Maintaining an Acceptable Level of Safety Performance at the Airport: Case Study of Split Airport

Abstract

This research investigates the process of determining and maintaining the Acceptable Level of Safety Performance (ALoSP) at an airport, utilizing a case study conducted at Split Airport. The study illustrates how the ALoSP framework, originally developed for State-level application under ICAO Annex 19, can be systematically adapted and implemented at the organizational level within the Safety Management System (SMS) of an aviation service provider. The aim of the study is to systematically demonstrate the process by which an airport defines, monitors, and maintains its ALoSP through the application of Safety Performance Indicators (SPIs), Safety Performance Targets (SPTs), and alert thresholds within the framework of Safety Performance Management (SPM). Main results show that Split Airport consistently maintained its safety performance at an acceptable level throughout a ten-year monitoring period (2015–2024), with a small number of deviations observed in certain safety performance indicators. The findings highlight the airport’s robust safety culture, strong data-driven monitoring, and proactive use of both leading and lagging SPIs to anticipate and prevent safety issues. The study confirms that the ALoSP framework can successfully support continuous safety improvement and regulatory compliance at the organizational level, offering a practical example for other aviation service providers.

1. Introduction

A Safety Management System (SMS) in aviation is a systematic method for overseeing safety that integrates key organizational structures, responsibilities, policies, and procedures [1]. Its main objective is to maintain the safe operation of aircraft through effective safety risk management. SMS functions as a continuous improvement framework focused on identifying hazards, collecting and analyzing safety data, and consistently assessing risk levels [2,3,4]. The ultimate goal is to proactively detect and reduce potential risks before they lead to accidents or incidents.

An effective SMS should be aligned with the organization’s regulatory standards and safety objectives [5,6,7]. It includes essential elements that facilitate hazard identification and risk management, ensuring that the necessary tools and information are accessible for operational purposes. These tools should be appropriately tailored to the organization’s specific activities and scaled to match its operational size and constraints, thereby supporting well-informed safety decisions [8,9].

The International Civil Aviation Organization (ICAO) provides the regulatory basis for aviation safety management systems (SMSs) through its Standards and Recommended Practices (SARPs). These are specifically presented in ICAO Annex 19—Safety Management and further elaborated in the ICAO Doc 9859—Safety Management Manual (SMM) [1,10]. Organizations are required to ensure that their SMS conform to these fundamental regulations, as well as any supplementary standards applicable to their specific operations. For example, airports must comply not only with ICAO Annex 19 and ICAO SMM but also with the requirements outlined in ICAO Annex 14—Aerodromes [11].

The ICAO SMS framework defines four core components and twelve key elements that represent the fundamental requirements for developing an effective and operational SMS [10,12].

The safety policy establishes the fundamental principles, strategies, and procedures that direct an organization’s SMS toward achieving its safety objectives. It reflects senior management’s commitment to integrating and continually improving safety throughout all areas of the organization [13]. Building on this policy, the organization must set clear safety objectives that provide a basis for monitoring and assessing safety performance.

To achieve safety objectives, operators are required to manage the safety risks linked to their operations. This is accomplished through Safety Risk Management (SRM), a process that entails identifying hazards, evaluating the related risks, and implementing appropriate mitigation strategies [10,14].

The Safety Assurance (SA) component is centered on evaluating the effectiveness of safety risk controls. Operators are required to implement systems that continuously track safety performance and measure how well the established controls function [4,15,16]. Safety performance should be assessed against the predefined indicators and targets of the SMS to confirm that the organization’s safety objectives are being achieved [17].

Safety promotion plays a vital role in fostering a strong safety culture and advancing the organization’s safety objectives [10,18]. A positive safety culture is demonstrated through the collective values, attitudes, and behaviors of personnel toward safety. This culture can be enhanced through continuous training, the development of technical competencies, and the encouragement of open communication and exchange of safety-related information.

