Photo-first vs clipboard: a better field data collection workflow for electrical systems

Why capturing structured data in the field beats reconstructing it at your desk. Written by a licensed PE.

If you have ever walked an electrical room with a clipboard in one hand and your phone in the other, you know the clipboard-and-camera workflow. Hundreds of nameplate photos, a sheet of legal pad covered in identifier numbers and tiny pencil notes, and three days back at the office trying to reassemble what you saw into a coherent one-line and an equipment list that is actually usable.

It works. I did it this way for years. But the reconstruction step at the desk is where most of the pain lives. By the time you are looking at a photo of a transformer nameplate five days after the site visit, you cannot remember if the 480V panel fed from breaker 12 was the one near the server room or the one next to the elevator machine room. You guess. Sometimes you guess wrong.

This post compares the old clipboard-and-camera workflow with a photo-first workflow that captures structured data per piece of equipment in the field, and shows how to run the new workflow even without a purpose-built tool.

The old workflow, step by step

1. Show up with a clipboard, a sharpie, a phone, and a stack of labels.
2. For each piece of equipment: write an identifier on a label, stick it on the nameplate, take a photo of the nameplate with the label in frame, and write a line on the clipboard with the identifier, equipment type, and any note you think you will need later.
3. Back at the office: dump all photos into a folder, sort by identifier, open a fresh spreadsheet, type every field from every nameplate photo into the sheet one at a time.
4. Build the one-line from the spreadsheet, trying to remember which panels fed which.
5. Produce the deliverable three to ten business days after the site visit.

Steps 3 and 4 are where the workflow breaks down. Transcribing 200 to 500 nameplate photos into a spreadsheet takes hours. Guessing connections between equipment you did not explicitly map on-site takes more hours. And if any field on the spreadsheet is ambiguous, you either go back to the photo, call the client, or drive back to the site.

What goes wrong, specifically

A few things reliably fail on the clipboard-and-camera workflow:

• Photos that do not match labels. Somewhere between taking the photo and looking at it three days later, the label fell off, the photo got mis-named, or your handwriting on the clipboard became unreadable.
• Connections that were not explicitly captured. You knew at the time that Panel LP-3 fed from the 60A breaker in the MCC, but nothing on the clipboard captured that cleanly and the photo of Panel LP-3's nameplate does not show its upstream source.
• Missed fields. Interrupting rating, percent impedance, locked rotor code. The photo sometimes does not show them clearly. You would have zoomed in on the nameplate if you had known at the time.
• Feeder lengths. A clipboard note says "~100 ft." It was actually 147 feet.
• Ambiguous equipment naming. Panel A and Panel A are two different panels in two different buildings, but you named both the same on the clipboard and now you are not sure which photo belongs to which.

None of these are fatal individually. Collectively they add hours of rework per site visit.

The photo-first alternative

A photo-first workflow flips the order of operations. Instead of taking a nameplate photo and writing a label ID on a clipboard, you capture the structured equipment data on your phone at the point of capture. The photo is the anchor, but the fields next to it are what the office workflow actually consumes.

The steps look like this:

1. For each piece of equipment, open a structured form on your phone (or a row in a template spreadsheet, or a tool built for this).
2. Snap the nameplate photo.
3. Fill in the fields for that equipment type right there, in the field, while you are standing in front of the equipment: bus rating, voltage, trip ratings, interrupting ratings, feeder conductor size and length to the upstream bus, connected downstream loads.
4. Explicitly record the upstream source (which breaker on which panel feeds this equipment).
5. Move to the next piece of equipment.
6. Back at the office: export the structured data to whatever tool you use next (spreadsheet, SKM, ETAP, EasyPower, whatever). Zero transcription. Zero guessing about connections.

The extra time per piece of equipment is maybe 30 seconds. The time saved at the desk is hours per site visit.

Running a photo-first workflow without a dedicated tool

You can approximate the photo-first workflow with Google Forms, Airtable, or a well-structured spreadsheet on your phone. The key is that the schema must be defined before you show up on-site, and you must fill it out in the field, not afterward.

A minimum viable schema looks like this (one row per piece of equipment):

• Equipment ID (e.g., LP-3)
• Equipment type (panel, switchgear, transformer, motor, generator, transfer switch, disconnect)
• Nameplate photo (attachment)
• Panel-door photo with schedule, if applicable (attachment)
• Voltage
• Bus rating or nameplate rating
• Main breaker trip + interrupting rating (if applicable)
• Upstream source equipment ID
• Upstream source breaker position (if applicable)
• Feeder conductor size
• Feeder conductor length (measured)
• Notes

For panels and switchgear specifically, you also need a nested table of every branch breaker (trip rating, frame size, AIC, poles, connected load). A flat spreadsheet makes this awkward. Airtable handles it with linked tables. A dedicated tool handles it natively.

Why the connection data is the real win

The single most valuable thing a photo-first workflow captures that clipboard-and-camera tends to miss is the connection graph. Each piece of equipment has an upstream source. If you write that down at the moment of capture, while you are standing in front of the equipment and can physically see where the feeder comes from, you end the walk with a complete connectivity map.

Without that map, you cannot build a one-line from the data. You can build an equipment list from photos, but the relationships between the pieces of equipment have to be reconstructed from memory, notes, and cross-referencing photos. That is the part that takes days and that is where errors creep in.

What this workflow delivers

If you execute a photo-first workflow well, you walk off the site with:

• A structured equipment list with every field you need for an arc flash study, panel schedule, upgrade design, or maintenance document
• A nameplate photo per piece of equipment, anchored to the equipment identifier
• Measured feeder lengths, not estimates
• An explicit upstream-source reference for every piece of equipment, which means the one-line builds itself without a reconstruction step
• Notes on spare breakers, phantom loads, and anything that did not match the existing documentation

The deliverable back to the client goes out the same week as the site visit, not three weeks later.

How I run this now

I spent a long time running the photo-first workflow manually in Airtable and eventually built AmpSketch specifically around this pattern. The tool is designed around the walk itself: open it on your phone at each piece of equipment, snap the nameplate, the AI reads the visible fields, you verify and fill anything missing, and the one-line builds on your screen as you go. Upstream connections are captured by design because the tool asks which breaker this equipment is fed from before saving.

A CSV export of the structured equipment list for hand-off to arc flash study software is on the roadmap, along with a more complete on-device arc flash workflow. The core capture and one-line build work today in any browser with no signup.