Scribd
Upload
2 views

L-1 Introduction

ChE 208: Process Data Analytics & Monitoring focuses on leveraging process data to improve manufacturing practices through robust design and continuous learning. The course emphasizes the need for effective data analytics to transform raw data into actionable information, addressing challenges in measurement, analysis, and interpretation. Key objectives include troubleshooting, optimizing systems, and ensuring compliance through advanced data techniques.

Uploaded by

mohak.2025.2026
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
2 views

L-1 Introduction

ChE 208: Process Data Analytics & Monitoring focuses on leveraging process data to improve manufacturing practices through robust design and continuous learning. The course emphasizes the need for effective data analytics to transform raw data into actionable information, addressing challenges in measurement, analysis, and interpretation. Key objectives include troubleshooting, optimizing systems, and ensuring compliance through advanced data techniques.

Uploaded by

mohak.2025.2026
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

ChE 208: Process Data Analytics

& Monitoring
Introduction
Department of Chemical Engineering, IIT Indore
Principles of Good Manufacturing Practice
• Emphasis on robust process design that can handle uncertainty / variations.
▪ Based on best available systemic knowledge
▪ Approximate characterization of uncertainty

• Continuous learning leveraging information from measurements

• Measurement available at different points in a product life cycle.


▪ Product development at lab scale
▪ Product scale-up at pilot scale
▪ Product manufacturing at full plant scale

• Need of the hour: Improved manufacturing leveraging process data

Process data analytics tools help to exploit data content towards improved, safe and productive
practices
Department of Chemical Engineering, IIT Indore
What is process data analytics ?
• Analytics: Discovery and purposeful utilization / exploitation of meaningful
patterns and information in the data
• At the faster scale, process control strategies have been effectively
utilizing data for navigating the plant.
▪ Raw approach to utilize only measurements.
▪ Can we do more?
• New generation approaches for data analytics focus on
▪ Data →Information
▪ Information →Knowledge
▪ Knowledge →Wisdom
Department of Chemical Engineering, IIT Indore
The path from data to wisdom

Listen carefully to your data →There is an interesting and useful story to be revealed !!
Department of Chemical Engineering, IIT Indore
Where are we in the data analytics area?

▪ Societal systems in general and manufacturing units in particular are


“data-rich” but “information-poor”

▪ We are in the age of peta-bytes: kb, Mb,Gb, Tb and now Pb& Eb

▪ More importantly, nature of the data is very diverse –numbers, images,


acoustics …

▪ How do we generate meaningful interpretations from this data?

▪ How do we derive useful procedures to operate and run our plants better?

Department of Chemical Engineering, IIT Indore


The peta byte & exa byte size
▪ Walmart’s data storage runs into
several peta-bytes

▪ the total data uploaded to YouTube


each day is1.62 petabytes (PB)

▪ A typical chemical manufacturing


archive runs into exa-bytes

Department of Chemical Engineering, IIT Indore


The digital tsunami …..

Tapping into a data stream is like drinking water from a fire hose pipe ….
Department of Chemical Engineering, IIT Indore
Combustion: Illustrative example of sensor
diversity

Department of Chemical Engineering, IIT Indore


Main Message
▪ Need a paradigm shift in the way we operate our plants:
• Listen carefully to your plant
• Interpret the signals into meaningful information
• Build knowledge using the information
• Evolve safe, improved, optimal & compliant operating policies
▪ Data deluge: Just too much of data and no wisdom
• Need to rescue the data from the graveyard !! (archives)
▪ Do not let data overload lead to information overload
• Look for nuggets in the hay and do not make more hay
▪ More importantly, translate data into actionable information
Department of Chemical Engineering, IIT Indore
The 5 main uses of data in process engineering
systems
1. To learn more about a process
2. To troubleshoot a system
3. To make predictions from the data
4. To optimize a system proactively
5. To monitor a system from (realtime) data

Department of Chemical Engineering, IIT Indore


Typical objectives of process data analytics
▪ Reveal (lower) dimension latent relationships
Dimensionality reduction
▪ Discover patterns and clusters in the data
Establish local and global cause-effect relationships
▪ Build soft-sensors for inferencing of variable /metric
Sensor fusion based approaches
▪ Generate reduced variance estimates for closed loop control
Complementing available first principles information
▪ Generate predictive models for prognostics and diagnostics
Facilitate fault detection and diagnosis
▪ Continuous performance assessment
Grade performance and ensure compliance
Department of Chemical Engineering, IIT Indore
Basic challenges in Process data analytics
▪ Design problem
• What to measure, and how (frequency, location etc.)? –sensor location/ design
• How to transform and visualize the data? –outliers, nonlinear effects
• How to present the data in a form suitable for analysis –streaming / aggregated

▪ Analysis problem
• Deal with complexity in size, collinearity in the data
• Accommodate heterogeneity and nonlinearity
• Address uncertainty in data (low SNR, multiple memberships, )
• Complement modeling knowledge with appropriate relative importance

Department of Chemical Engineering, IIT Indore


About the Course
• How to improve processes using the large quantities of data that are
routinely collected from process systems ?
• Data Analytic techniques to quantify variability and then to compare
variability
• Understand how variation in one variable is related to another variable
• Deal with multiple variables and extract information from them

Department of Chemical Engineering, IIT Indore

You might also like