How ArgosAI’s camera-based detection works

ArgosAI develops innovative technologies to automate airside operations at airports. ArgosAI’s portfolio of computer vision and artificial intelligence-based technologies enables airports to accelerate their digital transformation as they manage one of the world’s most complex and safety-critical operations.

How it works

Artificial intelligence

Our patented AI algorithms analyse camera imagery to detect FOD and other configured airside anomalies. The images and detection results are available for operator review.

Hardware

Each camera repeatedly scans an assigned area. Station locations and camera coverage are designed for the airport’s required monitoring area and scan cycle. Increasing the number of stations can reduce the area assigned to each camera and shorten the scan cycle.

Software

Our software manages camera operation, AI modules, data storage and user access. It monitors system components and reports faults. Integration requirements and interface compatibility are assessed for each project.

User interface

We have developed easy to use and still easy to customize web based interactive user interface modules.

Compare our systems

A-FODA-PACA-PASS
Main applicationFOD detection on runways, taxiways and other large airside surfacesAircraft stand checks before arrivalAdditional airside inspection and supervision
Automated detection and checksForeign object debrisFOD, boarding bridge position, vehicles, equipment and height limitsBroken lights and birds/wildlife, depending on the selected modules
Operational supportObject location, size, detailed imagery and live videoAutomatic stand-clearance updates to VDGSRemote inspection, recorded imagery and operational analytics

For automated FOD and stand-clearance checks before aircraft arrival, explore A-PAC — Automated Pre-Arrival Control.

Explore A-FOD. Explore A-PAC stand checks.

A-FOD performance and verification

Detection performance is specified for each installation, including the target objects, monitored area, maximum scan cycle, visibility requirements and false positive rates.

A-FOD generates an average of 1.5–2.7 false positive alerts per runway per hour. Operators review each alert using live video and recorded imagery, typically in around 15 seconds. Almost all false positives are identified and dismissed remotely, without dispatching personnel onto the runway or closing it.

Performance is verified through airport acceptance testing. Tests follow the applicable standards and can be extended to address the airport’s operational requirements.

FAQ

How do A-FOD, A-PAC and A-PASS differ?

A-FOD detects FOD on runways, taxiways and other large airside surfaces. A-PAC automates aircraft stand checks before arrival and sends stand-clearance status to the integrated VDGS. A-PASS provides additional airside functions, including remote inspection, bird and wildlife detection, and operational analytics.

What information does an A-FOD alert provide?

Each A-FOD alert includes the detected object’s location, size, detailed images and live video. Operators can also review recorded imagery to assess the alert and decide whether action on site is needed.

How quickly does A-FOD detect newly appearing debris?

A-FOD’s cameras repeatedly scan their assigned areas. Typical installations are designed for maximum scan cycles of 2 minutes during the day and 4 minutes at night. Debris may be detected earlier within the cycle; these values are not fixed waiting periods before detection.

The scan duration depends on the area assigned to each camera. More sensor stations can reduce that area and shorten the scan cycle.

Does A-FOD operate at night?

Yes. A-FOD operates during daylight and at night. The maximum scan cycle for each condition is defined by the installation design.

What visibility does A-FOD require?

The visibility requirement depends on the maximum camera observation distance in each installation. A-FOD maintains the specified detection performance while visibility remains at or above that distance.

For example, if the furthest observation point is 230 metres from a camera, the installation’s visibility requirement is 230 metres. Detection performance may degrade when visibility falls below this threshold.

How are false positive alerts handled?

A-FOD generates an average of 1.5–2.7 false positive alerts per runway per hour. Operators review each alert using live video and recorded imagery, typically in around 15 seconds. Almost all false positives are identified and dismissed remotely, without dispatching personnel onto the runway or closing it.

Can A-PAC integrate with our VDGS?

A-PAC integrates with most VDGS systems and automatically communicates stand-clearance status. Additional integrations can be developed as required.

Can A-PAC use existing cameras?

A-PAC can reuse existing cameras when they meet ArgosAI’s requirements. It uses a wide-angle camera and a pan-tilt-zoom (PTZ) camera. Camera suitability is assessed for the stand and the checks required.

What infrastructure is required?

ArgosAI supports three deployment options: on-premises servers, edge GPU computers installed in cabinets near the cameras, or the cloud. Infrastructure requirements depend on the airport’s preferred option and operational needs.

Which standards does A-FOD meet, and how is performance verified?

A-FOD meets the requirements of FAA AC 150/5220-24 and EUROCAE ED-235A. Airports verify performance through acceptance tests using the specified test objects and criteria. Additional tests can be performed to address the airport’s operational requirements.