On a utility-scale solar park, the question "where are we?" has thousands of small answers. Every mounting table has posts, a substructure and panels, and each of them moves through its own status. Most site teams still collect this on paper or in a shared spreadsheet and turn it into a report by hand. This guide shows how to automate that chain from the field to a report that is ready every morning.
Why manual progress reporting breaks at utility scale
A 90 MW park can have more than 4,000 mounting tables and over 140,000 panels. When foremen write status on paper and someone types it into Excel at night, three things happen: numbers arrive a day late, typing errors slip into totals, and the person doing the typing becomes a bottleneck. Reports to the client or investor are then built on data nobody fully trusts.
Step 1: define what you capture per table
Automation starts with a clear data model. For each table, decide which fields are mandatory and which values are allowed. A proven structure:
- Table ID and zone, unique and pre-loaded, so crews choose instead of type.
- Post status for every post position, with a fixed list such as rammed, drilled or ballasted.
- Substructure assembled yes or no.
- Panels mounted, with panel type (for example 615 W or 620 W).
- Position, perimeter or central, if it affects workload or reporting.
Step 2: capture offline, sync when possible
Solar sites are rarely covered by stable mobile internet. Choose tablet forms that store records locally and upload automatically once a device reaches Wi-Fi, for example at the site office. Let each zone sync on its own and make uploads resumable, so one failed transfer does not block the rest.
Step 3: validate before anything is counted
The biggest gain from automation is not speed, it is trust. Run these checks on every incoming record before it reaches a total:
- Missing post entries on a table marked as complete.
- Panel counts above the physical maximum of the table.
- Duplicate table IDs from two devices.
- Status regressions, such as a post going from rammed back to empty.
Flag problems to the zone lead the same day. Fixing one wrong entry on Tuesday is easy; finding it in a monthly total is not.
Step 4: fill the Excel your team already uses
Many EPCs have a progress workbook that managers, planners and clients know well. Replacing it creates resistance. A better approach is to let the automation write into it: per-zone sheets get updated rows, and summary sheets get totals and percentages. A workflow tool such as n8n combined with a small Python script can update even macro-enabled .xlsm files reliably.
Step 5: publish a daily report and a live dashboard
Once data is clean and aggregated, reporting is the easy part. Two outputs cover most needs:
- A daily PDF report sent automatically early in the morning, with yesterday's progress, active and idle zones and the change in every key metric.
- A live dashboard with overall and per-zone progress, daily throughput and a projected completion date.
Add alerts for milestones and for days with zero progress in a zone. They often reveal material or crew problems faster than any meeting.
Build or buy?
Dedicated PV construction platforms exist and suit large EPCs that want one system for tasks, defects and documentation. A custom automation is the better fit when you want to keep your existing workbook and reporting format, need a specific data model, or want to start small for a single park. Both approaches rely on the same foundation: structured capture, validation and automatic aggregation.
Checklist
- Write down the fields and allowed values per table.
- Pre-load all table IDs per zone.
- Pick an offline-capable form tool and test it on site.
- Define validation rules and who receives the error alerts.
- Connect the aggregation to your existing workbook.
- Schedule the daily report and set up milestone alerts.
We built exactly this chain for an 89.4 MW park; you can read how in our Alpin Sun case study.
Not sure which process in your company to start with? Here is how to find it: automate business processes, where to start.