Skip to main content
Skip to footer
Contact Us
We are taking you to another website now.
June 29, 2016

Predicting parts failure is crucial to the success of a product. Leading manufacturers would affirm that the Internet of Things (IoT), combined with intelligent analytics, generates critical insights into product performancereducing the significant expenses associated with equipment failure. At a time when revenue leakage poses a huge challenge, smart predictive maintenance is possibly the key to driving product success.

Gartner predicts that by 2020, the IoTs installed base will touch a whopping 26 billion units. In recent times, we have witnessed the proliferation of condition-based maintenance through embedded sensors for Boeing aircraft, or the use of telematics as a maintenance tool in fleets.

How is this changing the face of customer service?
The world of connected products is essentially an ecosystem that supports greater customization; the flipside to this is the demand for better, tailored customer service. Manufacturers need to offer a robust maintenance policy to their customers, while driving down associated costs.

The first step, then, is to embed RFID chips on machine parts, allowing maintenance personnel to easily scan and order replacements. Manufacturers can leverage this to collect data related to product failure, or wear and tear. Predicting parts failure boosts the overall equipment effectiveness (OEE), and ensures accelerated preemptive action. Most importantly, this results in fewer downtimes, enhanced productivity, and greater customer satisfaction.
The evolution of analytical models
The use of analytical models for predicting parts failure has seen some changes over time. Conventional analytical models make use of the classification of repairable and non-repairable parts to predict failure, based on the period of use. Models that use historical data to determine the need for predictive maintenance have been useful in the past when manufacturers did not have access to sensor data.

On the other hand, composite analytical models make use of sensor data to not only make predicting parts failures effective but also to improve prediction accuracy. These models use historical reliability data as well as advanced data mining models. Their adoption is growing with advances in technology.

The future: molding data into insights
We believe that with the use of the IoT and analytics, manufacturers can successfully clear several of the hurdles that prevented predictive maintenance. Long-term data on usage of products can be used to send alerts to users on warranty compliance issues. It can also help manufacturers make design improvements, and introduce product differentiation. Precise part failure predictions are the hallmark of good warranty management practices as well.

What sets apart a manufacturer with excellent part failure prediction capability from the rest? Our white paper Harnessing the Power of IoT in Manufacturing tries to answer this question. Leading manufacturers have been accurately predicting part failures, even in the absence of embedded sensors. Selecting the right analytical model can help manufacturers accurately predict part failuresand increase customer satisfaction.


Ravishankar Kandallu is a statistical modeler and data scientist within the Business Intelligence and Analytics Center of Excellence in the Manufacturing business unit at Tata Consultancy Services (TCS). He has worked on analytical modeling projects in the domains of warranty analytics, forecasting, and network optimization for high-tech, manufacturing, and consumer products industries. He has a doctorate in supply chain management from the Indian Institute of Technology (IIT) Madras.


Thank you for downloading

Your opinion counts! Let us know what you think by choosing one option below.