Define the Right Problem: Innovation Starts With a Better Problem Statement

Most teams react to a failure as if the failure itself is the target. A defect appears, scrap happens, pressure rises, and the organization moves fast to “fix what happened.”
Innovation begins earlier than that.
Innovation starts when you define the right problem: the solvable gap created by the failure. One failure can create multiple problems, and each problem can lead to a different improvement path. When the problem is framed well, the team gains a clean runway for both root cause analysis and inventive solutions.
If you want a deeper explanation of the “failure vs. problem” distinction, link this sentence to your post “What is the problem?”.

Case study: defects after a chemistry batch change

Imagine your company produces high-quality glass discs — clean, transparent, and visually perfect.

Your manufacturing flow includes formation, thermal treatment, etching, thin-film deposition, polishing, and more. The final step is washing with pure cleaning chemistry. That last wash step is sensitive and can significantly affect the final yield.

In a typical setup:

• The disc is held from the back side
• The disc rotates
• Cleaning chemistry dispenses through a nozzle toward the center of the rotating disc.

What happened (facts + evidence)

Quality control flags an excursion: several discs show unusual, unknown defects. The affected discs are scraped.

Engineering investigates and finds a correlation: the defects begin appearing after a new batch of cleaning chemistry is introduced. Comparative chemical analysis shows both the previous batch and the suspected batch remain within specification limits. One component concentration is slightly elevated in the suspected batch, still within the spec window.

Now comes the fork in the road:

• Do we treat “the batch” as the problem?
• Or do we define the right problem that the process must solve, regardless of batch variation?

The symptom-fix trap (blaming the batch)

The wrong “problem”

A common framing sounds like this:

“The new batch of chemistry is contaminated. The composition does not match typical chemistry.”

The typical direction

That framing pushes predictable actions:

• Remove the batch from production, return it or dispose of it, request a replacement batch
• Tighten specification limits so it never happens again

What you gain (and what you lose)

This path can restore production quickly. It also brings three long-term penalties:

1. Cost rises (tighter specs, more rejects, supplier conflict)
2. Excursions repeat (same symptom, new label; different component next time)
3. Innovation stalls (the process stays fragile; learning stays shallow)

When teams feel “done” after disposing of a batch, they miss the most valuable part of the failure: the mechanism lesson that can improve the process.

Any failure is an opportunity for innovation.

The correct workflow:
problem statement → root cause analysis → innovation

Let’s run the same case using a structured workflow (as supported inside the PRIZ Guru platform / PRIZ Innovation Platform). The goal stays practical: define the right problem, perform root cause analysis, then generate solutions that improve robustness and performance.

Phase 1 — Problem statement (5 steps)

A good problem statement is built from observable facts and measurable consequences. In PRIZ, it’s defined in three steps.

1) Current situation (what you see)

Describe the deviation from normal operation—the failure as observed:

Current situation:
Defects were found on the surface of the wafer after washing with the chemistry and rinsing with water.

2) Disadvantage (why it matters)

State the operational impact in plain terms (yield, scrap, delivery risk, customer commitments):

Disadvantages:

• The wafer contaminated with defects cannot be sold to the customer. The defective wafers will be scrapped.
• The process strongly depends on chemicals. Small change, even within spec limits, causes contamination of the wafers and a drop in the yield.
• The process and products are at high risk of failure due to even insignificant changes.

3) Ideal Final Result (IFR)

Ask yourself, “If I can get the perfect result, with a snap of the fingers, what would that be?”

IFR:

Stable yield after using the chemistry from different vendors with no cost increase.

4) Gaps

The question that we need to answer here is “If we know what our ideals are, what prevents us from getting there?”

Gaps:

A certain type of chemistry creates and leaves defects on the silicon wafer during wet cleaning

5) The problem statement (the solvable gap)

Now, combine all the previous steps into an actionable statement that a team can solve:

Problem statement:
The current wet process for wafer cleaning results in defects on the wafer surface, leading to scrapped wafers and decreased yield.

That sentence is the pivot. It defines the right problem.

It focuses the team on what must change in the process. It keeps the analysis stable even when the suspected “culprit” changes (batch A, batch B, supplier X, supplier Y).

Quick quality-check for your problem statement:

• It describes an outcome you can influence (not a suspect you want to blame).
• It stays true even if your first hypothesis is wrong.
• It points toward mechanism learning (how the process creates the result).

Phase 2 — Root cause analysis (cause–effect chain + mechanism)

With the right problem defined, the team can now do root cause analysis without drifting into supplier speculation.

PRIZ supports RCA through a 5+ Whys: a structured “Why?” chain that helps teams build causal logic on the left side, and capture solution ideas on the right side as the chain develops.

