🎓 ChemE × Data Engineering Course - START HERE

A self-paced course teaching Chemical Engineers data engineering through process control

---

What's Inside

16 Lessons · 35 YAML Examples · ~20-25 Hours

Every lesson includes: - Theory recap from Seborg textbook - Runnable YAML examples (use odibi run to execute) - Data engineering concepts explained - Validation patterns and quarantine - Exercises with solution hints - Real plant operation connections - Progressive difficulty from beginner to expert

Part I time: ~5 hours (beginner-friendly)

---

🚀 Quick Start (5 Minutes)

Step 1: Install Odibi

pip install odibi
odibi --version

Step 2: Run Your First Example

# Navigate to course examples
cd examples/cheme_course

# Run first simulation
odibi run L00_setup/tank_data.yaml

# View the generated data
python -c "import pandas as pd; print(pd.read_parquet('data/tank_data.parquet').head())"

Step 3: Start Learning Open L00: Setup and follow along.

You're learning! 🎉

---

📚 Your Learning Path

Part I: Foundations

L00: Setup & Basics (45 min)
🎯 Install Odibi, run first pipeline, understand data formats
📂 Examples: tank_data.yaml, tank_data_parquet.yaml, tank_realistic.yaml, multi_entity.yaml

L01: CV/MV/DV and Time Series (45 min)
🎯 Controlled/Manipulated/Disturbance variables
📂 Examples: mixing_tank_ph.yaml, heat_exchanger.yaml

L02: Degrees of Freedom + Balances (60 min)
🎯 DoF analysis, mass/energy balances
📂 Examples: tank_mass_balance.yaml, cstr_energy_balance.yaml

L03: First-Order Dynamics (45 min)
🎯 prev() and ema() stateful functions
📂 Examples: tank_temperature_fo.yaml, ema_filtering.yaml

L04: FOPTD Transfer Functions (45 min)
🎯 First-Order Plus Time Delay modeling
📂 Examples: tank_foptd.yaml, heat_exchanger_foptd.yaml, foptd_parameterized.yaml

L05: Second-Order Systems (60 min)
🎯 Damping ratio, overshoot, settling time
📂 Examples: pressure_underdamped.yaml, valve_actuator.yaml, damping_comparison.yaml

---

Part II: Feedback Control & System ID

L06: PID Basics (60 min)
🎯 P/I/D actions, pid() function, anti-windup
📂 Examples: tank_pi.yaml

L07: PID Tuning Methods (60 min)
🎯 Ziegler-Nichols, Cohen-Coon tuning
📂 Examples: ziegler_nichols.yaml, cohen_coon.yaml

L08: Disturbance Rejection (60 min)
🎯 Load rejection, feedwater disturbances
📂 Examples: load_disturbance.yaml, feedwater_disturbance.yaml

L09: System Identification (90 min)
🎯 Step test, pulse test for parameter estimation
📂 Examples: step_response.yaml, pulse_test.yaml

---

Part III: Advanced Control Strategies

L10: Interacting Control Loops (60 min)
🎯 MIMO systems, loop interactions
📂 Examples: dual_temperature.yaml, pressure_flow.yaml

L11: Cascade Control (60 min)
🎯 Primary/secondary loops, fast inner loops
📂 Examples: temperature_cascade.yaml, level_flow_cascade.yaml

L12: Feedforward Control (60 min)
🎯 Anticipatory control, ratio control
📂 Examples: simple_feedforward.yaml, ratio_control.yaml

L13: Nonlinear Systems (60 min)
🎯 Valve characteristics, pH neutralization
📂 Examples: valve_nonlinearity.yaml, ph_neutralization.yaml

L14: Model Predictive Control Intro (90 min)
🎯 Prediction, optimization, constraints
📂 Examples: mpc_basics.yaml, constrained_control.yaml

---

Capstone: Real-World Digital Twin

L15: CSTR Digital Twin (2-3 hours)
🎯 Complete reactor model with mass, energy, kinetics
📂 Examples: cstr_full_model.yaml, optimization.yaml

---

💪 What You'll Learn

Data Engineering Skills

• Generate realistic plant time-series data
• Use CSV, Parquet, and Delta Lake formats
• Validate data quality with range checks
• Implement quarantine patterns
• Build reproducible simulations
• Handle multi-entity pipelines

Process Control Skills

• Map process control to data schemas (CV/MV/DV)
• Implement mass and energy balances
• Model first and second-order dynamics
• Design and tune PID controllers
• Perform system identification
• Build cascade and feedforward controllers
• Handle nonlinear processes
• Apply basic model predictive control

Chemical Engineering

• CSTR modeling (kinetics, thermodynamics)
• Heat exchanger dynamics
• Tank level control
• pH neutralization
• Operating point optimization

Career Skills

• Portfolio of 35+ working examples
• Production-ready data pipelines
• Bridge between ChemE and data roles

---

🎯 What You'll Be Able to Do

After completing this course, you can:

