Wait, what? Why does one broken part sometimes stop an entire system, while another broken part barely changes the output?

The difference can be structure.

Some systems depend on one critical route. Others have alternative pathways, duplicated functions, stored reserves or components that can temporarily take over. A system with those features may continue operating after a local failure, although perhaps at lower performance.

This preserved Hougang Primary 5 Science URL now owns one precise job: redundancy, backup pathways and system resilience. The old duplicated 2019 tuition advertisement, stale locality claims, grade promises and unrelated image stack have been removed.

This page is deliberately distinct from the existing P5 owners. Bottleneck reasoning asks what limits performance. Counterfactual reasoning asks what changes if one part is removed. Feedback asks how effects return. This page asks:

When one part fails, what structure allows the rest of the system to keep functioning—and what becomes the new weak point?

Redundancy means more than one way to perform a needed job

In systems thinking, redundancy means there is more than one component, route or resource capable of supporting a required function.

A simple structure is:

Function F can be supported by route A or route B.

If A fails and B still works, the function may continue.

The system is not necessarily unaffected. B may become overloaded, slower or less efficient. Redundancy creates resilience, not invulnerability.

Single point of failure

A single point of failure is a component whose failure stops a critical function because no alternative route is available.

Primary 5 students do not need the formal engineering term to reason about it. They can ask:

• Is this the only pathway?
• Is this the only source?
• Is this the only component performing the function?
• If it fails, can the system route around it?

If the answer is no alternative, the part is highly critical.

Parallel pathways create different failure behaviour

Suppose a system has two parallel routes carrying the same type of flow.

If one route is blocked:

• flow through that route falls;
• the other route may continue carrying flow;
• total output may decrease rather than fall to zero;
• the remaining route may become the new bottleneck.

This is different from a single serial chain where one break can interrupt the entire route.

Series versus parallel thinking

A useful structural distinction is:

• Series dependency: A must work before B, and B before C.
• Parallel dependency: A or B can each provide a route toward C.

In a series chain, one failure can propagate strongly. In a parallel structure, one failure may be absorbed partially.

Students should learn to see the architecture before predicting the effect of failure.

Resilience is the ability to keep functioning after disturbance

A resilient system can absorb some disturbance and continue performing its essential function.

Resilience can come from:

• alternative pathways;
• extra capacity;
• stored resources;
• components with overlapping roles;
• feedback that counteracts disturbance;
• ability to reduce non-essential activity temporarily.

Different systems use different combinations of these mechanisms.

Backup capacity can be hidden until failure occurs

A system may look inefficient because not every component is operating at maximum output.

But unused capacity can become valuable during disturbance.

Example structure:

• Normal condition: Routes A and B each carry part of the load.
• Disturbance: Route A fails.
• Response: Route B carries more.
• Outcome: Total function decreases only partly.

The system had reserve capacity.

This connects resilience to trade-offs: extra capacity may cost resources during normal operation but reduce failure risk.

Resilience and bottlenecks

Redundancy can move the bottleneck.

Suppose two pathways feed one shared outlet.

If one pathway fails, the other may compensate—but the shared outlet may now become limiting.

The system’s weakest point changes with state.

• Before failure: both pathways operate comfortably.
• After failure: the surviving path carries more load.
• New condition: another component reaches capacity first.

Resilience analysis therefore includes a new-bottleneck check.

Resilience and feedback

A feedback response can help stabilise a disturbed system.

General pattern:

disturbance → system variable changes → compensating response increases → variable moves back toward a workable range

Feedback does not create redundancy by itself, but it can add resilience by changing how remaining components respond.

This is one reason dynamic systems can survive disturbances that would break a rigid one-path model.

Resilience and stores

Stored resources can buffer short disruptions.

A system may continue briefly even after an input stops because some resource is already stored.

This creates a time distinction:

• immediate effect may be small;
• later effect becomes larger as the store is depleted.

Students should not conclude “the part was unnecessary” simply because the system did not fail instantly.

Resilience and alternative food sources

Food webs provide a useful example of redundancy.

If a consumer has several food sources, reduction of one source may not have the same effect as losing its only food source.

Ask:

• Are alternative food sources shown?
• Do they provide enough food?
• Will competition increase for the remaining sources?
• Does the population decline partially or severely?

Network redundancy changes the size of the cascade.

Redundancy can reduce efficiency

Keeping backup capacity has costs.

• more material;
• more energy;
• more maintenance;
• more complexity;
• more space.

A system designed only for maximum efficiency may remove spare capacity. A system designed for resilience may accept some redundancy.

This creates a trade-off:

efficiency under normal conditions ↔ ability to absorb failure

Primary 5 students can reason about the trade-off without advanced engineering language.

Redundancy is not duplication without purpose

Two components are not useful backups merely because there are two of them.

For redundancy to increase resilience:

• both must be capable of supporting the function;
• one must remain available when the other fails;
• they should not share the exact same single failure cause;
• the downstream system must be able to use the backup route.

If both backups depend on one shared critical component, that shared component remains a single point of failure.

Shared dependencies can hide fragility

A system may appear to have two independent routes while both depend on the same upstream source.

Example:

• Route A and Route B are separate;
• both require Source S;
• failure of A alone is absorbed;
• failure of S disables both.

