to be reducing
  • Search maintenance logs stored within the system to determine preferred repair procedures
  • Root cause asset failure allowing you to take corrective actions.

Our platform contains connectors that allows easy extraction of data from various sources such as ERP, Customer Relationship Management (CRM), Asset Management System (EAM), SCADA, PLC, shopfloor production and quality data, unstructured data as well as sensor (such as pressure, vibration, temperature, etc.) data and pools it into a data lake. This data lake contains data in a format which can be utilized by our data models. Our data analysts decide which models would fit the data and the questions that they would answer. Some of the models that are used are:

  • Regression models to predict remaining useful lifetime
  • Classification models to predict failure within a given time period
  • Flagging anomalous behavior
  • Survival models for prediction of probability of failure over time

Our Artificial Intelligence (AI) and Deep Learning (DL) models are employed to detect patterns in historical data and detect anomalies in them based on live data. Some of the key details that our platform can provide are mentioned below:

  • Models calculate asset health scores and predict remaining lifespan of an asset
  • Predict and pre-empt failure of assets and quality issues
  • Explore asset performance data to learn the cause of failure
  • Suggest optimized maintenance recommendations to operations
  • Provide real time, role based, interactive dashboards that track assets in real time and provide insights
  • Customize solutions for your specific use cases for maintenance and predictions

Enabling Data Sciences for the Enterprise

Ideas2IT takes a pragmatic approach to the Internet of Things (IoT) by building on the equipment and data you already have and leveraging your current investments in technology and data to enable action based on data analytics.

The team at Ideas2IT works closely with clients following the six below steps to plan implementation of our platform.

  • Establish use cases and business validations for predictive maintenance.
  • Identify and prioritize data sources for a few assets as a starting point
  • Collect a niche set of data from disparate sources and pool it into data lake
  • Determine where to run the analytics (edge vs distributed vs cloud analytics)
  • Combine and analyze data to provide precise insights
  • Operationalize and suggest actions to required parties.

Additionally, Factory.360’s modular structure allows flexible deployment of independent modules and easy integration with platforms that you already use. The platform can be installed both on premise and in cloud.

Predictive Maintenance Use Case

A typical electricity provider supplies electricity to millions of commercial and residential consumers and needs to ensure maximum availability and meet SLAs. They need to know how much energy would be required by various locations to ensure that they have enough energy to meet the city’s needs across the year. Our data analytics team worked on a predictive analytics platform for a leading electricity supplier providing them detailed insights and predictions of energy usage across various locations.

 

We collected reliable and fault tolerant data from various smart meters installed across the city. Based on the data, we provided the following insights

  • Displayed the grid of meters by district/region/state or as individual meters or another cluster
  • Calculated the consumption and revenue by geographic clusters allowing users to interactively slice and dice growth and profitable areas.
  • Estimated/forecasted the future consumption of meters or a cluster of meters in a given geographic region (district/city/street etc)
    • to inform the grid what the likely demand is going to be.
    • estimate the revenue from each geographic cluster for a future time period

 

 

 

Through the use of our Factory.360 platform, Ideas2IT can help enterprises and partners deliver their product at a reduced time and cost, enabling them to take their products to market much more quickly and increasing the viability of their Idea. Our core business team will be your product partners ensuring that you achieve your IoT needs.





Leave a Reply

Your email address will not be published. Required fields are marked *

SHARE linkedin twitter