Safety Performance Management (SPM) is essential for evaluating and controlling an organization’s overall safety performance. It helps organizations address critical questions: what are the most significant safety risks, which safety objectives must be met, which risks should be prioritized, how to track progress toward achieving objectives, and what safety data and information are needed for decision-making [19,20,21,22,23]. When effectively implemented, SPM enables an organization to determine whether its operations and processes align with and support its safety objectives. This is achieved mainly through the establishment of Safety Performance Indicators (SPIs), which serve as measurable metrics for tracking safety performance [10]. The overall SPM framework integrates elements of safety assurance, data from Safety Data Collection and Processing Systems (SDCPS), safety analysis methods, and safety promotion initiatives.

A key aspect of an effective airport SMS is the continuous monitoring of SPIs, along with corresponding Safety Performance Targets (SPTs) and alert levels, which collectively help determine whether the operator’s safety objectives are being achieved.

As per ICAO SMM, Acceptable Level of Safety Performance (ALoSP) represents “the minimum level of safety performance of civil aviation in a State, as defined in its State Safety Program, or of a service provider, as defined in its Safety Management System, expressed in terms of Safety Performance Targets and Safety Performance Indicators” [10].

ALoSP does not represent a single numerical value but rather a collection of performance targets and indicators that define what is deemed “acceptable” for an aviation system, taking into account its specific context such as size, complexity, available resources, and operating environment [24,25,26].

ALoSP represents a balance between assessing the absence of negative outcomes (such as accidents or serious incidents) and the presence of safety-supporting mechanisms (such as reporting systems, safety culture, and training). It is closely linked to the implementation of a State Safety Program (SSP) and service providers’ SMSs, where achieving ALoSP indicates that the system is not only compliant but operates at an acceptable level of safety performance [27]. Establishing realistic SPIs and SPTs requires sufficient safety data, analytical tools, and modeling resources (e.g., historical records and occurrence databases) [28,29]. However, the limited availability of such data poses a significant challenge in determining ALoSP.

Key steps in determining ALoSP include understanding the operational system and context, selecting relevant SPIs and SPTs, setting-up alert or trigger thresholds, collecting and using available data, comparing actual performance against targeted one, reviewing and adjusting as necessary.

The concept of ALoSP was originally developed and implemented at the state level, as an integral part of the SSP, in accordance with ICAO Annex 19 and ICAO SMM [1,10]. At this level, the ALoSP represents a reference framework by which a State defines the minimum acceptable level of safety performance in civil aviation, expressed through SPIs and SPTs. In this way, the State establishes the boundary between “acceptable” and “unacceptable” risk, taking into account the complexity of the national system, traffic flows, infrastructure and organizational capacities, and regulatory requirements [10].

However, although the ALoSP is primarily intended for national aviation authorities, in practice the concept is increasingly being transferred to the level of aviation service providers—for example, airports, airline operators, air navigation service providers and aircraft maintenance organizations—who integrate it within their own SMS. This application arises from the need to systematically monitor safety performance within individual organizations, and not only at the State level [22,30,31,32,33,34].

Service providers should define their own safety performance that is consistent with the national ALoSP, in order to ensure vertical alignment between the SSP and the SMS [10]. In other words, the organizational ALoSP represents an operational translation of the State concept—a set of internal indicators, targets and thresholds that reflect how effectively safety is managed in the specific operational context of the provider concerned.

In such cases, the ALoSP within SMS allows the organization to measure its own safety performance against set targets, identify deviations from acceptable limits, proactively manage risks within the operational environment, and demonstrate compliance with State safety expectations to regulatory authorities.

Examples of such practices can be found in some airports [35,36] and air navigation service providers [36,37], where locally defined ALoSP frameworks are a key tool for managing and monitoring the effectiveness of safety measures.

Recent regulatory developments place increasing emphasis on aviation safety performance, particularly through ICAO’s Global Aviation Safety Plan (GASP) [38,39] and Global Aviation Safety Roadmap [40], and through the European Union Aviation Safety Agency (EASA) documents such as the EASA European Plan for Aviation Safety and the EASA Acceptable Level of Safety Performance (ALoSP) [37,41], which highlight the obligation of service providers to develop measurable, data-driven SPIs and align them with national safety objectives. The role of predictive analytics, risk-based decision-making, and harmonized performance frameworks across aviation organizations, represent a shift from compliance-based oversight to continuous monitoring and performance improvement. These principles directly support the organizational application of ALoSP explored in this study.

