Prevent accidents or equipment failure with Spectro-AIForeign Object Debris Detection
Detect FOD with drone or security-camera imagery. Deploy the same use case through SAI-HUB for on-premises operations or as DJI On-Board AI on supported enterprise aircraft.
Automated FOD Detection for Runways, Aprons and Airfield Surfaces
Foreign Object Debris can be small, irregular and easy to miss during conventional inspections. SpectroAI has trained a dedicated AI algorithm to detect a range of FOD classes directly from operational imagery.
- Detection-focused AI: trained to identify common airfield debris and surface-condition classes.
- Multiple image sources: run the use case from supported DJI drone workflows or fixed security cameras through SAI-HUB.
- Two deployment options: process on your own Brain-Box with SAI-HUB, or run the algorithm directly on supported DJI aircraft with DJI On-Board AI.
- Visual verification: detected objects can be reviewed in imagery instead of relying on a text-only alert.
Watch the FOD Detection Showcase
This short video shows the FOD use case in practice. Watch it to see how detections are surfaced and how the workflow can be applied to real airfield inspections.
Prefer YouTube? Open the full video in a new tab.
Watch on YouTubeDeploy FOD Detection Where It Fits Your Operation
Choose between a locally processed SAI-HUB deployment or DJI On-Board AI on supported enterprise aircraft.
SAI-HUB
SAI-HUB is our software deployment for DJI Dock 2/3, DJI Enterprise drones and security cameras. The solution runs on premises, with data processed and stored locally on the Brain-Box.
View SAI-HUB ClassesDJI On-Board AI
Run the FOD algorithm directly on supported DJI enterprise platforms: Matrice 4E, 4T, 4D, 4DT and Matrice 400 with H30/H30T payloads.
View On-Board ClassesFOD Detection with SAI-HUB
Use existing airfield camera infrastructure or DJI Enterprise drone workflows while keeping processing and storage on site. The Brain-Box runs the detection workload locally, making SAI-HUB suitable for operations that require an on-premises architecture.
- Runs with DJI Dock 2/3 deployments, DJI Enterprise drones or security-camera feeds.
- All data is processed and stored on the Brain-Box.
- One deployment covers both debris classes and surface-condition classes such as cracks and repaired cracks.
20 Trained Classes
Run the FOD Algorithm Directly on the Drone
DJI On-Board AI brings the FOD model onto supported Matrice platforms, allowing detections to happen at the edge during the mission.
- Purpose-built class set for airborne FOD inspections.
- Processing happens on the supported DJI platform rather than on a separate local server.
- Designed for teams that want the detection model embedded directly into the aircraft workflow.
10 FOD Classes
From Inspection Route to Detection Review
The FOD model is designed to slot into repeatable inspection workflows, whether imagery comes from a scheduled drone mission or fixed cameras.
Capture
Collect runway, taxiway, apron or other airfield imagery with a supported drone or camera source.
Process
Run detections either locally through SAI-HUB and the Brain-Box or directly through DJI On-Board AI.
Detect
The trained model identifies supported FOD classes and highlights objects in the imagery.
Review
Operators can visually verify detections and use the result as input for inspection and removal workflows.
Choose the Architecture That Matches Your Airfield
SAI-HUB
DJI On-Board AI
Bring Automated FOD Detection Into Your Inspection Workflow
Tell us how you inspect your airfield today. We can help you choose between SAI-HUB on-premises processing and DJI On-Board AI for supported aircraft.