APMonitor Courses

APMonitor AI Engineering Assistant

Clara Sterling

A page-aware engineering assistant helping students, instructors, and engineers work through process dynamics and control, optimization, machine learning, and scientific computing—one clear explanation at a time.

How to ask Clara [email protected]
  • Evidence-led
  • Page-aware context
  • Math + code rendered
Clara Sterling, AI Engineering Assistant
Available through the Clara chat bubble on APMonitor course pages

Objective + purpose

Turn course material into working understanding.

Clara connects engineering fundamentals with the page you are studying. Her role is to explain concepts, walk through derivations, and help debug code—coaching you toward answers rather than replacing the learning.

Objective

Help learners master dynamics, control, and optimization concepts and build correct, well-reasoned engineering solutions.

Teaching purpose

Show the reasoning: derivations, tuning rules, and code walk-throughs with equations and examples rendered right in the chat.

Human purpose

Support students and instructors while respecting academic integrity—Clara coaches through graded work instead of doing it.

Technical profile

Skills across dynamics, data, and computation.

Clara's context spans the APMonitor courses—from physics-based modeling and controller tuning through dynamic optimization, machine learning, and the hands-on Temperature Control Lab.

Process dynamics + control

  • FOPDT modeling
  • Laplace transforms
  • Transfer functions
  • PID tuning
  • Cascade control
  • Feedforward
  • Stability analysis
  • TCLab exercises

Optimization + machine learning

  • Dynamic optimization
  • Model predictive control
  • Moving horizon estimation
  • Parameter estimation
  • Regression
  • Classification
  • Neural networks
  • Gekko

Scientific computing

  • Python
  • MATLAB
  • Jupyter notebooks
  • NumPy + SciPy
  • Differential equations
  • Numerical methods
  • Data visualization
  • Arduino + TCLab

Available in context

Ask from the page you are studying.

Look for Clara's portrait in the lower-right corner of APMonitor course pages. Open the chat and ask about a derivation, an exercise, a plot, or the code on the page. Clara combines her engineering context with the content of that page, renders equations and code in her answers, and remembers prior exchanges for follow-up questions.

How to chat with Clara

  1. Open the chatSelect Clara's portrait in the lower-right corner of the page.
  2. Ask about what you seeRefer to an equation, exercise, plot, code block, or concept on the current page.
  3. Read the streamed responseAnswers stream in with rendered math and formatted code. The Send button returns after the stream pauses.
  4. Ask a follow-upContinue naturally with questions such as “Why?”, “Can you derive that?”, or “Show it in Python.”

Good questions to ask

Understand the theoryAsk what an equation means, how a derivation works, or why a tuning rule takes the form it does.
Work through the labClarify TCLab setup, step tests, model fitting, or how to run the digital twin without hardware.
Debug the codePaste an error or describe unexpected results from the Python, MATLAB, or Gekko examples on the page.
Clara SterlingAI Engineering Assistant
How do I get a FOPDT model from my TCLab step test?
Fit three parameters from the step response—gain, time constant, and dead time: τp dy/dt = −y + Kp u(t − θp) Kp is Δy/Δu at steady state, θp is the delay before the response starts, and τp is the time to reach 63.2% of the total change after the delay.
What should I check if my fit looks off?
Confirm the test started at steady state, the heater step was large enough to clear sensor noise, and you waited long enough to see the new steady state. Then compare the model prediction against the data before using it for tuning.
Ask a follow-up…

Learning with integrity

Clear about reasoning. Honest about limits.

  • Shows derivations and reasoning instead of just final answers.
  • Coaches through graded work rather than completing it.
  • Never fabricates data, plots, or references.
  • Admits uncertainty and suggests a simulation or measurement to check.
  • Separates established theory from assumptions and approximations.
  • Supports instructors and the course AI policy.