• Build simulations generating millions of rows of realistic process data
• Implement PID controllers from Seborg textbook
• Tune controllers using industry methods (Z-N, Cohen-Coon)
• Model complex systems (CSTR, heat exchangers, cascades)
• Explain to data engineers: "This is a controlled variable"
• Explain to process engineers: "This is a Parquet file"
• Build production-ready data pipelines
• Create digital twins of chemical processes

---

📊 Course Statistics

Metric	Count
Total Lessons	16 (L00-L15)
YAML Examples	35
Course Time	20-25 hours
Exercises	40+ hands-on problems
Seborg Coverage	Chapters 1-20

---

🛠️ Course Structure

Where Everything Lives

YAML Examples:

examples/cheme_course/
├── L00_setup/           # 4 examples
├── L01_cv_mv_dv/        # 2 examples
├── L02_dof_balances/    # 2 examples
├── L03_first_order/     # 2 examples
├── L04_foptd/           # 3 examples
├── L05_second_order/    # 3 examples
├── L06_pid_basics/      # 1 example
├── L07_tuning/          # 2 examples
├── L08_disturbances/    # 2 examples
├── L09_system_id/       # 2 examples
├── L10_interacting_loops/ # 2 examples
├── L11_cascade/         # 2 examples
├── L12_feedforward/     # 2 examples
├── L13_nonlinearity/    # 2 examples
├── L14_mpc_lite/        # 2 examples
├── L15_cstr_digital_twin/ # 2 examples
└── README.md            # Quick reference

Lesson Documentation:

docs/learning/cheme_data_course/
├── START_HERE.md        # This file
├── index.md             # Course overview
├── lessons/
│   ├── L00_setup.md through L15_cstr_digital_twin.md
│   └── (16 lesson files with theory + exercises)
└── solutions/
    └── index.md         # Solutions hub

---

📖 Key Concepts & Patterns

Row Number Counter (Essential Pattern)

# Use this for step changes and time-dependent logic
- name: row_num
  data_type: int
  generator:
    type: derived
    expression: "prev('row_num', -1) + 1"

# Then create step changes:
- name: setpoint
  data_type: float
  generator:
    type: derived
    expression: "50.0 if row_num < 100 else 60.0"

PID Controller

- name: controller_output
  data_type: float
  generator:
    type: derived
    expression: >
      pid(
        process_variable,
        setpoint,
        Kp,    # Proportional gain
        Ki,    # Integral gain
        Kd,    # Derivative gain
        dt,    # Sample time (seconds)
        min,   # Output minimum
        max,   # Output maximum
        true   # Anti-windup enabled
      )

EMA Smoothing

- name: smoothed_value
  data_type: float
  generator:
    type: derived
    expression: "ema('raw_value', alpha, default)"

---

🎓 Teaching Philosophy

1. Hands-On First

Don't just read - run code, generate data, see results immediately.

2. ChemE Problems → Data Skills

Learn Parquet by simulating tanks. Learn validation by modeling reactors.

3. Incremental Complexity

L00: Simple CSV files → L15: Production digital twin pipelines

4. Real-World Focus

Every lesson connects to actual plant operations.

5. Portfolio Building

By the end, you have 35+ working examples for LinkedIn/interviews.

---

🤔 FAQ

Q: Do I need the Seborg textbook?
A: No! Lessons recap key concepts. But it helps for deeper theory.

Q: Can I skip lessons?
A: Start with L00-L03 to learn basics. Then pick topics you need.

Q: How long does the full course take?
A: 20-25 hours total. Part I (L00-L05) takes ~5 hours.

Q: Can I use this to teach others?
A: Absolutely! Share, improve, contribute back.

Q: What if I get stuck?
A: Check solutions, consult docs, or ask questions.

Q: Can I run these on Databricks?
A: Yes! All examples work on local Pandas or Databricks Spark.

---

🔗 Additional Resources

Course Materials: - Course Overview - Philosophy and structure - Seborg Textbook Mapping - All chapters mapped to Odibi - Process Simulation Guide - Deep dive on stateful functions - Solutions Index - Exercise solutions

Framework Guides: - Chemical Engineering Simulation - Thermodynamics Transformers - Unit Conversion - Custom Functions Reference

---

👉 Get Started Now

Recommended Path

1. Read Course Overview (10 min)
2. Install Odibi (5 min)
3. Start L00: Setup (45 min)
4. Progress through L01-L05 at your own pace (4 hours)
5. Advance to Part II (L06-L09) for control topics (4 hours)
6. Master Part III (L10-L14) for advanced strategies (5 hours)
7. Build L15 digital twin capstone project (2-3 hours)

Alternative Paths

Path A - Just the Basics (5 hours): L00 → L01 → L02 → L03 → L04 → L05

Path B - PID Focus (8 hours): L00 → L01 → L03 → L06 → L07 → L08

Path C - Advanced Only (6 hours): L00 (setup) → L10 → L11 → L12 → L13 → L14

Path D - Digital Twin Sprint (4 hours): L00 → L02 → L15

---

👉 Start with L00: Setup & Basics

---

Built with ❤️ for Chemical Engineers learning Data Engineering
Part of the Odibi Framework - Explicit over implicit, Stories over magic