A continuous-time linear system is asymptotically stable when the roots of its characteristic polynomial lie in the open left half of the complex plane. Directly calculating every root can hide the structural question: how many roots have positive real part? The Routh–Hurwitz criterion answers from the coefficients. A Routh array is built row by row, and sign changes down its first column count roots in the open right half-plane, subject to careful treatment of zero cases. Mathematics matters because polynomial algebra, complex roots, determinants, signs and limiting arguments turn a difficult root-location problem into an auditable table. The criterion says something precise about a declared linear model; it is not a guarantee that an actual machine is safe under every disturbance.
The practical search question is not simply “what is this named method?” but “which mathematical structure makes the method work, what does it guarantee, and what evidence would show that it is a poor choice?” This article keeps those questions together so students can connect mathematics education with computing, engineering, data reasoning and careful everyday decisions without turning one technique into a career promise.
Quick navigation: Why it matters · Mechanism · Worked example · Deeper ideas · Investigations · Limits · Learning route · Parent guidance · FAQs · Sources
Why this mathematics matters
Mathematics is important here because it turns an apparently visual or computational trick into a chain of reasons. The representation tells us what information is kept. The equation or invariant explains each legal step. The complexity analysis describes the work. The boundary conditions say when a conclusion should not be trusted. That combination is a practical form of problem-solving literacy.
A strong learner should be able to move in both directions: from a real task to a mathematical model, and from a formal guarantee back to its real meaning. Naming an algorithm is not enough. The useful achievement is to explain why the next operation is valid, check a small instance independently, and notice which assumptions the implementation has added.
Source: MIT OpenCourseWare lecture on system stability and the Routh–Hurwitz criterion.
Primary or implementation reference: Hurwitz’s 1895 paper on roots in the negative half-plane.
How the mechanism works
Write the real characteristic polynomial in descending powers of s and place alternating coefficients across the first two rows of a Routh table. Each later entry is formed from entries in the two rows above by a signed cross-product divided by the leading entry of the row immediately above. Continue until the constant row. When no special case occurs, count sign changes in the first column after normalising only by positive factors. That count equals the number of roots in the open right half-plane. No sign change and no imaginary-axis root gives the usual stability conclusion. A zero leading entry is replaced by a positive symbolic epsilon and interpreted by a limit; an entire zero row signals a symmetric root pattern and requires an auxiliary polynomial.
For p(s)=a_n s^n+…+a_1 s+a_0 with a_n>0, the Routh table encodes the Hurwitz determinants in a recursive form. The theorem relates positivity conditions on those determinants to every root having negative real part. The first-column sign-change rule is stronger than a simple coefficient check: positive coefficients are necessary for many stable polynomials written with positive leading coefficient, but they are not sufficient in general. Rows of zeros occur when an even or odd auxiliary polynomial has roots symmetric about the origin, including possible imaginary-axis pairs. Because roots on the imaginary axis are not in the open left half-plane, marginal cases must be reported separately rather than called asymptotically stable.
Keep a four-column notebook labelled object, rule, guarantee and caveat. In the first column name the inputs and stored state. In the second write one update precisely. In the third state what remains true after the update. In the fourth record a tie, approximation, domain restriction or hardware assumption. This routine prevents symbols from drifting away from their meaning.
Worked example and core calculations
Test p(s)=s^3+2s^2+3s+4. The first two Routh rows are s^3: [1,3] and s^2: [2,4]. The first entry of the s^1 row is (2×3−1×4)/2=1, and its next entry is zero. The s^0 row is [4]. The first column is 1,2,1,4: all positive, so there are no sign changes and therefore no roots in the open right half-plane. No special zero row appears, so the criterion identifies an asymptotically stable polynomial. A numerical root calculation is useful as a check, but it is not the proof used by the table.
How to audit the calculation
- Recompute every intermediate value directly from the definition.
- Keep labels, indices, signs, units and normalisation visible.
- Test a case with a known exact answer.
- State tie rules, random seeds, tolerances and boundary conventions.
- Compare with a slower transparent baseline or trusted implementation.
- Separate agreement of numbers from suitability of the real-world model.
Small examples are not childish; they are diagnostic instruments. A five-line trace can reveal a reversed inequality, a missing scale factor or an inconsistent label convention before a large dataset hides the same error behind plausible output.
Four deeper ideas
Stability can be counted without explicit roots
The array extracts half-plane information from coefficients, avoiding fragile symbolic formulas for high-degree roots.
