Can Street-Level Imagery Replace Traditional Asset Inventories?

Can street-level imagery combined with computer vision and AI replace — or substantially reduce — traditional field-based municipal asset inventories?

Imagery and AI can substantially reduce field inventory work for visible, countable, above-ground assets. They are much less able to replace independent verification where dimensions, hidden attributes, safety, regulatory compliance or legal defensibility matter. Crowdsourced dashcam imagery, commercial mobile mapping and increasingly capable computer-vision systems now make a network-wide visual survey far cheaper and more frequent than a traditional periodic field programme — but the correct framing is re-tasking field staff to verification, not eliminating them. This review sets out what the accuracy and cost evidence supports, which asset classes are ready for automated inventory, and what to ask before procurement.