ROI Case File No.353 | 'InspectTech's Invisible Risk'

Chapter One: The Danger of Speculation—Divergence Between Blueprints and Reality

The day after resolving the invoice management case at Global Invoice Solutions, another consultation arrived regarding building inspection visualization. Volume 29, "The Pursuit of Reproducibility," Episode 353, tells the story of clarifying ambiguity.

"Detective, we conduct building renovation work. However, we cannot determine the condition of pipes and ducts above ceilings. We want to visualize them without opening inspection ports. Because the actual conditions often differ from original construction blueprints, significantly impacting construction plans and estimates."

Kenta Yamada, technical manager from InspectTech, born in Kawasaki, visited 221B Baker Street with a confused expression. In his hands were blueprints of a 30-year-old building and, in stark contrast, a report noting "Water leakage accidents: 12 annually."

"We specialize in renovation work for commercial buildings and office buildings. Forty-five employees. Annual revenue of 1.8 billion yen. We receive approximately 120 renovation projects annually."

InspectTech's Current State: - Established: 2005 (Building renovation work) - Employees: 45 - Annual revenue: 1.8 billion yen - Annual renovation projects: Approximately 120 - Problems: Unknown condition of ceiling pipes/ducts, divergence between blueprints and reality, speculative proposals

Deep impatience permeated Yamada's voice.

"The renovation work flow is as follows. First, we receive a request from customers: 'The pipes may be deteriorating. Please inspect them.' Next, we check the original construction blueprints. However, pipes are often not positioned according to blueprints.

For example, blueprints show 'pipe runs straight from Point A to Point B,' but in reality it's 'Point A through Point C to Point B.' Because pipe routes were changed during original construction, but blueprints weren't updated."

Typical Problem Cases:

Case 1: Pipe Route Inconsistency - Blueprint: Straight route (5 meters) - Reality: Detour route (8 meters) - Impact: Pipe replacement costs increase 60%

Case 2: Duct Size Discrepancy - Blueprint: 300mm diameter - Reality: 400mm diameter - Impact: Parts ordering errors, construction delays

Case 3: Unknown Pipe Deterioration Locations - Blueprint: 30 years old, full replacement recommended - Reality: Only partial deterioration (leakage from connections) - Impact: Excessive estimates, customer distrust

Yamada sighed deeply.

"Currently, we make speculative proposals. 'It's been 30 years, so let's replace all pipes.' 'Inspection is necessary every X years.' However, we don't know if full replacement is truly necessary.

And 12 water leakage accidents occur annually. After accidents, when we open inspection ports to check, we find 'just replacing this spot would have been sufficient.' If we'd known in advance, we could have prevented them."

Chapter Two: The Expectation of Infrared—But Requirements Remain Vague

"Mr. Yamada, you want to visualize ceiling spaces using infrared or laser. What kind of system specifically are you envisioning?"

My question showed a confused expression on Yamada's face.

"Honestly, I don't know. I heard 'detecting temperature differences with infrared can identify leakage locations.' But I haven't organized specifically what (What), why (Why), when (When), where (Where), who (Who), and how (How) to inspect."

Current Understanding (Unclear Requirements Type): - Expectation: Visualization with infrared/laser - Problem: Specific requirements vague

I explained the importance of clarifying requirements with five Ws and one H.

"The problem is vagueness about 'what you want to achieve.' 5W1H—What, Why, When, Where, Who, How. Organize requirements with these six questions. Transform vague expectations into concrete specifications."

⬜️ ChatGPT | Catalyst of Conception

"Don't proceed with ambiguity. Clarify requirements with 5W1H. Specifications emerge from questions."

🟧 Claude | Story Alchemist

"Invisible risks always become visible through 'questions.' Establish the six questions."

🟦 Gemini | Compass of Reason

"5W1H is the foundation of requirements definition. Draw the complete picture with six questions, solidify specifications."

The three members began their analysis. Gemini displayed the "5W1H Framework" on the whiteboard.

The 6 Elements of 5W1H: 1. What: What is the inspection target? 2. Why: Why is inspection necessary? 3. When: When to conduct inspection? 4. Where: Where to conduct inspection? 5. Who: Who conducts inspection? 6. How: How to conduct inspection?

"Mr. Yamada, let's answer each question one by one."

Chapter Three: Clarification Through Questions—Six Answers Generate Specifications

Phase 1: Clarifying What (1 week)

Question: What is the inspection target?