To transfer this idea, ask what would remain true if labels, scale or input order changed. Then identify the step whose justification would fail if the stated assumption were removed. This turns the paragraph into a reusable reasoning pattern rather than a fact to memorise.
Signs encode a topological crossing count
A change of sign in the first column corresponds to a root count in the right half-plane, not merely an attractive table pattern.
To transfer this idea, ask what would remain true if labels, scale or input order changed. Then identify the step whose justification would fail if the stated assumption were removed. This turns the paragraph into a reusable reasoning pattern rather than a fact to memorise.
Did you know? Zero cases carry information
A zero first element and an entire zero row mean different things and demand different mathematical repairs.
To transfer this idea, ask what would remain true if labels, scale or input order changed. Then identify the step whose justification would fail if the stated assumption were removed. This turns the paragraph into a reusable reasoning pattern rather than a fact to memorise.
A criterion belongs to a model
The result concerns the characteristic polynomial of a linear time-invariant model; delays, nonlinearities, uncertainty and saturation need further analysis.
To transfer this idea, ask what would remain true if labels, scale or input order changed. Then identify the step whose justification would fail if the stated assumption were removed. This turns the paragraph into a reusable reasoning pattern rather than a fact to memorise.
Twelve practical investigations
1. Stable cubic
Use s^3+2s^2+3s+4. Build every row.
What the mathematics reveals: No first-column sign change means no right-half-plane roots. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Check that no special zero case occurs. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
2. One unstable root
Choose a polynomial with one positive real root. Compare factorisation and table.
What the mathematics reveals: Sign changes count unstable roots. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Do not infer exact root values. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
3. Two unstable roots
Use a conjugate pair with positive real part. Expand then build the array.
What the mathematics reveals: Complex pairs contribute two roots. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Keep real coefficients. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
4. Zero leading entry
Construct a table with a zero first entry. Introduce epsilon and take the positive limit.
What the mathematics reveals: The limiting signs preserve the count. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Never divide by zero. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
5. Entire zero row
Use a polynomial with symmetric factors. Form the auxiliary polynomial.
What the mathematics reveals: Symmetry creates the special row. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Differentiate the auxiliary polynomial correctly. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
6. Imaginary-axis pair
Test a factor s^2+w^2. Identify marginality.
What the mathematics reveals: Boundary roots prevent asymptotic stability. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Do not label the system robust. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
7. Parameter range
Put a gain K in the coefficients. Derive first-column inequalities.
What the mathematics reveals: The table can delimit stable gains. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Preserve strict inequalities. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
8. Coefficient scaling
Multiply the polynomial by a positive constant. Compare the table signs.
What the mathematics reveals: Root locations do not change. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Negative scaling flips signs unless normalised. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
9. Numerical roots
Calculate roots with trusted software. Compare right-half-plane counts.
What the mathematics reveals: Independent methods validate the trace. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Root rounding near zero can mislead. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
10. Discrete-time contrast
Map a pole inside the unit circle. Compare the geometry.
What the mathematics reveals: Left-half-plane rules do not transfer unchanged. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Use the appropriate theorem. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
11. Model uncertainty
Vary one coefficient within a range. Rebuild boundary conditions.
What the mathematics reveals: Nominal stability is not robust stability. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: State the uncertainty set. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
12. Control design
Relate K to a feedback denominator. Test candidate gains.
What the mathematics reveals: Algebra guides design choices. Evidence to collect: write the input, predict the next state, carry out the calculation, and compare the result with an independently produced reference. Explain the exact invariant, equation, inequality or counting rule that supports the conclusion. Then vary one assumption and record whether the result changes smoothly, discontinuously or not at all.
Limit to keep visible: Also assess performance and saturation. A useful report places this boundary beside the result rather than hiding it in a final disclaimer.
Limits and misconceptions
The standard criterion assumes a real polynomial and addresses continuous-time root location relative to the imaginary axis. Discrete-time systems require a unit-circle criterion such as Jury or Schur–Cohn. Coefficient uncertainty can move roots even when the nominal table is stable. A zero row or zero first element must not be handled by arbitrary rounding. Stability does not imply acceptable overshoot, settling time, robustness or actuator effort. A model may omit friction changes, delays, nonlinear saturation or sensor faults. Numerical implementations need scale-aware arithmetic because extremely different coefficient magnitudes can make table entries ill-conditioned.
Five misconceptions to challenge
- Positive coefficients prove stability: They are not sufficient for a general polynomial.
- No sign change means every real root is negative: Roots may be complex; the theorem concerns their real parts.
