AI is pattern recognition on structured data. If your data is messy, your AI will be hit or miss.
You can have the smartest model in the world — but if your inspection data isn’t structured around assets and code, your AI is guessing.
How to Structure Asset & Question Data
A checklist for building the data foundation that makes AI actually useful.
Why Data Structure Matters
Structured data isn’t glamorous, but it’s the lever that unlocks AI.
The Problem with Unstructured Data
| Unstructured | Structured |
|---|---|
| Free-form notes in a text box | Standard question with typed answer |
| “Checked the sprinkler heads, looked good” | Asset ID #4521, Status: Pass, Photo: attached |
| PDF form with handwritten entries | Digital form tied to building and asset |
| Email thread about a deficiency | Deficiency record linked to asset, severity, and recommendation |
Free-form notes create guesswork. Structured fields create action.
The Data Structure Blueprint
1. Unique IDs for Everything
Every entity needs a unique identifier that persists over time:
Why it matters: IDs let AI connect the dots across inspections, buildings, and time.
2. Standard Question Sets Tied to Code
Replace free-form fields with structured questions:
Why it matters: Standard questions = comparable data = patterns AI can learn from.
3. Typed Answers, Not Free Text
Wherever possible, use structured answer types:
| Instead of… | Use… |
|---|---|
| “Pressure was fine” | Pressure: 125 PSI |
| “Some corrosion noted” | Corrosion: Moderate (dropdown) |
| “Needs service” | Severity: Critical / Major / Minor |
| “Talked to the customer” | Contact: John Smith, Date: 2026-01-15 |
Why it matters: Typed data can be analyzed, compared, and acted on. Free text can't.
4. Photos with Context
A photo without context is just a file. A photo with context is evidence.
Why it matters: AI can analyze photos — but only if it knows what it's looking at.
5. Historical Data Linked Over Time
The real value emerges when AI can see patterns across inspections:
Why it matters: AI can catch anomalies, predict failures, and ensure consistency — but only with historical continuity.
Data Hygiene Checklist
Run through this quarterly to keep your data AI-ready:
Assets
Inspections
Deficiencies
Customers & Buildings
Quick Wins: Start Here
If you’re not sure where to begin, start with these three:
1. Asset IDs
Pick one building. Ensure every asset has a unique ID. Use that as your template.
2. One Standard Question Set
Pick your most common inspection type. Standardize the questions. Roll it out.
3. Photo Requirements
Require a photo for every deficiency. Link it to the asset. Enforce it.
One asset ID, one clear question, one photo done right — repeated a thousand times — builds an AI you can trust.
Key Takeaways
- AI is pattern recognition on structured data. No structure = no patterns.
- Unique IDs are the foundation. Everything connects through IDs.
- Standard questions beat free text. Comparable data = useful AI.
- Photos need context. Tag them, link them, require them.
- Historical continuity unlocks the magic. Same IDs, same questions, over time.
Your Action Item
Pick ONE data source to structure this week:
Start small. The discipline compounds.
Learn More About Data & AI
Watch Episode 4 for a deeper dive into how structured data supercharges AI.