Answers: 1. Pipes (water supply pipes, drainage pipes) 2. Ducts (for air conditioning) 3. Electrical wiring 4. Pipe deterioration locations (rust, corrosion, connection failures) 5. Locations with water leakage risk

Priority: 1. Pipe deterioration locations (highest priority) 2. Duct position/size 3. Electrical wiring (low priority)

Reasoning: - 12 water leakage accidents occur annually - Pipe deterioration is the greatest risk - Ducts necessary for improving renovation estimate accuracy

Phase 2: Clarifying Why (1 week)

Question: Why is inspection necessary?

Answers: 1. Risk avoidance: Prevent water leakage accidents (currently 12 annually) 2. Estimate accuracy improvement: Resolve discrepancies between blueprints and reality 3. Customer trust improvement: Data-based proposals instead of speculation

Specific Expected Effects: - Water leakage accidents: 12 annually → 3 (75% reduction) - Estimate accuracy: ±30% → ±5% (error reduction) - Customer satisfaction: NPS 52 → NPS 72

Phase 3: Clarifying When (1 week)

Question: When to conduct inspection?

Answers: 1. Renovation planning stage (before creating estimates) 2. Regular inspection time (15, 25, 35 years after construction) 3. Water leakage accident occurrence (emergency response)

Inspection Frequency: - Renovation work: 120 annually - Regular inspection: 40 annually - Emergency response: 12 annually - Total: 172 annually

Phase 4: Clarifying Where (1 week)

Question: Where to conduct inspection?

Answers: 1. Above ceilings (pipes, ducts) 2. Inside walls (pipes, wiring) 3. Below floors (pipes)

Inspection Target Areas: - Above ceilings: Highest priority (high water leakage risk) - Inside walls: Next priority (necessary for renovation work) - Below floors: Low priority (easy access)

Specific Scope: - Average inspection area: 500㎡/project (above ceilings) - Inspection time: 3 hours/project (conventional inspection port opening method)

Phase 5: Clarifying Who (1 week)

Question: Who conducts inspection?

Answers: 1. InspectTech engineers (15 in technical department) 2. Partner company inspection staff (external outsourcing also considered)

Required Skills: - Infrared camera operation - Laser scanner operation - Data analysis skills - Pipe/duct knowledge

Training Plan: - Training for 15 internal engineers (40 hours/person) - Technical instruction by external instructors

Phase 6: Clarifying How (2 weeks)

Question: How to conduct inspection?

Answers: 1. Infrared thermography: Detect leakage locations by temperature differences 2. Laser scanner: Acquire 3D shapes of pipes/ducts 3. AI image analysis: Automatically detect rust/corrosion locations

Technical Specifications:

Technology 1: Infrared Thermography - Principle: Visualize temperature differences (leakage locations are cooler than surroundings) - Detection accuracy: ±0.5℃ - Detection distance: Within 3 meters - Advantages: Non-destructive, rapid (30 min/100㎡) - Disadvantages: Cannot detect when no temperature difference

Technology 2: Laser Scanner - Principle: Measure 3D shapes with laser light - Accuracy: ±2mm - Scan speed: 500,000 points/second - Advantages: Acquire accurate pipe/duct position/size - Disadvantages: Expensive (equipment cost 8 million yen)

Technology 3: AI Image Analysis - Principle: Detect rust/corrosion from infrared images/laser data - Training data: 5,000 past inspection images - Detection accuracy: 92% - Advantages: Higher accuracy than human visual inspection - Disadvantages: 8% false detection

System Configuration: - Infrared cameras: 2 units (3 million yen/unit) - Laser scanner: 1 unit (8 million yen) - AI analysis software: 1 license (1.2 million yen annually) - Tablet devices: 5 units (150,000 yen/unit) - Total initial investment: 20.75 million yen

Chapter Four: Execution as Verification—Results After 6 Months

Phase 7: Pilot Implementation (Months 1-3)

Target: 10 renovation projects - Inspection target: Ceiling pipes/ducts - Engineers: 3 (training completed) - Investment: Pilot equipment rental (500,000 yen/month)

Results After 3 Months:

Inspection Time: - Before (inspection port opening method): 3 hours/project - After (infrared/laser method): 1.2 hours/project - Reduction: 1.8 hours/project (60% reduction)

Estimate Accuracy: - Before: ±30% - After: ±8% - Improvement: 73% error reduction

Specific Cases:

Case 1: Pipe Route Discovery - Customer: 28-year-old office building - Blueprint: Pipe runs straight route (5 meters) - Reality: Laser scan discovers detour route (8 meters) - Effect: Corrected estimate, presented 60% cost increase in advance, customer satisfied

Case 2: Water Leakage Risk Detection - Customer: 32-year-old commercial building - Blueprint: Full replacement recommended - Reality: Infrared detected temperature drop at only 2 connection points (leakage risk) - Effect: Proposed partial replacement (1.2 million yen) instead of full replacement (4.8 million yen), customer satisfied

Customer Feedback: "Previously, we were told 'it's been 30 years, so full replacement.' But InspectTech showed us the actual condition with infrared. 'Just replace these two spots.' Construction costs reduced 75%. Trustworthy."

