Property Imagery Assessment

There is no property imagery assessment. Underwriters rely on the applicant's self-reported property description. When an underwriter wonders 'what does this building actually look like?' they either drive by personally or take the agent's word for it. Roof condition, building maintenance, and surrounding hazards are unknown until someone physically visits.

None — AI cannot assess property condition because no imagery or assessment records exist.

Property photos exist but are ad hoc. Some agents submit exterior photos with applications. The occasional loss control inspection includes building photographs. Images sit as email attachments or in a shared drive folder labeled by address. There is no consistent image collection process and no standard for what to photograph. Quality varies wildly — some photos are clear exterior shots, others are blurry images taken from a car window.

AI could potentially identify obvious property characteristics (building size, roof type) from some photos, but the inconsistent quality, angles, and coverage make systematic analysis unreliable.

Property imagery follows a standard collection protocol. Aerial imagery from commercial providers (Nearmap, EagleView) is purchased for property submissions above a coverage threshold. Required views and minimum resolution standards are documented. Images are stored in the underwriting workbench linked to the application. But the imagery assessment is still manual — underwriters eyeball the photos and note observations in free-text fields.

AI can perform basic computer vision analysis — identifying roof material, building footprint, and obvious hazards. Cannot produce structured assessments because there is no defined assessment schema to populate.

Property imagery assessments are structured records with defined scoring criteria. Each assessment rates roof condition (1-5 scale), building maintenance, vegetation clearance, and surrounding hazard proximity using standardized rubrics. Assessment records link to the source imagery and the application. An underwriting manager can query 'show me all assessed properties in hurricane-prone zones with roof condition scores below 3' and get an instant answer.

AI can perform automated property assessments following the defined rubric, flag high-risk properties, and integrate assessment scores into the overall risk score calculation.

Property imagery assessments are schema-driven with a formal property feature taxonomy. Each assessment links to geospatial hazard databases, municipal building records, and historical imagery comparisons. An AI agent can query 'compare the current roof condition to the assessment from two years ago and flag properties where deterioration exceeds the threshold for this coverage type.' The assessment is a rich, multi-layered record.

AI can perform fully autonomous property assessments with temporal comparison, hazard correlation, and risk scoring integration. Human review limited to edge cases and disputed assessments.

Property imagery assessments update continuously from real-time sources. Satellite imagery refreshes monthly. Drone surveys trigger automatically after catastrophe events. IoT sensors on monitored properties stream structural condition indicators. The assessment record is always current — there is no concept of a 'stale' property assessment because monitoring is continuous.

Fully autonomous property assessment. AI monitors, assesses, and updates property condition records continuously without human intervention.

Ceiling of the CMC framework for this dimension.