Research on organizational-level performance management further reinforces the need for robust SPI frameworks. Numerous studies have demonstrated how safety culture, training systems, and internal reporting mechanisms influence safety performance outcomes [27,42,43,44,45]. For example, some studies examine how organizations optimize safety performance through integrated risk management processes, while some discuss the interaction between performance indicators and organizational safety culture [46]. Such research shows that SPIs not only measure outcomes but also reflect internal maturity and managerial capability.

Empirical comparisons of airport safety performance indicators have emerged in recent years. Some studies evaluate airport incident patterns and SPI-based monitoring, showing how traffic volume, infrastructure complexity, and reporting culture affect SPI outcomes [47]. These studies outline the importance of contextualizing SPI data through operational interpretations of deviations and alert-level analyses.

This paper presents how airport manages their safety performance, with a detailed description of each step of determining and maintaining ALoSP, following the ICAO prescribed methodology and organizational SPM framework. This includes monitoring the safety performance indicators, defining target values, calculating and setting-up alert thresholds for unacceptable performance, and finally comparing achieved level of safety performance against targeted level of performance, i.e., determining airport’s ALoSP. In order to show how organizational ALoSP is determined and maintained, a case study was conducted at Split Airport.

2. Data Collection and Methodology

This chapter describes collected data and methods used to conduct the analysis of airport’s safety performance and its process of maintaining an acceptable level of safety performance.

2.1. Data Collection

This paper analyzes the safety performance of an airport—specifically, how it establishes, achieves, and maintains an acceptable level of safety performance. To carry out this analysis, data on the airport’s safety performance were required. Such data were collected from the case-study airport, i.e., Split Airport, to evaluate its safety performance, calculate the achieved safety levels, and assess whether it meets acceptable standards.

Split Airport, also called Saint Jerome Airport, is one of nine airports in Croatia, situated in the Resnik area west of Kaštel Štafilić, approximately 25 km from Split. Its primary operations involve passenger, cargo, and aircraft handling for both domestic and international air traffic. Key components of the airport’s infrastructure include maneuvering areas such as the runway (05/23) and apron, as well as the passenger and cargo terminals, control tower, access roads, parking areas for buses and cars, and various service and commercial facilities [13,48].

Split Airport began operations on 15 November 1966. Passenger traffic increased steadily each year until 1988, when growth halted as a result of the economic crisis. The airport was closed in September 1991 due to the war but reopened in April 1992 [48].

In 2019, Split Airport became Croatia’s second busiest airport after Zagreb Airport, serving 3.3 million passengers. To accommodate rising traffic, especially in summer, a major expansion completed that year tripled the terminal’s size and increased capacity to 5 million passengers annually. The original terminal was renovated for some international departures, while the new facilities handle check-in, arrivals, and domestic flights. The project also included a covered bridge over state road D409, connecting the terminal to new parking, bus, and rental car areas [13,48].

In 2020, Split Airport experienced a major 79% decline in traffic due to strict pandemic restrictions. Passenger numbers began to recover gradually (Figure 1), reaching 50% of 2019 levels by 2021. That year, Split Airport became the busiest airport in Croatia for the first time, handling 1.6 million passengers and surpassing Zagreb Airport. Recovery continued, with 2.9 million passengers in 2022, 3.4 million in 2023, 3.6 million passengers recorded in 2024, and 3.7 million in 2025 (recorded by October 2025) [49,50].

Split Airport upholds high safety standards and is committed to continuously enhance its safety management system. The airport applies both reactive and proactive approaches to collect and analyze safety data. Safety management is supported by advanced software, Galiot Aero SMS [50,51], which enables robust and effective proactive safety management.

This analysis utilized a dataset containing actual safety performance indicators from Split Airport. As part of its safety assurance framework, the airport has defined a comprehensive set of safety performance indicators (SPIs) along with corresponding safety performance targets (SPTs). These indicators are monitored monthly to track safety performance [50]. The list of SPIs and SPTs of the Split Airport SMS are presented in the following

Table 1. List of SPIs and SPTs in observed dataset at the Split Airport [50].

