TOOL / Free self-assessment

Industrial data readiness assessment for AI in manufacturing

28 questions across seven dimensions, from connectivity and ISA-95 context to data quality, stewardship and the use-case pipeline. You get a 0-100 index, a stage and a plan built from your answers.

10 minutes 28 questions · 7 dimensions Answers stay in your browser

ABOUT

About this assessment

Most AI projects in manufacturing stall on data, not models: tags without context, historians that only cover half the lines, timestamps that drift, and nobody who owns the asset hierarchy. This assessment checks whether your plant data can support predictive maintenance, quality analytics and AI assistants at scale, not just one heroic pilot. Answer for a typical plant. Pick the level you could demonstrate today. "Not sure" scores as level 1 and is flagged for follow-up.

  • An industrial data readiness index from 0 to 100 and a maturity stage
  • Scores for 7 dimensions against a target for scaling AI beyond pilots
  • Your three weakest areas with next actions for the answers you chose
  • A 30/60/90-day plan for data foundations

MODEL

What the assessment covers

DIM-01

Inventory & connectivity

Knowing which data sources exist and whether machine, process and quality data can be collected reliably.

4 questions · target 3/5

DIM-02

Context & models (ISA-95)

Asset hierarchy, naming and context so data from different lines and plants means the same thing.

4 questions · target 3/5

DIM-03

Data quality

Completeness, accuracy, timeliness and trust in the data used for decisions and AI.

4 questions · target 3/5

DIM-04

Ownership & stewardship

Named owners and stewards for data, with clear rights and duties.

4 questions · target 3/5

DIM-05

Security & access

Controlled, secure access to plant data for people and applications.

4 questions · target 3/5

DIM-06

Platform & tooling

The data platform, integration and tools that let you build and run analytics and AI repeatably.

4 questions · target 2.5/5

DIM-07

Skills & use-case pipeline

People who can turn data into results, and a steady pipeline of valuable use cases.

4 questions · target 3/5

SCALE

Maturity stages

  1. 0–20

    1. Fragmented

    Data is scattered, hard to reach and poorly understood. AI pilots will mostly measure data problems. Start with inventory, one connected line and owners.

  2. 21–40

    2. Emerging

    Some lines are connected and some data is trusted. Focus on context, naming and fixing quality at source.

  3. 41–60

    3. Structured

    Core data is connected and contextualized for key lines. Focus on a shared model, platform and stewardship across plants.

  4. 61–80

    4. Scalable

    Data foundations support repeatable use cases. Focus on reuse, monitoring and value tracking.

  5. 81–100

    5. Optimized

    Data is a managed asset across plants. Keep governance lean and aligned with new use cases.

EXHIBIT 1

Sample result

Fictional 3-plant automotive parts supplier, 600 staff

Inventory & connectivity2.5/5
Context & models (ISA-95)1.8/5
Data quality2.0/5
Ownership & stewardship1.5/5
Security & access2.3/5
Platform & tooling1.8/5
Skills & use-case pipeline2.0/5

Fictional company. Illustrative scores only.

All 28 questions

Inventory & connectivity

  1. Do you have an inventory of data sources (PLCs, historians, MES, LIMS, CMMS, ERP, spreadsheets)?
  2. What share of critical machines and lines are connected for data collection?
  3. Do you have a process historian or time-series store for machine and process data?
  4. Is data collection designed with OT security in mind? (IEC 62443-3-2 (zones and conduits))

Context & models (ISA-95)

  1. Is there a consistent asset hierarchy (enterprise, site, area, line or work center, unit)? (ISA-95 / IEC 62264 equipment hierarchy)
  2. Do tags and data points have consistent naming, units and descriptions?
  3. Can you link process data to production context (order, product, batch, shift, recipe)? (ISA-95 / IEC 62264)
  4. Do you have a shared information model or unified namespace?

Data quality

  1. Do you know how complete and accurate your key production data is?
  2. Are timestamps and time synchronization reliable across systems?
  3. How are manually entered data (downtime reasons, scrap codes, inspections) captured?
  4. Do data issues get fixed at the source?

Ownership & stewardship

  1. Does each key data domain (equipment, process, quality, production, maintenance) have an owner?
  2. Who decides whether data can be used for a new analytics or AI purpose?
  3. Is master data (materials, equipment, recipes) maintained centrally and consistently?
  4. Are there rules for sharing data with OEMs, vendors and customers?

Security & access

  1. How do analysts and data scientists get access to plant data?
  2. Is sensitive data (recipes, customer specs, personal data) protected in analytics and AI tools?
  3. Are cloud and edge data connections approved and monitored? (IEC 62443-3-3 FR5 (Restricted data flow))
  4. Is personal data in plant systems (operator IDs, badge data, video) identified and handled lawfully? (GDPR (where applicable))

Platform & tooling

  1. Is there a common data platform for analytics across plants?
  2. How are integrations between OT and IT systems built and maintained?
  3. Can you move a working analytics or AI model from one line to another?
  4. Are models and analytics in production monitored and maintained?

Skills & use-case pipeline

  1. Do you have people who understand both the process and the data?
  2. How are AI and analytics use cases selected?
  3. Do pilots move into production?
  4. Are operators and maintenance staff involved in data and AI projects?

Related guides and templates

FAQ

Frequently asked questions

What does industrial data readiness mean?

It means your plant data can support analytics and AI reliably: sources are connected, data has context such as asset, order and batch, quality is known, owners exist, access is secure, and there is a platform and a pipeline of use cases. A readiness score helps you fix foundations before spending on models that will stall on data.

Do I need a unified namespace before starting AI?

No. A unified namespace is one way to share contextualized data, and it pays off at scale. Many mid-market plants start with a well-structured historian and a consistent asset hierarchy on one line. The assessment checks whether you have a shared information model, but a lower score there is not a blocker for a first use case.

How does this relate to the AI governance assessment?

This assessment focuses on data foundations for AI. The AI governance readiness assessment covers ownership, policy, risk, vendors and oversight. Data governance appears in both, from different angles. Running both gives a fuller picture of whether you can scale AI safely.

Is my data stored anywhere?

Answers stay in your browser's local storage so you can resume later. Nothing is sent to a server. Use "Start over" to clear them.