Lessons from planning fiber and SCADA connectivity across 20+ AEP energy stations in the Indiana–Michigan region.
The seductive math of the desk
From the desk, a fiber run between two substations is a straight line in Google Earth and a clean length calculation in a spreadsheet. From the field, it's a railroad crossing that requires a permit no one budgeted for, a road widening that moved a handhole, and a customer's fence that wasn't there on the last aerial image.
Every hour spent at the desk in place of an hour in the field feels efficient. It isn't. Desk-verified designs generate rework; field-verified designs generate revisions before construction starts — a much cheaper failure mode.
What field verification actually catches
- Pole ownership and make-ready costs that never appear in GIS layers.
- Existing attachments (power, comms, streetlight) that constrain the sag and clearance envelope.
- Access constraints — locked gates, seasonal roads, active construction, environmental buffers.
- Local jurisdiction quirks: which township wants a bond, which county wants engineered plans, which railroad requires the specific insurance rider.
- The delta between the as-built and the map. It is almost always larger than the model assumes.
The habit that scaled
For AEP telecom planning across 20+ energy stations, the working rule became: no design package leaves engineering without a site survey signed by the engineer who drew it. Not a subcontractor's photo dump — the actual designer, on the actual site, with the drawing in hand.
It slowed the front end by days. It removed weeks from the back end. That trade is almost always the right one.
The one-line rule
If you can't defend the design from the ground it will sit on, it isn't a design yet — it's a hypothesis.