- An imaginary-axis root is asymptotically stable: It is a marginal boundary case, not open-left-half-plane stability.
- A zero row may be ignored: It signals a special symmetric root structure and needs an auxiliary polynomial.
- Stable mathematics guarantees a safe product: Physical validation, robustness and operating constraints remain separate.
Another broad misconception is that advanced mathematics automatically creates intelligence, admission, income or employment. It does not. Deliberate study can strengthen modelling, abstraction, calculation and explanation, while real pathways also depend on interests, communication, domain knowledge, opportunities and circumstances outside one topic.
A learning route for students
Move from a visible trace to a symbolic rule, then to code and critique. For every step below, prepare three pieces of evidence: a correct hand example, a boundary or failure case, and a short explanation using the relevant invariant. Do not advance merely because software returned a number.
Step 1: Form a characteristic polynomial
Begin with the smallest non-trivial instance that lets you form a characteristic polynomial. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 2: Place coefficients in a routh table
Begin with the smallest non-trivial instance that lets you place coefficients in a Routh table. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 3: Calculate recursive table entries
Begin with the smallest non-trivial instance that lets you calculate recursive table entries. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 4: Count first-column sign changes
Begin with the smallest non-trivial instance that lets you count first-column sign changes. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 5: Handle a zero leading entry
Begin with the smallest non-trivial instance that lets you handle a zero leading entry. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 6: Handle an entire zero row
Begin with the smallest non-trivial instance that lets you handle an entire zero row. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 7: Derive parameter inequalities
Begin with the smallest non-trivial instance that lets you derive parameter inequalities. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 8: Distinguish marginal and asymptotic stability
Begin with the smallest non-trivial instance that lets you distinguish marginal and asymptotic stability. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 9: Check roots numerically
Begin with the smallest non-trivial instance that lets you check roots numerically. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
Step 10: Separate model stability from physical safety
Begin with the smallest non-trivial instance that lets you separate model stability from physical safety. Predict the result before calculating. Write each state change clearly enough that another student can reproduce it. Compare with a trusted implementation only after the hand trace is complete.
Next, change one feature—size, ordering, sign, tolerance, labelling or data distribution—and explain which part of the reasoning changes. Finish by teaching the step without code and by naming one input for which the method is unhelpful. This predict–calculate–vary–explain cycle is a strong test of transferable understanding.
A four-week practice plan
- Week 1 — vocabulary and diagrams: define each object, redraw the mechanism and reproduce the worked example.
- Week 2 — equations and manufactured tests: derive the key relation and test inputs with known answers.
- Week 3 — implementation and evidence: use a trusted library, record parameters and compare with a transparent baseline.
- Week 4 — transfer and critique: apply the method to a new context, report limits and identify when a different method is better.
Guidance for parents and educators
Ask “what stays true?” and “what could make this conclusion fail?” before asking for speed. Invite the student to draw the input and output, estimate the result and locate the first step that disagrees with a reference. These prompts reveal conceptual understanding more clearly than memorised terminology.
Advanced enrichment should sit beside secure school foundations in algebra, functions, geometry, statistics and computational thinking. It need not accelerate a student into a fixed career identity. A healthy goal is curiosity with discipline: the learner becomes comfortable reading unfamiliar notation, checking an example, revising a model and communicating uncertainty.
Evidence prompts for a useful conversation
- Which quantity is observed and which is inferred?
- Which step saves work, and why is it allowed?
- Is the claim exact, approximate, probabilistic or empirical?
- What units and conventions are in use?
- Which input makes the method struggle?
- What simple baseline could check the answer?
- What would count as evidence that the parameters are poor?
Mini-project assessment
A strong mini-project contains the original question, a small reproducible dataset, a hand-worked example, code or a spreadsheet trace, at least one boundary test, a comparison with a baseline, and a paragraph on limits. Assess the reasoning trail as well as the final output. Repairing a model after a surprising result is valuable mathematical progress.
Six evidence checks for a finished project
Check 1: Form a characteristic polynomial
Ask the student to demonstrate how to form a characteristic polynomial without relying on a hidden software step. The evidence should include a labelled input, the exact rule used, one intermediate state and a result that can be checked independently. If the answer changes after labels or order change, the student should say whether that reflects the mathematics or only a convention.
Then request one deliberately awkward example. The learner should predict the difficulty before running the method, explain what the output does and does not establish, and name a reasonable alternative or safeguard. This check joins correctness with judgement: a technically valid calculation is useful only when its assumptions and interpretation remain visible.