Mark	Full Name of the SPI	Short Name of the SPI	Set Target (SPT 1 for SPIs)
SPI-1	Communication occurrences	Communication	≤1/10,000	≤0.0001
SPI-2	LIRF and loadsheet crosscheck omissions	LIRF and loadsheet crosscheck	≤1/10,000	≤0.0001
SPI-3	Incidents related to wrong figures for loadsheet	Wrong figures for loadsheet	≤1/10,000	≤0.0001
SPI-4	Anti-collision occurrences	Anti-collision occurrences	≤1/1000	≤0.001
SPI-5	Occurrences related to wildlife	Wildlife	≤1/1000	≤0.001
SPI-6	Dangerous goods incidents	Dangerous goods	≤1/10,000	≤0.0001
SPI-7	Occurrences related to baggage loading/unloading	Baggage loading/unloading	≤1/1000	≤0.001
SPI-8	Occurrences related to passenger handling at the gate	Passenger handling at the gate	≤1/1000	≤0.001
SPI-9	Occurrences related to passenger handling (disembarking/embarking)	Passenger handling (disembarking/embarking)	≤1/1000	≤0.001
SPI-10	Personnel or passenger injuries	Personnel or passenger injuries	≤1/1000	≤0.001
SPI-11	Occurrences related to personal protective equipment	Personal protective equipment	≤1/1000	≤0.001
SPI-12	Training deficiencies	Training deficiencies	≤1/1000	≤0.001
SPI-13	Engine start-up incidents	Engine start-up	≤1/1000	≤0.001
SPI-14	Occurrences related to fuel handling	Fuel handling	≤1/1000	≤0.001
SPI-15	Aircraft damage occurrences	Aircraft damage	≤1/100	≤0.01
SPI-16	Aircraft marshaling occurrences	Aircraft marshaling	≤1/1000	≤0.001
SPI-17	Aircraft chocking incidents	Aircraft chocking	≤1/1000	≤0.001
SPI-18	Aircraft conning incidents	Aircraft conning	≤1/1000	≤0.001
SPI-19	Runway incursions/excursions	Runway incursions/excursions	≤1/10,000	≤0.0001
SPI-20	Occurrences related to taxiing to/from apron	Taxiing to/from apron	≤1/1000	≤0.001
SPI-21	Apron maintenance incidents	Apron maintenance	≤1/1000	≤0.001
SPI-22	Occurrences related to FOD presence	FOD presence	≤1/1000	≤0.001
SPI-23	Occurrences related to maneuvering area maintenance	Maneuvering area maintenance	≤1/1000	≤0.001
SPI-24	Vehicle maintenance incidents	Vehicle maintenance	≤1/1000	≤0.001
SPI-25	Occurrences related to ground traffic (GSE) and vehicle driving	Ground traffic (GSE) and vehicle driving	≤1/1000	≤0.001

¹ Value of the Acceptable Level of Safety Performance (ALoSP) is equal or less than value of Safety Performance Targets (SPT), i.e., fixed (set) target value of airport’s SPIs expressed in specific number of occurrences versus specific number of aircraft operations.

.

2.2. Methodology

A key element of SPM is the establishment of SPIs, which assess how effectively an organization manages safety. SPIs are tailored to each organization’s specific operations and needs, helping management evaluate progress toward safety objectives [10,52,53]. These indicators may include data on accidents, incidents, non-conformities, and related occurrences.

SPIs can be either quantitative (numerical) or qualitative (descriptive or subjective). They are also categorized as leading or lagging indicators [26,54,55,56,57,58]. Leading SPIs, also known as process or active indicators, monitor actions and processes designed to prevent incidents, reflecting a proactive approach to safety and supporting adaptation to operational changes. Lagging SPIs, or outcome-based indicators, measure past events such as accidents and incidents, and are used to evaluate an organization’s overall safety performance.

Each SPI should clearly define what it measures, its goal, calculation method, responsible personnel, data sources, and the frequency of reporting and analysis. Two thresholds are established for every SPI: a safety performance target (SPT), which represents the desired performance level, and an alert or trigger level (three versions), which indicates when the SPI values exceed desired limits [10].

Determining target and alert levels involves calculating the mean (average) and standard deviation of SPI values over a defined period, typically one year, during which the indicators are tracked and recorded monthly [10,24].

