REACHbyMortoff

Case study · AI Vision – Health & Safety

Aircraft Proximity Monitoring – automated no-touch monitoring

From sample-based audits to continuous, video-evidenced monitoring of an aviation safety rule – pilot project.

Customer
International airport ground handling company
Industry
Airport ground handling
10 cmno-touch rule, continuously monitored
6–8ground handling units in the pilot
~1 minutevideo clip for every event

In brief

Challenge

Compliance with the 10 cm no-touch rule could only be checked by sample-based audits, without video evidence.

Solution

Proximity sensing on ground handling equipment, automatic camera selection, event-triggered video clip, AI anonymisation and a REACH incident interface.

Result

The pilot proved that compliance with the rule can be monitored continuously, objectively and in an auditable way.

Starting point

At an international airport ground handling company, the no-touch policy stipulates that no ground handling equipment may come closer than 10 cm to the body of the aircraft. A breach of the rule poses aviation safety, financial and contractual risks at the same time.

  • Aviation safety risk: every day the equipment comes extremely close to aircraft.
  • Financial risk: repairing damage caused by negligence can cost several hundred thousand euros.
  • Contractual risk: the aircraft belong to the airlines, and the contracts strictly regulate damage and SLA breaches.

Compliance with the rule could previously only be checked by sample-based audits, without video evidence.

What was not visible

The sample-based audit drew conclusions about the whole operation from a few turnarounds. Minor contacts could go unnoticed, and there was no objective video evidence for reconstruction.

Unanswered questionAnswered today by
How many no-touch violations actually occurred?Automatic detection – every event
Which ground handling unit was involved?REACH incident – unit involved
When and under what circumstances did it happen?Video clip before and after
Did the event cause actual damage?Objective video evidence
Are there recurring risk patterns?Search and analysis

Implemented solution

The goal was to check and document compliance with the rule continuously rather than by sampling. The solution consists of five components:

  1. Proximity sensing: generates an event within 10 cm.
  2. Camera selection: the image of the most relevant camera.
  3. Video clip: 30 seconds before, 30 seconds after.
  4. AI anonymisation: data protection and auditability.
  5. REACH incident: monitoring and audit.

Detection and camera selection

Proximity sensors mounted on the ground handling equipment continuously monitor the distance. Within 10 cm, the edge IoT gateway automatically generates an event, which is sent encrypted to the REACH server over a private APN. Cameras are mounted on the vehicles – including passenger stairs and ground power units (GPUs): REACH selects the most relevant one and automatically requests the footage.

Video clip and AI anonymisation

For every detected no-touch violation, a video clip of roughly one minute is created: 30 seconds before the event and 30 seconds after. The clip is uploaded to central Amazon cloud storage, where AI anonymises privacy-sensitive visual information while preserving auditability.

REACH incident monitoring and audit

In the REACH interface every event is available, together with its video, in a single click:

  • identifying the event: time and unit involved
  • replaying the video before and after the incident
  • assessing and auditing the event
  • flagging a false detection
  • searching and analysing past events

Results

With 6–8 ground handling units, the pilot proved that the no-touch policy can be monitored continuously with technology. Previously, conclusions about the whole operation were drawn from observing a few turnarounds; now every detected proximity event is recorded automatically as a structured incident.

  • Faster investigation: not just an alert, but objective video evidence of what happened before and after.
  • Fewer hidden events: clear unit, time and context – no violation, however minor, goes undocumented.
  • Preventive safety: analysing recurring patterns supports process improvement and compliance.

The solution reduces aviation safety and SLA risks and can contribute to preventing damage worth several hundred thousand euros.

Planned extensions

  • automatic monitoring of entries into and exits from the apron zone
  • checking whether the vehicles arrive at the aircraft within the contractual SLA time
  • geographic expansion: the whole Budapest operation, then the equipment of further airports

These are not part of the current pilot: the pilot creates a stable basis for further functional and geographic expansion.

Summary

Sample-based checks turned into continuous, auditable monitoring: automatic detection, video evidence and searchable incidents. Compliance with an aviation safety rule can be monitored in an automated, continuous and auditable way.

Any questions?

Facing a similar challenge?

Tell us about the problem, and after a short assessment we'll show you how REACH would solve it in your plant.

Let's talk