← Dynamic Systems Engineering Foundations · Semester 1, Week 4

State Estimation: Reasoning Under Uncertainty

Theme

Every real-world system contains hidden variables. Engineers must estimate the internal state of a system using incomplete, noisy, and sometimes contradictory observations. State estimation is the bridge between observation and intelligent decision-making.

Learning Objectives

By the end of this week, students should be able to:

Distinguish observable variables from hidden variables.
Explain why perfect information is rarely available.
Estimate system state from incomplete evidence.
Express confidence quantitatively or qualitatively.
Revise estimates as new evidence becomes available.

Core Vocabulary

State · State Variable · Observable Variable · Hidden Variable · Measurement · Sensor · Noise · Signal · Estimate · Uncertainty · Confidence · Update · Calibration

Fundamental Principles

Principle 1 — The complete state of a system is rarely observable.

Principle 2 — Measurements are imperfect representations of reality.

Principle 3 — Intelligence consists largely of estimating what cannot be directly observed.

Principle 4 — Confidence should increase or decrease as evidence changes.

Principle 5 — Every new observation is an opportunity to refine the system model.

Anchor Example — Thermostat

Observed: Current room temperature.

Hidden: Heat stored in walls. Heat escaping through insulation. Outdoor temperature changes.

Lesson: The thermostat regulates effectively despite not observing every relevant variable.

Anchor Example — Human Learning

Observed: Homework. Quiz scores. Participation.

Hidden: Conceptual understanding. Motivation. Fatigue. Long-term retention. Misconceptions.

Discussion: Why teachers estimate understanding rather than directly measuring it.

Anchor Example — AI Assistant

Observed: User prompt. Generated response.

Hidden: Internal token activations. Intermediate reasoning processes. Confidence calibration. Relevant latent knowledge.

Discussion: Why evaluating an AI requires observing behavior rather than assuming internal correctness.

Anchor Example — Traffic

Observed: Vehicle counts. Average speed.

Hidden: Driver intentions. Upcoming lane changes. Road hazards beyond view. Weather conditions farther ahead.

Task: Estimate traffic conditions using only partial observations.

Anchor Example — Small Business

Observed: Sales. Inventory. Revenue.

Hidden: Customer satisfaction. Employee morale. Brand reputation. Future demand.

Task: Identify which hidden variables most strongly influence long-term performance.

AI Laboratory

For each anchor system: list all observable variables, list all hidden variables, ask AI which hidden variables are most important, compare the AI's estimates with your own reasoning, and revise your model based on evidence.

Reflection Questions

Which variables am I assuming rather than observing?
What additional measurements would improve my estimate?
How uncertain is my current model?
How should my confidence change if new evidence appears?

Assignment

Choose a real-world system and create a two-column table: Observed Variables, Estimated Hidden Variables. For each hidden variable, explain why it matters, what evidence could improve the estimate, and how uncertainty in that variable affects predictions.

End-of-Week Competency

A successful student can distinguish measurements from underlying system state, estimate hidden variables from incomplete evidence, and continuously refine those estimates as new information becomes available. This ability prepares students for mathematical modeling, cybernetics, and Dynamic Systems Engineering in the following weeks.

Next: Week 5 — Mathematical Models: The Language of Dynamic Systems