Formula (1) is applied to determine the mean (average) of the observed SPI values, where μ(SPI) represents the average SPI value, SPI_i denotes individual observed SPI values, and n is the total number of observations in the dataset:

$$\mu \left(SPI\right)={\displaystyle \frac{{\sum }_{i=1}^n{SPI}_i}{n}}.$$

(1)

The following formula illustrates how to calculate the standard deviation for the observed set of SPI values:

$$\sigma \left(SPI\right)=\sqrt{{\displaystyle \frac{{\sum }_{i=1}^n{{(SPI}_i-\mu \left(SPI\right))}^2}{n}}},$$

(2)

where σ(SPI) represents the standard deviation of the observed SPI values.

The target level, or SPT, is determined to be equal to or lower than the mean (average) of the observed SPI values, as shown in the following formula:

$$Target\le \mu \left(SPI\right),$$

(3)

where μ(SPI) represents the mean (average) value of the observed SPI values.

Besides being determined as equal or lower than the mean (average) of the observed SPI values, SPTs may be set as a fixed target set by an operator, regardless of μ(SPI), which allows aviation service providers to set SPT based on operational characteristics or strategic priorities, i.e., if mean-based approach is not appropriate [10].

Alert settings function as statistical warnings to detect unusual trends in SPI data. Three alarm levels are activated depending on how many consecutive values exceed the defined thresholds, indicating possible issues that may require mitigative or corrective action. The first alarm is triggered after three or more consecutive exceedances, the second after two or more, and the third after one or more.

The values for the three alarm levels are determined using the following formulas:

$$Alert1=\mu \left(SPI\right)+\sigma \left(SPI\right),$$

(4)

$$Alert2=\mu \left(SPI\right)+2\sigma \left(SPI\right),$$

(5)

$$Alert3=\mu \left(SPI\right)+3\sigma \left(SPI\right)$$

(6)

where μ(SPI) represents the mean (average) value and σ(SPI) represents the standard deviation value of the observed SPI values.

The relationship between monthly SPI calculations, annual mean values, alert levels, and the determination of ALoSP follows a linear and sequential structure. First, each SPI is calculated monthly by dividing the number of safety-related events by the number of aircraft operations for that month. These monthly ratios form a 12-month dataset from which an annual mean and standard deviation are derived, representing the baseline safety performance for that indicator. Using these two statistics, three alert levels are generated to signal abnormal deviations. During the subsequent year, actual monthly SPI values are compared continuously against both the annual mean (or set SPT) and alert thresholds. SPTs may be set as a fixed target set by an operator, or as the mean (average) value of the observed SPI values (from the preceding period). At the end of the monitoring cycle, the achieved level of safety performance is assessed by comparing the annual SPI mean with the predefined SPT. If the achieved annual SPI does not exceed the SPT and no alert-level breaches occur, the SPI is considered to have achieved its acceptable level of safety performance, contributing to the airport’s overall safety performance.

Figure 2 shows a flowchart of relationship among monthly SPI calculations, annual mean values, alert levels, and the determination of ALoSP.

The safety performance of each SPI is assessed at the end of every monitoring cycle (typically a twelve-month period) based on two criteria: whether the target was met and whether any alerts were triggered.

Using this information, the organization can maintain and continuously enhance its safety performance to an acceptable level.

3. Determining an Acceptable Level of Safety Performance at the Airport: Case Study of Split Airport

As part of its safety performance management, Split Airport has established a comprehensive set of SPIs and corresponding SPTs, which are monitored on a monthly basis [50]. Sample of data on SPI-8 Passenger handling at the gate and its related SPTs, from January 2014 to December 2024, is presented in

Table 2. Example of monthly data on SPI-8 Passenger handling at the gate at Split Airport for the period 2014–2024 with number of recorded events, number of operations, calculated values of SPI-8, and accompanying safety performance target (SPT) set by the airport operator.