Phase 8: Full Implementation (Months 4-9)

Company-wide Rollout: - Applied to 120 annual renovation projects - All 15 engineers completed training - Production equipment purchased: 20.75 million yen

Results After 9 Months (Annual Conversion):

Inspection Time Reduction: - Before: 3 hours/project × 120 projects = 360 hours/year - After: 1.2 hours/project × 120 projects = 144 hours/year - Reduction: 216 hours/year - Personnel cost reduction: 216 hours × 4,500 yen = 972,000 yen/year

Order Rate Improvement Through Estimate Accuracy Improvement: - Before: Order rate after estimate presentation 68% - After: Order rate after estimate presentation 84% - Improvement: +16 points - Effect: 120 projects annually × 16% = 19 additional orders - Sales increase: 19 projects × average 15 million yen = 285 million yen/year - Gross profit increase (gross margin 20%): 285 million yen × 20% = 57 million yen/year

Water Leakage Accident Reduction: - Before: 12 annually - After: 3 annually - Reduction: 9/year - Accident response cost reduction: 9 projects × 800,000 yen = 7.2 million yen/year

Customer Satisfaction Improvement: - Before: NPS 52 - After: NPS 74 - Improvement: +22 points

Annual Effect (Total): - Inspection time reduction: 972,000 yen - Gross profit increase: 57 million yen - Accident response cost reduction: 7.2 million yen - Total: 65.172 million yen/year

Investment Recovery: - Initial investment: 20.75 million yen - Annual operation cost: 1.2 million yen (AI analysis software license) - Annual net effect: 65.172 million yen - 1.2 million yen = 63.972 million yen - ROI: (63.972 million yen - 20.75 million yen) / 20.75 million yen × 100 = 208% - Investment recovery period: 20.75 million yen ÷ 63.972 million yen = 0.32 years (3.9 months)

Phase 9: Continuous Improvement (Month 10-)

5W1H Re-evaluation:

New Questions: 1. What: Should electrical wiring be added to inspection targets beyond pipes/ducts? 2. Why: Electrical wiring deterioration also poses fire risk 3. When: Should emergency response inspection frequency increase? 4. Where: Should inspection scope expand to inside walls and below floors? 5. Who: Should technology be deployed to partner companies to expand inspection structure? 6. How: How to improve AI detection accuracy from 92% → 98%?

Next Actions: - Expand AI training data from 5,000 → 20,000 images - Additional introduction of electrical wiring inspection equipment (5 million yen) - Conduct technical training for 3 partner companies

Organizational Transformation:

Engineer A's Voice: "Previously, we opened inspection ports for visual confirmation. Entering ceiling spaces was difficult. Dark, narrow, dusty. Took 3 hours per project.

But after using infrared cameras and laser scanners, we can inspect without opening inspection ports. Completes in 1.2 hours. And when we show 3D images to customers, they're surprised: 'There are pipes here!' I feel trusted."

Yamada's Reflection:

"Until we conducted 5W1H, we had only vague expectations like 'infrared should enable something.' However, by answering the six questions—What, Why, When, Where, Who, How—concrete specifications emerged.

What to inspect (pipe deterioration locations), why to inspect (water leakage accident prevention), when to inspect (renovation planning stage), where to inspect (above ceilings), who inspects (internal engineers), how to inspect (infrared/laser/AI).

We achieved 65.172 million yen annual effect, ROI 208%, investment recovery in 3.9 months. And customer satisfaction improved by 22 points.

Making invisible risks visible. That is our mission."

Chapter Five: The Detective's Diagnosis—Questions Generate Specifications

That evening, I contemplated the essence of 5W1H thinking.

InspectTech had vague expectations like "infrared should enable something." Concrete requirements were not visible.

By answering six questions with 5W1H, vague expectations transformed into concrete specifications. What (pipe deterioration locations), Why (water leakage accident prevention), When (renovation planning stage), Where (above ceilings), Who (internal engineers), How (infrared/laser/AI).

"Don't proceed with ambiguity. Clarify requirements with 5W1H. Answer the six questions: What, Why, When, Where, Who, How. From there, specifications emerge."

The next case will surely depict another moment of clarifying requirements through questions.

"What, Why, When, Where, Who, How. Clarify requirements with six questions. Transform vague expectations into concrete specifications. Questions open the path to realization."—From the Detective's Notes

Related Files