Industrial IoT Predictive Maintenance Solution

© 2017 Cognizant

© 2017 Cognizant

Industrial IoT

Predictive Maintenance Solution

© 2014 Cognizant

Business Challenge

Our Solution

Solution Detail

Benefits

Agenda

© 2014 Cognizant

Business Challenge

Machine and Maintenance Cost

No automated Way to Predict

High Labor Cost

Lost Production Time

Loss in Revenue

Low Yield – Industrial Throughput

© 2014 Cognizant

Our Solution

Predictive Maintenance Solution–

• Real time connectivity with factory floor Sensors to gather real time machine data.

• Machine data is matched against pattern library to understand faulty pattern for a given time interval.

• Any match of faulty pattern in a given time interval, an action is initiated.• Email Notification

• Work Order Initiation in Maximo

• Web Service Initiation

• Sensor data is visualized on Mash zone NextGen.

© 2014 Cognizant

This solution is built using below Integration products and their respective components.

• Software AG – Apama (For Complex Event Processing)

• Software AG – Universal Messaging (To receive events from sensors and send to Apama)

• Software AG – Mashzone NextGen (For visualization of sensor data including fault patterns.)

© 2014 Cognizant

Edge Analytics - Solution Detail

PLC

Controls and monitor

Process Values

OPC UA

OSI PI

Edge Gateway

Edge Analytics

(Apama CEP)

Process Data

Univ

ers

alM

essa

gin

g

JMS

Mashzone

NextGen

Integration Server Data Lake

Ale

rts

Sensor

Graphs

Alarms

Pattern

Detection

Real Time Dashboard

Process Variables of Extruder:

• Extruder pressure

• Extruder Temperature

• Extruder Screw speed

• Product Flow

Process Variable of Coater

• Coater Speed

• Gap between rollers

• Gauge (thickness) of roll

(Output)

Process Data

© 2014 Cognizant

Solution Benefits

Reduction in

Equipment Cost

Reduction in Labor Cost

Reduction in

Production Cost

Increase in Revenue

Increase in Employee Efficiency

New Business Model

© 2014 Cognizant

Event Patterns

Variable Set Point Set Point Range Low Alarm(5% of SP) High Alarm(10% of SP) Static Alarm

Unwinder web tension 1.32 0.8 - 1.5 1.254 1.452 High/Low Unwinder web tension

Extruder Pressure (bar) 150 110 - 190 142.5 165 High/Low Extruder Pressure

Extruder Temperature (°C) 145 110 - 170 137.75 159.5 High/Low Extruder Temperature

Extruder Screw speed (RPM) 120 90-150 114 132 High/Low Extruder Screw speed

Extruder Product Flow (g/s) 2.22 2.0-2.5 2.109 2.442 High/Low Extruder Product Flow

chill roll water temperature 25 17° to 27°C 23.75 27.5 High/Low chill roll water temperature

Coater Speed (RPM) 110 90-150 104.5 121 High/Low Coater Speed

Coater Gap (mm) 0.8 0 .5 - 2 0.76 0.88 High/Low Coater Gap

Winder web tension 1.25 0.8 - 1.5 1.1875 1.375 High/Low Winder web tension

Output Thickness (mm) 0.8 0 .5 - 2 0.76 0.88 High/Low Output Thickness

Effect Pattern Detection for: Current Setup - Labelling Paramters Possible Cause Severity

Wrinkles on the Windup Roll

Cooling (variations)

Inadequate tension control at

the rewind

Coater chill roll water temperature -

Variations(High and Low alarms of the SP)

Winder web tension - low alarm of the SP

In 60 second time window

Water cooling passages

clogged

Idler rollers not in train High/Low

Fault Pattern

Threshold Breach

© 2014 Cognizant

Edge Analytics – Sensor Feeds

© 2014 Cognizant

Edge Analytics – Alarms