HOW ARTIFICIAL INTELLIGENCE AND OTHER TECHNOLOGY CAN SOLVE SAFETY ISSUES ON THE APRON
Petr Zhigalin
August 26th 2020

Safety is a cornerstone of the aviation industry. Ground operations are complex and regularly result in incidents and accidents. Many of them can be prevented.

According to IATA’s safety report, 15% of aircraft accidents were attributed to ground damage in 2018. Sixty-one percent of ground damages are caused by GSE (ground support equipment).

Ground damages lead to additional expenses, decrease punctuality and operational efficiency. What’s more important, they may lead to injuries and fatalities.

Expenses caused by incidents and accidents on the apron

According to Allianz, the cost of apron incidents at airports is $10bn per year.

When incidents and accidents happen on the apron, airlines and airports have to cover expenses such as:

  1. Expenses caused by unavailability of damaged aircraft or GSE. Airlines have to delay or cancel flights and, as a result, issue ticket refunds, provide meals, pay for overnight hotel stays, and access to business lounge;

  2. Growing insurance costs;

  3. Expenses associated with repair due to unclaimable damage.

Through the use of modern technologies, airlines and airports can detect and, in some instances, even prevent safety violations and accidents, which will not only decrease overall costs, but most importantly, will create a safer environment for employees.

Safety issues on the apron

Among the most widespread safety issues that can be prevented are:

    • Speeding

    • Unauthorized persons on the stand and onboard aircraft

    • Unsecured and unevenly distributed luggage in carts

    • Movement in restricted areas

    • Staff not wearing personal protective equipment (PPE)

    • Aircraft stand is not clear

    • Using vehicles without headlights and taillights

    • None or incorrect placement of chocks and cones

    • Collisions of aircraft with infrastructure or with another aircraft

Speeding on the apron

Speed limits apply to all apron vehicles apart from emergency vehicles. Many different vehicles, however, violate speed limit rules, including tugs, catering and fueling trucks, apron buses, and others. This increases the risk of collisions and other accidents.

As an example, take a pushback tug during pushback. If needed, it would be hard to immediately stop the aircraft during pushback because of its weight. Adherence to speed restrictions is crucial in this situation, however, we have observed incidents where vehicles were tracked at up to 60% above the indicated speed limit.

Below is the pushback speed limit evaluation chart recently prepared by Assaia. Violations were detected in 38.5% of pushbacks.

How to reduce speed violations

Airports and airlines can use different technologies to address the problem of speed violations on the apron.

One option is to install speed control system (SCS), a piece of hardware that limits speed. The benefit is that it is possible to program speed limits for a particular zone to make sure that GSE can travel with higher speed while outside aircraft stand. The system, however, needs to be installed on all ground handling vehicles to ensure 100% coverage.

IoT-based systems (Internet of Things) can also help to measure and manage the speed of GSE. It can be achieved by installing beacons and creating a mesh network with data exchange between devices. Both SCS and IoT-based systems require significant investment and calibration.

Another option to address the problem of speed violations is to use artificial intelligence (AI) and computer vision (CV). AI and CV can use existing apron surveillance cameras to measure the speed of GSE in real time and send alerts to ramp officers. They also provide data on overall speed performance on the apron which can be used for training purposes. The benefit is that existing camera infrastructure can be used and thus no expensive hardware needs to be separately procured.

Unauthorized persons on the stand and onboard aircraft

Only those involved in turnaround handling shall be present on the stand and onboard aircraft. General rule states that pedestrian and vehicular activity on the apron should be kept to minimum required for operations.

It’s impossible to check the ID badge of each turnaround stakeholder and evaluate the level of security and validity of the ID. As a result, unauthorized personnel sometimes access the stand and aircraft.

How to prevent unauthorized personnel from accessing the stand and aircraft

First, the stand perimeter could be programmed to scan ID badges. The system can notify security officers when an unauthorized person attempts to enter the stand.

