Research and Development Business Capability Model: AI-Ready Edition

This Research and Development Business Capability Model is for CIOs, architects and transformation leads who have to decide where AI goes in R&D and in what order. It describes what R&D does in 10 L1, 40 L2 and 158 L3 capabilities, independent of the lab systems and product lifecycle tools in use. Every L2 capability carries an AI profile with a target AI mode, a human oversight level and a risk tier, and the prioritization model scores and ranks all 40 with ratings and weights you set. A 40-question readiness assessment shows what has to be in place before the first pilot.

About the R&D Business Capabilities Map

A capability model describes what a function does, not how it is organized or which systems run it. This one breaks research and development into 10 level 1 areas, 40 level 2 capabilities and 158 level 3 capabilities, each with a definition, so that an activity like testing (RD-5.2) has one name and one ID that research, engineering and quality teams can all point to. The 161 KPIs at L2 give each capability a measure that does not depend on a particular tool or team.

The AI layer sits on that structure. Each L2 capability has a full AI profile: where AI could do useful work, what stays with a person, the agents that could support it, the data it needs, the prerequisites, the risks and the effect on the workforce. Each L3 carries ratings for work pattern, target AI mode, AI fit and risk tier. The testing profile in the public sample shows the level of detail: AI drafts test plans and traces each test to the requirement it covers, while the Test and Validation Lead approves the plan and the Quality Lead approves any conclusion that bears on safety or quality.

A plain capability list stops at names and definitions. Here the ratings feed a prioritization model with a risk-adjusted score and rank for each L2 capability, an AI heatmap shows the pattern across the whole model, and a 40-question weighted readiness assessment checks whether the organization can support what the ranking suggests. The ratings describe a typical large enterprise; you replace them with your own.

Why It Matters

R&D is usually described in the vocabulary of its projects, labs and tools: the requirements system, the test bench, the stage-gate calendar. That makes it hard to compare one program with another or to see where the same work is done twice by different groups. AI then arrives tool by tool, a literature assistant in one team and a test-generation feature in another, with no shared view of which capabilities it touches.

Decisions about what AI may do in R&D have consequences, because a test conclusion can release or reject a design. Without a shared map, each team settles automation and oversight case by case, and nobody can say which capabilities are already covered, which carry high risk, or what order makes sense. The model gives those decisions one structure: one list of capabilities, one set of ratings and one ranking that can be reviewed and changed.

What's Inside

• 10 L1, 40 L2 and 158 L3 R&D capabilities, with definitions to L3 and 161 KPIs at L2
• A full AI profile for every L2 capability: opportunities, what stays human, agents, data, KPIs, AI measures, prerequisites, risks and workforce impact
• AI ratings for every L3: work pattern, target AI mode, AI fit and risk tier
• A prioritization model with a risk-adjusted score and rank for every L2 capability, using your own ratings and weights
• An AI heatmap of the whole model
• A 40-question weighted AI readiness assessment
• A Welcome document, the capability map deck, standalone capability definitions and KPIs documents, the Excel workbook, the user guide
• Bonus files: an R&D glossary, A Practical Guide to Business Architecture, a capability modeling overview, an AI-ready capability profile template and capability relationship mapping templates
• Free corrections, and new versions free for 180 days

Every file names your organization and carries your license ID. Ratings describe a typical large enterprise and are a starting point for your own.

Where It Is Used

AI opportunity assessment. Score the 40 L2 capabilities with your own ratings and weights and take the top of the ranking into discovery. The heatmap shows where high-fit, lower-risk work sits, so first pilots are chosen on evidence rather than on which team asked first.

Agreeing oversight before pilots start. Use the target AI mode and human oversight on each profile to agree, before anything is built, who approves what. For testing, that means deciding up front which conclusions a quality lead must sign and which a test lead can accept.

Application portfolio rationalization. Map lab, design, simulation and test applications to the L2 and L3 capabilities they support. Capabilities served by several tools, or by none, give a consolidation or investment case its structure.

Tool evaluation scoped by capability. Write requirements against named capabilities instead of a supplier's feature list. Evaluators score every candidate on the same capabilities, and the AI profile tells them which data and prerequisites the tool will depend on.

R&D operating model redesign. Use the capability definitions to assign ownership when research, product engineering and quality groups are reorganized. The workforce impact section of each profile shows which roles change as AI takes on drafting and analysis.

Performance baseline for AI changes. Pick KPIs from the 161 at L2 to measure a capability before and after an AI change. A pilot is then judged on the capability it was meant to improve, not on usage of the tool.

Who Works With It

Business and enterprise architects. Use the model as the reference structure for R&D, tailor names and definitions to the organization, and connect capabilities to applications, data and projects.

Head of R&D or chief technology officer. Read the ranking and heatmap to decide where AI goes first and which capabilities need investment in data or skills before it can.