Year	Number of Events	Number of Aircraft Operations	SPI-8 Value (Calculated Ratio)	SPT [Set by the Airport Operator]
January-2014	0	438	0	0.001
February-2014	0	392	0	0.001
March-2014	0	514	0	0.001
April-2014	0	1032	0	0.001
May-2014	0	1942	0	0.001
June-2014	1	2554	0.000392	0.001
July-2014	5	3872	0.001291	0.001
…	…	…	…	…
October-2024	0	3502	0	0.001
November-2024	0	798	0	0.001
December-2024	0	624	0	0.001

. All monthly data of all 25 SPIs in the form of recorded events and calculated SPI values (expressed in events per operations), covering the period from January 2014 to December 2024, are shown in Appendix A.

Table 3. Calculation of achieved level of safety performance of SPI-8 Passenger handling at the gate at Split Airport in 2018.

Monitoring Period	Number of Aircraft Operations	Number of Events (SPI-8)	SPI-8 (Preceding 1-Year Period)	Mean of the SPI (Based on Preceding 1-Year Period)	Alert 1 (Based on Preceding 1-Year Period)	Alert 2 (Based on Preceding 1-Year Period)	Alert 3 (Based on Preceding 1-Year Period)	SPT ≤ (Set by the Airport Operator)	SPI-8 [Events per Operations]	Mean of SPI—Achieved Level of Safety Performance
January-2018	586		0.000000	0.000666	0.001498	0.002330	0.003161	0.001	0.000000	0.000906
February-2018	496		0.000000	0.000666	0.001498	0.002330	0.003161	0.001	0.000000	0.000906
March-2018	640		0.000000	0.000666	0.001498	0.002330	0.003161	0.001	0.000000	0.000906
April-2018	1378	1	0.000726	0.000666	0.001498	0.002330	0.003161	0.001	0.000673	0.000906
May-2018	2644	3	0.001135	0.000666	0.001498	0.002330	0.003161	0.001	0.001737	0.000906
June-2018	3594	6	0.001669	0.000666	0.001498	0.002330	0.003161	0.001	0.000494	0.000906
July-2018	5216	7	0.001342	0.000666	0.001498	0.002330	0.003161	0.001	0.001090	0.000906
August-2018	5078		0.000000	0.000666	0.001498	0.002330	0.003161	0.001	0.000974	0.000906
September-2018	378	1	0.002646	0.000666	0.001498	0.002330	0.003161	0.001	0.001041	0.000906
October-2018	2116	1	0.000473	0.000666	0.001498	0.002330	0.003161	0.001	0.002201	0.000906
November-2018	654		0.000000	0.000666	0.001498	0.002330	0.003161	0.001	0.002667	0.000906
December-2018	554		0.000000	0.000666	0.001498	0.002330	0.003161	0.001	0.000000	0.000906

shows the calculation (as per formulas in Section 2.2) of achieved level of safety performance of SPI-8 Passenger handling at the gate at Split Airport in 2018 with number of recorded events, number of operations, calculated values of SPI-8 in each month in the preceding 1-year period (expressed in events per operations), i.e., in 2017, mean of the SPI (based on preceding 1-year period, i.e., 2017), Alert 1, 2 and 3 (based on preceding 1-year period, i.e., 2017), calculated values of SPI-8 in each month in 2018 (expressed in events per operations), and mean of SPI in 2018, i.e., achieved level of safety performance in 2018.

Calculations of achieved levels of safety performance for SPI-8 Passenger handling at the gate, at Split Airport, for the period 2015–2024, are presented in

Table 4. Calculation of three alert levels and achieved levels of safety performance at Split Airport for the period 2015–2024 for SPI-8 Passenger handling at the gate.

Year	Alert 1 (Based on Preceding 1-Year Period)	Alert 2 (Based on Preceding 1-Year Period)	Alert 3 (Based on Preceding 1-Year Period)	SPT ≤ (Set by the Airport Operator)	Achieved Level of Safety Performance
2015	0.001674	0.002535	0.003395	0.001	0.000686
2016	0.001844	0.003001	0.004159	0.001	0.001319
2017	0.002871	0.004422	0.005974	0.001	0.000213
2018	0.000637	0.001060	0.001484	0.001	0.000672
2019	0.001493	0.002314	0.003136	0.001	0.000837
2020	0.002149	0.003460	0.004771	0.001	0.000223
2021	0.000719	0.001215	0.001711	0.001	0.000218
2022	0.000604	0.000991	0.001378	0.001	0.000179
2023	0.000457	0.000734	0.001012	0.001	0.000123
2024	0.000290	0.000457	0.000624	0.001	0.000152

. The three alert levels are also calculated for each year, according to the formulas outlined in Section 2.2. Analogously, the calculations of achieved levels of safety performance with accompanying three alert levels for all other SPIs monitored at Split Airport, for the period 2015–2024, are presented in Appendix A.