Alternatively, each stakeholder could be equipped with an IoT beacon or a body-worn camera preventing people from entering irrelevant stands and aircraft.

Another option to prevent unauthorized personnel from accessing the stand and aircraft is to use computer vision. If an airline chooses a specific color for its safety vests, CV can analyze the vest’s color and determine whether the staff member is authorized to enter the stand. If needed, it can send an alert to a ramp officer. In this case, it’s enough to use existing airport cameras and the latest AI software.

Unsecured and unevenly distributed luggage in carts

If luggage is unevenly distributed in carts, it creates a risk of jackknifing and tumbling. Also, if it is not secured in carts, it might fall out during transportation. If lost and not collected after, it can lead to FOD (Foreign Object Debris) on the ramp causing damage, injuries, and passenger complaints.

How to ensure luggage is secured and evenly distributed in carts

First, using semi-automated baggage systems that limit loads can help to avoid issues of unsecured and unevenly distributed luggage. They minimize human interaction with luggage, which means that staff gets less workload and risks. Also, luggage is less likely to fall out of the moving cart. It can prevent incidents on the apron, and passengers receive their luggage faster and in a good condition.

AI and CV can also solve the problem by detecting and reporting unsecured and unevenly distributed luggage. Incidents and damages could be avoided if the ramp officer gets notified about luggage mishandling.

Movement in restricted areas

Sometimes ground operations personnel cut corners by moving under aircraft, its wings, or close to engines. Safety regulations prohibit anyone but technical staff or flight crew to move in these areas, which are always coned. Violation of these safety measures may lead to aircraft damage and, most importantly, threatens people’s lives (e.g., risk of being sucked into a running jet engine).

GSE should also follow rules. They are not allowed to pass between cones, block traffic, drive in the wrong direction, move under the wing or any part of aircraft (except for service vehicles), interfere with moving aircraft, or drive under jetbridges.

safety.vehicle.gif employees.gif

How to prevent movement in restricted areas

One of the options to prevent employees and vehicles from entering restricted areas is to equip them with cameras or IoT beacons.

Alternatively, artificial intelligence and computer vision can be used to detect movement in restricted areas. Although it is difficult to prevent these events from happening as they normally happen fast, the automatically detected events can be used to train staff on the ground and prevent this type of at-risk behavior in the future.

Check Assaia’s video to see how artificial intelligence

can help to improve safety on the apron


Staff not wearing personal protective equipment (PPE)

Being struck by a vehicle is one of the biggest safety hazards on the apron. That is why high-visibility clothing such as safety vests are mandatory for all apron personnel. People sometimes ignore this rule.

In some cases, additional PPE (ear protection, goggles, gloves, etc.) are required. If PPE is not used, it might lead to injuries, loss of hearing, and other health-related issues.

How to ensure that staff use PPE

Airports and airlines should invest in training and promote a safety culture.

Additionally, they can require personnel to use body-worn cameras. If personnel know that they are being observed, they might take PPE more seriously. This method, however, requires regular analysis of video recordings.

PPE control can be done seamlessly by CV and AI in the cloud. It doesn’t require audits – each stand can be analyzed in real time. Employees might take PPE requirements more seriously if they know that technology detects violations 24/7.

Aircraft stand is not clear

An aircraft stand should stay clear of GSE when no aircraft is parked there. GSE should be stored in clearways and in parking boxes (jetbridge wheels). If the stand is not clear when the aircraft arrives, it leads to delays and sometimes collisions, which puts people's lives at risk. Staff is also not allowed to enter and stay in empty stands. It raises the risk of being struck by aircraft or getting sucked into an engine.

How to make sure the stand is clear

The most popular option is to conduct physical checks. “Follow me” vehicles check the stand where they will guide the aircraft to make sure it is clear. It requires a lot of time and peoples’ effort and doesn’t guarantee that the aircraft stand will stay clear after the check. It also increases the number of movements on the apron.