The true critical point is S.

This teaches students to look beyond visible duplication to dependency structure.

Resilience can be partial

Systems do not always fall into “working” or “failed”.

After disruption, a system may:

• continue normally;
• continue at reduced output;
• work more slowly;
• protect essential functions while reducing others;
• operate temporarily before stored resources run out;
• fail completely.

Primary 5 reasoning becomes more realistic when failure is treated as a spectrum.

Resilience can depend on disturbance size

A backup route may absorb a small failure and fail under a larger one.

For example:

• one pathway blocked → alternative route handles the extra load;
• two pathways blocked → remaining route reaches capacity;
• shared source lost → all routes fail.

Resilience has an operating range.

Ask:

How much disturbance can the system absorb before the backup structure is no longer enough?

Resilience and recovery are different

A system can resist a disturbance or recover after being disturbed.

• Resist: performance changes little during disturbance.
• Recover: performance falls but later returns.

These are different system behaviours.

A backup path may provide resistance. A repair or feedback process may support recovery.

The learner should identify which behaviour the evidence shows.

The failure-map method

1. Function: What must the system keep doing?
2. Pathways: Which parts support that function?
3. Critical: Which parts have no backup?
4. Redundant: Which parts have alternatives?
5. Disturb: Remove or block one part.
6. Shift: Where does the load move?
7. Bottleneck: What becomes limiting next?
8. Output: Does performance remain, fall partly or stop?
9. Time: Is the effect immediate or delayed?
10. Recover: Can the system return after the disturbance?

This connects structure, causality and system response.

The double-failure test

If one failure is absorbed, test a second failure mentally.

• Does the backup still work?
• Do both routes depend on the same source?
• Does capacity become insufficient?
• Does the system cross from reduced performance to total failure?

This reveals whether resilience is deep or only superficial.

The common-cause test

Two backups can fail together if they share the same vulnerability.

Ask:

• Do both pathways depend on the same resource?
• Are both affected by the same environmental condition?
• Do both rely on the same upstream component?

Redundancy is stronger when backups fail independently rather than together.

The graceful-degradation idea

A resilient system may lose performance gradually rather than collapse instantly.

For example:

• 100% output with all routes;
• 70% output after one route fails;
• 40% after another constraint appears;
• 0% only when the final critical path is lost.

The exact numbers are illustrative. The reasoning lesson is that partial function can persist.

Five Primary 5 resilience failure modes

1. One-break-equals-total-failure thinker

Assumes every broken component stops the whole system. Repair by mapping alternative pathways.

2. Backup-means-no-effect thinker

Assumes redundancy means performance is unchanged. Repair by checking capacity and load shifting.

3. Duplicate-means-independent thinker

Two routes are assumed independent even though they share one source. Repair with the common-dependency test.

4. Immediate-result thinker

No instant failure is taken as proof the part is unnecessary. Repair by checking stored resources and delayed effects.

5. Resilience-is-free thinker

Backup capacity is assumed to have no cost. Repair by connecting resilience to material, energy and complexity trade-offs.

A Phase 4 Primary 5 resilience lesson

• Function: define what must continue.
• Map: draw dependencies and routes.
• Critical: mark single points with no alternatives.
• Redundant: mark backup paths or overlapping functions.
• Disturb: remove one part.
• Shift: trace where the load or flow moves.
• Capacity: ask whether backups can carry the new load.
• Time: separate immediate and delayed effects.
• Common cause: test whether backups share a vulnerability.
• Trade-off: ask what the redundancy costs during normal operation.

Why small groups help resilience reasoning

Give three students the same system diagram and remove one component.

• One may predict total failure.
• One may find a backup route.
• One may notice the backup shares the same source.

The disagreement reveals which dependencies each learner can see.

What parents can practise at home

• Ask “is there another route?”
• Ask “what happens if this part fails?”
• Ask whether the system stops completely or only loses some performance.
• Ask what becomes the new bottleneck.
• Ask whether the backup shares the same source.
• Ask what extra cost comes from keeping backup capacity.

How to tell whether resilience reasoning is improving

• Single points of failure are identified.
• Parallel and serial dependencies are distinguished.
• Alternative pathways are checked before predicting collapse.
• Partial failure is separated from total failure.
• Load shifting and new bottlenecks are traced.
• Stored resources and delays are considered.
• Shared vulnerabilities are noticed.
• Resilience costs and trade-offs are recognised.

How this page fits the Hougang Science network

This eduKateSingapore page owns redundancy, backup pathways and resilience. It complements constraints and bottlenecks, counterfactual intervention, feedback loops and cascading effects, and trade-offs and competing constraints.

For the complete P3-to-PSLE map, use Hougang Primary Science Learning Library.

Official curriculum reference

The Ministry of Education’s Science Teaching & Learning Syllabus: Primary Three to Six develops Systems and Interactions alongside prediction, analysis and explanation. Resilience reasoning is used here as an age-appropriate systems-thinking scaffold within those curriculum boundaries.

---

Primary 5 Science becomes more realistic when failure is not automatically treated as collapse. Map the routes, find the single points of failure, identify backups, trace load shifting, test shared vulnerabilities and ask what the system can still do after one part is lost.