Check 2: Place coefficients in a routh table
Ask the student to demonstrate how to place coefficients in a Routh table without relying on a hidden software step. The evidence should include a labelled input, the exact rule used, one intermediate state and a result that can be checked independently. If the answer changes after labels or order change, the student should say whether that reflects the mathematics or only a convention.
Then request one deliberately awkward example. The learner should predict the difficulty before running the method, explain what the output does and does not establish, and name a reasonable alternative or safeguard. This check joins correctness with judgement: a technically valid calculation is useful only when its assumptions and interpretation remain visible.
Check 3: Calculate recursive table entries
Ask the student to demonstrate how to calculate recursive table entries without relying on a hidden software step. The evidence should include a labelled input, the exact rule used, one intermediate state and a result that can be checked independently. If the answer changes after labels or order change, the student should say whether that reflects the mathematics or only a convention.
Then request one deliberately awkward example. The learner should predict the difficulty before running the method, explain what the output does and does not establish, and name a reasonable alternative or safeguard. This check joins correctness with judgement: a technically valid calculation is useful only when its assumptions and interpretation remain visible.
Check 4: Count first-column sign changes
Ask the student to demonstrate how to count first-column sign changes without relying on a hidden software step. The evidence should include a labelled input, the exact rule used, one intermediate state and a result that can be checked independently. If the answer changes after labels or order change, the student should say whether that reflects the mathematics or only a convention.
Then request one deliberately awkward example. The learner should predict the difficulty before running the method, explain what the output does and does not establish, and name a reasonable alternative or safeguard. This check joins correctness with judgement: a technically valid calculation is useful only when its assumptions and interpretation remain visible.
Check 5: Handle a zero leading entry
Ask the student to demonstrate how to handle a zero leading entry without relying on a hidden software step. The evidence should include a labelled input, the exact rule used, one intermediate state and a result that can be checked independently. If the answer changes after labels or order change, the student should say whether that reflects the mathematics or only a convention.
Then request one deliberately awkward example. The learner should predict the difficulty before running the method, explain what the output does and does not establish, and name a reasonable alternative or safeguard. This check joins correctness with judgement: a technically valid calculation is useful only when its assumptions and interpretation remain visible.
Check 6: Handle an entire zero row
Ask the student to demonstrate how to handle an entire zero row without relying on a hidden software step. The evidence should include a labelled input, the exact rule used, one intermediate state and a result that can be checked independently. If the answer changes after labels or order change, the student should say whether that reflects the mathematics or only a convention.
Then request one deliberately awkward example. The learner should predict the difficulty before running the method, explain what the output does and does not establish, and name a reasonable alternative or safeguard. This check joins correctness with judgement: a technically valid calculation is useful only when its assumptions and interpretation remain visible.
Frequently asked questions
What does the Routh–Hurwitz criterion tell us?
How many characteristic roots lie in the open right half-plane, and therefore whether the declared continuous-time polynomial is stable.
Must every root be calculated?
No.
What does one sign change mean?
One root lies in the open right half-plane.
Are positive coefficients enough?
Not in general.
What if the first table entry is zero?
Use a symbolic positive epsilon and interpret the limit.
What does a whole zero row mean?
A symmetric root pattern that requires an auxiliary polynomial.
Are imaginary-axis roots stable?
They are boundary or marginal cases, not asymptotically stable.
Does this work for discrete time?
Not directly.
Can it choose a controller gain?
It can derive necessary stability ranges for a parametrised characteristic polynomial.
What should students know first?
Polynomials, complex roots, determinants, inequalities and linear systems.
Useful next reading and sources
Source: MIT OpenCourseWare lecture on system stability and the Routh–Hurwitz criterion.
Source: Hurwitz’s 1895 paper on roots in the negative half-plane.
Related eduKateSG reading: Durand–Kerner complex roots.
Related eduKateSG reading: Gauss–Seidel iterative solves.
Related eduKateSG reading: Broyden nonlinear systems.
Read sources with a question in mind: which definition, theorem, default or limitation is being supported? An authoritative link is useful evidence, but it does not replace a worked calculation or a clear statement of how the source applies to the example.
A cheerful final thought
Mathematics becomes friendlier when every symbol has a job and every claim has a boundary. Start small, make the state visible, test the awkward case, and celebrate the moment when a pattern becomes explainable. That is the practical benefit of learning mathematics: not certainty about every future, but a growing ability to ask better questions and build trustworthy answers.