01 · Business objective · critical foundation Is there a specific operational or business problem that AI is expected to improve? AI is the starting point; the problem and affected users are not yet explicit A useful area is known, but the decision boundary or users are still broad The problem, affected users, decision boundary and consequence of doing nothing are explicit I do not have enough information to answer.
02 · Business objective Are the expected benefits measurable against a current baseline? Benefits are described mainly as innovation or efficiency Target benefits exist, but baseline, owner or review date is missing Baseline, target, measurement owner and review date are defined I do not have enough information to answer.
03 · Data · critical foundation Can the necessary building and operational data be accessed lawfully and reliably? Required data sources or access rights are unknown or unavailable Some BMS, IWMS, CAFM, IoT, occupancy or energy data is accessible, with material gaps Required sources, access rights, retention and accountable owners are documented I do not have enough information to answer.
04 · Data Is data quality understood well enough for the intended use case? Completeness, accuracy, timeliness and identifiers have not been checked Sample checks exist, but exceptions and remediation ownership are incomplete Representative quality tests, thresholds, exceptions and remediation owners are recorded I do not have enough information to answer.
05 · Integration · critical foundation Can the use case integrate with the systems and workflows it depends on? Interfaces, formats and workflow hand-offs are unknown Key integrations are identified, but feasibility or ownership is not proven Interfaces, data contracts, owners, failure handling and write-back boundaries are documented I do not have enough information to answer.
06 · Infrastructure Can current infrastructure support the required availability, latency and connectivity? Hosting, connectivity, capacity and resilience needs are not assessed A likely architecture exists, but site constraints or resilience remain assumptions Capacity, connectivity, latency, resilience, monitoring and recovery needs are evidenced I do not have enough information to answer.
07 · Security & governance · critical foundation Have cybersecurity, privacy and safety risks been assessed for this use case? No use-case-specific security, privacy or operational safety assessment exists Material risks are recognised, but controls or accountable acceptance are incomplete Threats, personal data, access, suppliers, safety impacts and accountable controls are documented I do not have enough information to answer.
08 · Security & governance · critical foundation Are human accountability and AI operating boundaries explicit? The system may influence actions without named human accountability or escalation An owner exists, but review, override, incident and change controls are incomplete Ownership, review, override, escalation, monitoring and change authority are defined I do not have enough information to answer.
09 · People & adoption Do teams have the skills to operate, challenge and maintain the capability? Required business, data, engineering, security and assurance roles are not identified Some skills exist, but material dependency on individuals or suppliers remains Required roles, capability gaps, training and supplier responsibilities are documented I do not have enough information to answer.
10 · People & adoption Is the operational workflow ready to use AI output responsibly? The output has no agreed place in day-to-day work A workflow is proposed, but user testing, exceptions or adoption ownership is incomplete Users, workflow, exceptions, training, adoption owner and feedback route are defined I do not have enough information to answer.
11 · Financial viability Does the business case include the full cost and a credible route to value? Costs, implementation effort or benefit assumptions are largely unknown A budget exists, but integration, data, assurance, operation or exit costs are incomplete Full lifecycle cost, benefit assumptions, sensitivities, funding and stop criteria are recorded I do not have enough information to answer.
12 · Evidence & assurance · critical foundation Can readiness and supplier claims be tested with attributable evidence? Readiness depends mainly on presentations, demos or untested assumptions Important claims have sources, but gaps, contradictions or acceptance tests remain Claims, sources, gaps, tests, owners and acceptance thresholds are traceable I do not have enough information to answer.
