Every locate produces knowledge. For a few days it sits on the ground in paint and flags: where the gas main actually runs, where the duct bank swings wide of the as-built, where an unrecorded service crosses the boulevard.
Then the marks fade, the crew moves on, and most of that knowledge is gone. The next project pays to rediscover it. The difference between organizations that keep paying and those that stop is not better paint. It is a workflow that turns field evidence into a record — and a record into something operations can actually maintain.
Capture with the record in mind
A mark on the ground and a point in a geodatabase are different products, and the difference is decided at capture time. A position alone is not enough. The capture should carry the attributes the record will need later:
- what was found — utility type, size and material where observable;
- how it was found — the locate method and evidence source;
- when and where — date, location and the survey basis of the measurement.
Evidence source matters most. A record that remembers whether a position came from records research, a surface trace or a verified exposure keeps its meaning years later — the same logic that drives SUE quality levels.
Review before anything enters the geodatabase
Field data goes through quality review before it becomes part of the record. Positions are checked against the survey control, attributes against the field notes, and new features against what the record already claims. Where new evidence contradicts an old as-built, the conflict is resolved deliberately — not overwritten silently. A record that mixes reviewed and unreviewed data without distinction eventually earns the same distrust as no record at all.
Structure the geodatabase for maintenance
The record has to be built for the day someone updates it. In practice that means feature classes organized by utility system, consistent attribution across capture campaigns, and source and confidence fields on every feature. Structure is what lets the next locate, the next inspection or the next capital project add to the record instead of sitting beside it in another folder of drawings.
Publish where operations already work
A geodatabase no one can open changes nothing. Publishing the record through a Web-GIS puts it in front of the people who make daily decisions — operations, engineering, contractors — as layers they can query rather than drawings they have to request. Access, hosting, permissions and security are scoped per project, because every owner’s environment and obligations differ.
Keep an owner and an update path
Records decay the moment maintenance stops. Someone must own the dataset, and the update path must be defined: when new field evidence arrives, who reviews it, who commits it, and how often. A utility record with a named owner and a working update path keeps earning trust. Without them, it is only a snapshot of the day the project ended — and the next locate starts from zero again.
