How to Specify a UAV Building Inspection: Procurement Guide

Before you write the specification

Three questions will shape the entire procurement. What do you need to know? You are not procuring 'drone inspections.' You are procuring answers to specific questions about your buildings.

What buildings are included? Define the portfolio clearly. For each building, the provider will need address, building type, number of storeys, roof type, specific areas of concern, and access restrictions.

What will you do with the output? If you need the output for a CAPEX model, the specification must require costed, graded findings. If for regulatory evidence, it must require dated, located, verifiable records.

Scope of inspection

Define exactly what is inspected. Options typically include roof zones, facades, drainage, plant and equipment, and ground-level features relevant to the building envelope.

Be specific about exclusions. If you do not need internal common areas inspected, say so. Ambiguity in scope creates ambiguity in pricing.

Capture methodology

Specify minimum capability. Visual imagery: minimum 20MP, ideally 48MP+. Thermal imaging: if required, specify radiometric thermal capture with temperature data per pixel, not just a colour palette image.

For roof areas, require complete coverage with sufficient overlap for detail inspection. Random sample photographs are not an inspection.

Regulatory and operational compliance

Your specification should require CAA operator registration with a valid Operator ID, appropriate operational authorisation for the work proposed, minimum £1m public liability insurance for drone operations (£5m to £10m for institutional property), site-specific risk assessment and method statement, and pilot competence including a valid GVC or equivalent.

Do not accept 'CAA approved' as sufficient evidence. CAA does not approve or endorse individual companies. It issues operational authorisations.

Analysis and verification methodology

This is where specifications most commonly fail. Specify defect taxonomy, severity grading, location referencing, human verification of findings before delivery, cost estimation basis, and RICS alignment if required.

Automated AI detection can identify defects at scale, but unverified automated output contains false positives. The specification should require professional review of all findings before they are delivered as the formal record.

Deliverables and data handling

At minimum: building-level inspection reports, portfolio summary with comparative analysis, digital imagery archive, and specified data format. Clarify whether you expect ongoing access to the data after the project.

For institutional clients, confirm data ownership, storage location, retention and deletion policy, encryption requirements, and any compliance standards such as Cyber Essentials or ISO 27001.

Evaluating proposals

Look for inspection experience, not just flying experience. A company that has flown 10,000 flights but never analysed a building envelope is a data capture provider, not an inspection provider.

Ask who reviews the imagery and produces the findings. Compare on total delivered value, not unit capture cost. A provider quoting £200 per building for visual-only capture is not comparable to one quoting £600 for visual, thermal, AI-assisted analysis, professionally verified findings and platform access.

Common procurement mistakes

Specifying by the hour instead of by the outcome. Buying '20 hours of drone time' does not guarantee useful output.

Evaluating on price alone. Omitting analysis requirements: if the specification only covers the flight, you will receive photographs, not intelligence.

Assuming all drone companies are the same. The market ranges from sole-operator pilots with consumer equipment to full-service building intelligence providers with integrated analysis, reporting and platform capabilities.