Instead, airports and airlines can use AI, which notifies ramp officers if the stand is not clear. CV monitors stands in real time 24/7.

As an alternative, airports and airlines can employ different smart sensors including IoT to analyze the surface of the stand and report the presence of GSE or foreign objects. This requires following industry standards and equipping stand perimeter surface with weatherproof sensors.

Using vehicles without headlights and taillights

All GSE operators should turn on headlights and taillights at night and in some cases during the day. Lights are especially important during night shifts due to poor visibility. People also tend to be less focused due to fatigue. Visibility is also poor during precipitation or fog. Unfortunately, we still see cases when people forget to turn on the lights. GSE operators cannot see the road well if the lights are turned off. Such GSE might also be invisible to oncoming traffic. This increases chances of getting into accidents.

How to ensure headlights and taillights are turned on

To ensure headlights and taillights are turned on, all apron GSE could be equipped with auto light sensors. GSE operators will not have to keep the procedure of turning on the lights in mind. However, this requires investments since airports have many GSE.

The second option is to use AI to detect vehicles that don’t have lights on and notify ramp agents about the issue.

None or incorrect placement of chocks and cones

Coning restricted areas (engines, wings, etc.) signals to employees that they shouldn’t cross them. Chocks are placed against wheels to prevent accidental movement of aircraft.

Cones and chocks are placed right after aircraft’s arrival and removed right before its departure. Sometimes employees don’t perform these operations or perform them incorrectly.

Frequently GSE operators approach aircraft before chocks are placed. Similarly, ground handlers start servicing the aircraft before cones are placed.

These are serious safety violations that cannot be tolerated.

How to make sure chocks and cones are placed correctly

Body-worn cameras could help to solve this issue. Monitoring would ensure that ground handling personnel does its best to timely place both chocks and cones in correct places. Additionally, it might impact the behavior of GSE operators and ground handlers in surrounding areas - knowing that they might be seen in recordings, they would be less likely to enter the stand before the chocks and cones are placed. If chocks and cones are not placed or placed incorrectly, it will be possible to use recordings for training purposes.

CV and AI can also help – they can detect placement of chocks and cones and notify ramp officers if they are missing.

Collisions of aircraft with infrastructure or with another aircraft

Wingtip clipping is among the most frequent aircraft-to-aircraft collisions. It happens at stands and taxiways. Different situations cause wingtip clipping: understeering, maneuvering around another aircraft, failure to follow the stand or taxiway centerline, and others.

Aircrafts also often collide with infrastructure such as jetbridges and lighting towers. It happens due to understeering (83% of pilots understeer), inappropriate stand allocation, or human error.

How to avoid collisions with infrastructure and aircraft

Visual Docking Guidance Systems (VDGS) allow pilots to park correctly by providing them with visual information regarding current position of aircraft on stand. The downside of VDGS is that some of them recognize ramp officers and wing walkers as obstructions and provide pilots with a false stop signal, which leads to delays and unnecessary fuel burn.

CV can also help to solve the collision problem. It can be taught to determine if an aircraft is docking correctly by tracing front wheels of the aircraft.

There are also stand allocation systems that eliminate the problem of incorrect stand allocation and, as a result, collisions caused by such mistakes.

Increase safety on the apron

As this blog post shows, many safety issues related to ground operations exist today. Limited airport infrastructure will cause even more issues in the future as traffic grows, which will leave less space and time for maneuver.

It’s important to constantly explore new methods that can help to increase safety on the apron. Latest technologies can help to ensure that rules are followed, prevent some incidents and accidents and, as a result, make the apron a safer place. Moreover, IATA has recently shown that if the aviation industry continues to adopt new technologies to enhance safety, it can significantly reduce costs.

Petr Zhigalin has over 5 years of experience in aviation. He worked as a business aviation supervisor and as an airline duty manager. He has expertise in both ground handling and terminal operations. His current role includes analysis of aviation-related data generated by artificial intelligence.


Get in touch