This matters because a strong RCA produces more than a label (“bad batch”). Strong RCA produces:

• a causal chain you can test, and
• a mechanism you can redesign.

What the analysis reveals (the mechanism model)

A surface-level conclusion is tempting: “the batch caused it”. In many teams, that becomes the endpoint.

A mechanism-focused RCA goes further. Summarizing the causal logic and generated ideas produces a model like this:

1. Cleaning chemistry dispenses onto the disc through a nozzle.
2. The disc rotates; rotation drives chemistry outward.
3. Rotation and airflow create a low-pressure zone near the disc surface, increasing evaporation.
4. When chemistry leaves the disc mainly via centrifugal removal, particles wash away with the liquid.
5. When chemistry leaves the disc mainly via evaporation, particles remain, and defects persist.

This model shifts the team from “chemistry blame” to process robustness. It explains why a within-spec variation can still trigger defects: the process is operating near a boundary where evaporation wins.

This pattern appears in many processes: any step where contaminants must be removed by a fluid can fail when evaporation, airflow, orientation, temperature, pressure, or time shifts the balance away from removal. The “right problem” often lives in the mechanism: what causes the system to stop doing the cleaning job reliably.

Phase 3 — Move from RCA to solution space

Now the problem becomes concrete:

We want chemistry to leave the disc primarily via removal (wash-off), not evaporation.

That goal leads naturally to solution directions such as:

• increase local pressure or reduce airflow effects near the surface
• reduce air temperature
• adjust rotation profile or timing
• modify delivery geometry (coverage and residence time)

Soon, the team encounters a classic tradeoff:

• increasing chemistry volume improves cleaning,
• but chemistry volume increases the cost.

This is an engineering contradiction: improving one parameter worsens another.

Breaking contradictions with the 40 inventive principles

The 40 Inventive Principles come from TRIZ, a structured innovation approach developed from large-scale patent analysis. In practice, the principles work as thinking prompts: they suggest patterns that have solved similar contradictions in other contexts.

Important reminder:

Principles are not solutions. Principles help teams generate solutions.

Once you define the contradiction and select improving/worsening parameters, PRIZ can suggest a small set of principles to explore.

In this example, the platform suggests four:

• #7 Nesting dolls
• #15 Dynamism
• #13 The other way around
• #16 Partial or excessive action

Three example ideas generated from principles

Idea 1 (Principle #7 — Nesting dolls): Replace a single nozzle with a “shower”

A single nozzle creates a specific coverage pattern. A shower head creates many micro-nozzles and can be designed with variable hole sizes.

Concept: Replace the single nozzle with a shower-style distributor to increase uniformity and residence time across the disc surface.

Idea #1: Replace the nozzle with a shower-style delivery head.

Idea 2 (Principle #15 — Dynamism): Make the flow profile adaptive

A constant flow treats the entire wash interval the same. An adaptive profile can target when removal is most needed.

Concept: Use variable flow, especially near the end of the process, to improve particle removal and reduce the chance that evaporation “wins” during the final moments.

Idea #2: Use a variable flow profile during cleaning.

Idea 3 (Principle #13 — The other way around): Flip orientation

Why keep the disc “face up” during washing? Orientation influences drainage and how droplets and particles behave during spin and evaporation.

Concept: Wash with the disc “face down” to improve drainage and support removal.

Idea #3: Keep the disc “face down” during washing.

Exercise (Principle #16 — Partial or excessive action)

Read the principal explanation in the platform and generate at least two concepts that apply “slightly more / slightly less” to the variables that control evaporation vs removal (time, rotation, airflow, pressure, temperature, chemistry volume).

Checklist: define the right problem every time

Use this checklist to force clarity before your team invests time in debates and actions:

1. Write the problem statement as an outcome gap
   “X remains / fails / exceeds / drops after step Y under conditions Z.”
2. Confirm the disadvantage in business terms
   Yield, scrap, delivery risk, customer impact.
3. Run root cause analysis until you can describe a mechanism
   Not a label, an explanation that makes predictions.
4. Turn tradeoffs into contradictions
   Name what improves and what worsens.
5. Use inventive prompts to expand the solution space
   Principles help teams escape “more of the same.”

Two fast examples (to reinforce generalization):

• If solder voids remain after reflow cleaning, the right problem often lives in flux removal dynamics, not “bad paste.”
• If residues remain after wafer rinse, the right problem often lives in drying/evaporation conditions, not “bad chemical.”

Important reminders

Any failure is an opportunity for innovation.

The key to innovation success is the correct definition of the problem.