Overview
Ground Control AI automatically locates and tags GCP targets in your map images. This feature is enabled by default, and you do not need to take additional steps to use it. If you use a non-conventional GCP, you can still manually tag your points.
How to use the tagging workflow
After you submit a map for processing, DroneDeploy pre-processes the data. Once pre-processing is complete, you will receive an email notification. By default, both the plan creator and the data uploader receive these emails.
Open the email notification from DroneDeploy.
Click the link in the email to review the tagged GCPs.
Verify that the crosshairs align to the center of the target.
If needed, fine-tune the position manually by clicking the center precisely. DroneDeploy recommends at least 7 tags per GCP.
Click Submit GCP at the top right once satisfied with your tags.
Click Preview to review all tags one final time.
Click Continue to Processing in the top right corner to finish the map.
Key Features
Ground Control AI: Automates the identification of standard checkerboard targets to reduce manual labor.
Dual Views: Review points using both the Map view and List view for better context.
Point Switching: Change a point type from a GCP to a checkpoint directly within the tagging interface.
Limitations
Target Requirements: You must use standard, high-contrast checkerboard targets for the system to reliably tag points.
Processing Time: Most maps finish GCP AI processing in less than 90 minutes, though maps with over 2,500 images may take longer.
FAQs
Q: Who receives the GCP tagging emails?
A: By default, both the plan creator and the data uploader receive these emails. You can enable notifications for all project coordinators by contacting support@dronedeploy.com.
Q: What happens if I have a CSV error?
A: If your GCP CSV file contains an error, correct the file and re-upload it to the project.
Q: Can I skip a GCP?
A: Yes. If a target is no longer visible or obstructed, you can select the option to skip that GCP.
Video Demonstration (pre-January 2026 Release cycle)
v2.2