What does observability mean?

A system is observable if you can determine the initial state x(0) from the output y(t) over a finite time interval. Mathematically, this requires the observability matrix O to have full column rank (rank = n).

What is the observability matrix?

The observability matrix is O = [C; CA; CA²; …; CAⁿ⁻¹], a (n·p)×n matrix stacked from n block rows. A system is observable if and only if O has rank n.

What if my system is not observable?

An unobservable system means some states cannot be reconstructed from the outputs. You may need to add sensors, redesign the C matrix, or restructure the system model.

What matrix sizes are supported?

The tool supports systems with 1 to 8 states and 1 to 8 outputs. For larger systems, use MATLAB (obsv), Python-control (control.obsv), or SciPy.

How is rank computed here?

The tool uses Gaussian elimination with partial pivoting on floating-point entries. A row is considered non-zero if its leading element exceeds a small tolerance (1e-9).

Observability Helper

Name: Observability Helper
Availability: InStock
Author: Nham Vu

Enter state-space matrices A and C to compute the observability matrix and determine if your LTI system is fully observable.

System Dimensions

States (n)

Outputs (p)

Matrix A (n × n state matrix) rows separated by semicolons

Example: 0 1; -2 -3

Matrix C (p × n output matrix)

Example: 1 0

Quick Examples

Enter matrices and click "Check Observability"

Summary

Enter state-space matrices A and C to compute the observability matrix and determine if your LTI system is fully observable.

How it works

Enter the number of states (n) and outputs (p) for your system.
Input the n×n state matrix A row by row, with values separated by spaces or commas.
Input the p×n output matrix C in the same format.
Click "Check Observability" to compute the observability matrix O = [C; CA; CA²; …; CAⁿ⁻¹].
The tool computes the rank of O using Gaussian elimination and reports whether rank = n (observable) or rank < n (not observable).
The full observability matrix and its rank are displayed so you can inspect each block row.

Use cases

Verify a control system design is observable before implementing an observer or Kalman filter.
Check observability after removing a sensor to assess whether state estimation is still feasible.
Validate textbook state-space examples and homework problems.
Debug system models where state estimates diverge unexpectedly.
Teach or learn observability theory with instant numerical feedback.
Quickly test small systems (up to 8 states) without MATLAB or Python.

Frequently Asked Questions

Last updated: 2026-06-10 · Reviewed by Nham Vu