Data Input Requirements for Solar Thermography Pro

This article will cover the data input requirements for the Solar Thermography Pro data product

Updated over a week ago

Introduction

Making sure that the results we deliver to our customers are the quality they expect start with how the data is acquired. Making sure that we have all of the data that we need in, delivered in the way we expect will make sure processing runs as smoothly as possible.

However, capturing these thousands of images might be challenging. Therefore, we created this article, considering the best practices to save you time and frustration during the data acquisition.

This article will cover the following requirements for the Solar Thermography Pro data product:

  • Flight plans (RGB & Thermal)

  • Tips depending on your site type

  • Site As-built Document for sites larger than 5ha


Video Guide


Flight plan

We require visual (RGB) and thermal photos for the Solar Thermography Pro data product. A drone pilot will use an unmanned aerial system (a drone with a camera) to capture the data on-site.

Visual (RGB) Data Requirements

  • Ground Sample Distance (GSD): 1.0 - 1.5 cm per pixel at ground level.

  • Front Overlap: 75%.

  • Side Overlap: 75%.

  • Flight Line Direction: Along the long edge of the solar rows.

  • Flight Path: At least 2 flight lines should be outside the boundaries of the operation in all directions.

  • Gimbal Orientation: standard Nadir (-90 deg), with acceptable deviation up to 20 deg to avoid sun glare.

  • Image Format: JPEG.
    Each image should contain the following metadata: GPS location, relative altitude, gimbal pitch, gimbal yaw, gimbal roll, and local timestamp.

  • Image Quality: in focus, free from motion blur and minimal glare.

Thermal Data Requirements

  • Ground Sample Distance (GSD): 5.5 cm per pixel at the panel level.

  • Front Overlap: 75%.

  • Side Overlap: 75%.

  • Image Resolution: 640x512 pixels or greater.

  • Flight Line Direction: Along the long edge of the solar rows.

  • Flight Path: At least 2 flight lines should be outside the boundaries of the operation in all directions.

  • Gimbal Orientation: standard Nadir (-90 deg), with acceptable deviation up to 20 deg to avoid sun glare.

  • Image format: R-JPEG or Radiometric TIF.
    Each image should contain the following metadata: GPS location, relative altitude, gimbal pitch, gimbal yaw, gimbal roll, and local timestamp.

  • Image Quality: in focus, free from motion blur and minimal glare.

  • Environmental Conditions: Aim for a clear blue sky. If a cloud passes, wait 10-15 minutes for the panels to heat up again. Successive flights should not exceed 15 minutes unless it is a new flight block. Wind speed below five m/s, humidity ideally less than 60%.

  • Irradiance: β‰₯ 600 Watt/sqm

Special Considerations for Single-Axis Trackers

Trackers should be as horizontal as possible, up to a maximum 50-degree tilt. Avoid flying in the early morning or late afternoon when the sun is low at the horizon to minimise inter-row glare.

Special Considerations for Double-Axis Trackers

Trackers should be as horizontal as possible, up to a maximum 50-degree tilt. The entire tracker should be visible on at least one photo. Otherwise, increase the altitude.

Example of image alignment

Example of image alignment

Rooftops

Flying a rooftop? Check out this support article on correctly adapting your flight plan to fly rooftops.

Blocks

  • To ensure weather conditions are kept homogeneous across the entire solar inspection, sites larger than 5ha should always be divided into blocks.

  • To ensure a margin of error lower than 10% for the Ground Sample Distance [in cm/pixel], sites with elevation differences greater than 10 meters should always be divided into blocks. Also, you need to take off from the highest point within the block.

  • Blocks should be flown as separate flights from each other but as much as possible one block should contain one flight. If you need to switch batteries, there should be minimum downtime for the flights in the same block.

The Sitemark platform does not let you cut sites into blocks. Reach out to our support team to let them know, and they will cut the site into blocks on your behalf.


Site As-Built

For operations larger than 5 ha, an up-to-date and accurately geo-referenced as-built plan from your solar site should be uploaded on the Sitemark platform as a site overlay. This is crucial to ensure all of the geospatial data is mapped correctly. You must do this before the confirmation of the operation.

You can use this support article if you have not uploaded your site documents before to help you with this.

The site as-built needs to meet the following requirements:

  • It should be recent to match the site as closely as possible.

  • It must show clear physical reference points, such as roads, buildings, rivers, tables, panels, etc..

  • It should be accurately scaled so it can be geo-referenced on the map.

  • It should be uploaded in PDF or DXF as an overlay in Sitemark.

  • It should be uploaded before the confirmation of the operation.
    It is important that, if there is only a PDF, some site reference points on the as-built must be visible on the satellite image in the platform. So it can be accurately geo-referenced on the map.

  • It should cover all areas that need analysis because areas not covered in the as-built might not be analyzed.

  • It should not be hand-drawn or low-quality scanned site plans.

Examples of an acceptable as-built vs not acceptable

βœ… Good - every module is clearly visible

Example of a good quality as-built plan

❌ Not good - modules not visible

Example of a bad quality as-built plan

❌ Not Good - the resolution is too low, meaning we can not use any reference point from the as-built to locate on the map.

Another example of a bad quality as-built plan

If you do not have an as-built from your site available, we strongly recommend you use the Solar Site As-built data product from Sitemark.

Please contact your account manager or our support team for more information about this data product.

Did this answer your question?