R&D portfolio and program managers. Tie projects and AI initiatives to the capabilities they change, so the portfolio shows coverage and duplication.

Quality and test leads. Review the what-stays-human statements and risk tiers for testing and validation, and confirm where human approval must remain.

AI program leads. Use the profiles, prerequisites and readiness assessment to plan pilots and the data work that has to come before them.

A Blank Slate, Consultants, or a Reference Model

A usable R&D capability model needs more than a list of headings. It needs definitions to L3, KPIs, consistent naming and, for AI decisions, a profile and rating on each capability that the people running the work accept.

• Building it from a blank slate. Your own team runs workshops with research, engineering and quality leads, drafts the structure, argues over boundaries and writes definitions. The result is a model your people own, but much of the effort goes into generic structure every R&D function shares, and the AI ratings come later, if at all.
• Hiring consultants to build it. A firm brings method and facilitation and can move faster than an internal team. You still pay for elicitation and review cycles on structure common to any R&D function, and the deliverable may follow the firm's house format rather than one your tools can import.
• Starting from this model. You begin with the structure, definitions, KPIs and AI profiles already consistent with each other, and spend workshops on what is specific to you: renaming, merging, removing and re-rating. A first draft exists the day the files arrive.

Internal architects and outside advisers are best used for tailoring: testing the ratings against your data, your risk appetite and your operating model. The generic structure is the part least worth paying anyone to draft again.

First Steps

1. Read the Welcome document and the user guide to see how the files fit together.
2. Walk the capability map deck with R&D leads and mark capabilities to rename, merge or remove.
3. Replace the default ratings in the workbook for the capabilities you care about most, and set your weights.
4. Run the 40-question readiness assessment with the people who own data, technology and risk.
5. Take the top of the ranking and the matching AI profiles into a pilot shortlist.
6. Load the model into your architecture repository so projects and applications can be mapped to it.

Read the Public Sample

The sample follows one thread through the package, testing, and is 0.79% of the content. From the Testing (RD-5.2) AI profile:

• Work patterns: Analysis & insight, Content generation, Transactional processing
• Target AI mode: Augment
• Human oversight: Human-in-the-loop
• AI fit / Value / Feasibility / Data readiness: 4 / 4 / 3 / 3 (1–5)
• Risk tier: High
• Roadmap wave: Wave 2
• AI opportunity: Draft test plans from the requirements, design risks and earlier test results: methods, conditions, sample sizes, acceptance criteria and the trace link from each test to the requirement it covers
• What stays human: The Test and Validation Lead approves each test plan and its acceptance criteria, and accepts test results that do not bear on safety or quality; the Quality Lead approves every safety- or quality-relevant test conclusion that releases or rejects a design

Upgrade Credit

Start here and move up later: the full price of this edition counts toward the AI Transformation Blueprint for Research and Development if you upgrade within 90 days. Email support@ciopages.com with your order number for a one-time checkout code.

Frequently Asked Questions

What is the difference between the two editions?

The AI-Ready Capability Model covers the capability map, an AI profile for every L2 capability, prioritization, the heatmap and the readiness assessment. The AI Transformation Blueprint adds value streams redesigned for AI, the business data model, agent specifications, use cases, responsible AI principles and controls, the operating model and the roadmap, a leadership presentation template, and bonus files on roles, operating model archetypes and transformation.

Can I upgrade from the Capability Model to the Blueprint later?

Yes. The full price you paid for the Capability Model counts toward the Blueprint if you upgrade within 90 days. Email support@ciopages.com with your order number for a one-time checkout code.

What happens after 180 days?

Your license to use the files does not end. Corrections are free for as long as you use the files. New versions released within 180 days of purchase are free: we email you a new download link when one is released. There is no renewal after that: you can buy a new version at 25% off its price.

Can I load the content into our own AI tools?

Yes, within the license. The Enterprise license allows internal tools, including your own assistants and agents. Neither license allows training, fine-tuning or prompting AI systems for third parties.

Does the package cover AI laws and regulations?

No, by design. It is business architecture: capabilities, value streams, data and the AI layer describe what R&D does and how AI fits, and they hold whichever laws apply where you operate. The risk tiers, principles and controls are durable practices your compliance team maps to your own jurisdictions.

Why not just ask ChatGPT or another AI assistant?

An AI assistant will produce a credible list of R&D capabilities in a minute, and for getting oriented that is often enough. The trouble starts when the list has to support a decision. Ask twice and the structure changes; nothing checks that a capability, its ratings and the readiness questions refer to each other correctly; and the suggestions come without a risk tier, an oversight level or a statement of what stays human. This package is a fixed model with stable IDs, validated cross-references, a rated AI profile for each of the 40 L2 capabilities, a risk tier on each, and a scoring model you can weight and defend in a review. Your assistant is still useful on top of it, for drafting and exploring.