Figure 3 illustrates the monthly monitoring of SPI-8 values at Split Airport, showing whether they met the established target level and avoided the three defined alert levels. The figure presents the achieved level of safety performance for each year (from 2015 to 2024), revealing that, except for 2016, the achieved values consistently remained below the set target, indicating effective safety performance management at Split Airport. Analogously, illustrations of achieved levels of safety performance with accompanying three alert levels for all 25 SPIs monitored at Split Airport, for the period 2015–2024, are shown in Appendix A.

The set target level, achieved level of safety performance and three alert levels of all 25 safety performance indicators, are presented in Figure 4, for each year of the observed time period from 2015 to 2024. For example, as shown in Figure 4a,c,d,g,i–t,v–x in the ten-year period, achieved levels of safety performance were below the set target level for every year of the observed period. As per Figure 4b,e,f,h,u,y achieved levels of safety performance breached set target level in six areas of monitoring which points to the critical safety issues at Split Airport, i.e., regarding LIRF and loadsheet crosscheck (SPI-2), wildlife (SPI-5), dangerous goods (SPI-6), passenger handling at the gate (SPI-8), apron maintenance (SPI-21), and ground traffic (GSE) and vehicle driving (SPI-25). Breaches occurred in 2016, 2017, 2019, 2020, and 2024. Other years mark no breaches in any area of monitoring.

Throughout the monitoring period, SPI-1 to SPI-4 (communication, loadsheet errors, anti-collision events) demonstrate low-frequency but highly critical events that are predominantly influenced by human–machine interface issues, workload peaks, and seasonal traffic growth. Minor increases (e.g., SPI-2 in 2016–2017) correspond with intensive summer-season operations, indicating potential process saturation and reinforcing the need for periodic procedural refreshers and tighter cross-check mechanisms. Their safety implication is significant, i.e., even small fluctuations in these indicators represent vulnerabilities in core aircraft-handling processes that directly affect flight preparation and ground safety.

Wildlife occurrences (SPI-5) show periodic spikes (e.g., 2017) that correlate with seasonal migratory patterns, vegetation cycles, and construction-related disturbances in the airport environment. These deviations emphasize the dynamic nature of environmental hazards and the need for adaptive wildlife management programs, such as habitat control and enhanced monitoring technologies. FOD occurrences (SPI-22), although less volatile, similarly reflect the influence of apron congestion and maintenance workload, especially during peak seasons.

Breaches in 2019, 2020, and 2024 for SPI 6 (dangerous goods) reflect increased cargo diversification, temporary staffing shortages during the COVID-19 operational downturn, and procedural handovers among ground handlers. DG mishandling is a latent high-risk precursor to operational emergencies, and the airport intensified the DG refresher training and standardized procedures.

SPI-7 to SPI-12 (passenger and baggage handling, training, protective equipment, injuries) reveal patterns linked to personnel competencies, training cycles, and staffing variability. For example, SPI-8 in 2016 shows a sharp increase consistent with operational strain before terminal expansion (as per Table 4). Trends in these indicators show that increases correspond to organizational transitions, such as rapid seasonal recruitment or changes in terminal configuration. Their safety implication lies in the early-warning value, i.e., they enable detection of systemic safety issues before manifesting as more severe events (injuries, aircraft damage).

For SPI-13 to SPI-18 (engine start, fuel handling, aircraft ground movements) most values fall well below SPTs, indicating stable performance. Occasional spikes (e.g., SPI-14 in 2017) stem from coordination gaps between multiple contractors or weather-related fueling delays. These SPIs reveal how effectively standard procedures are executed under irregular operating conditions.

Indicators SPI-19 to SPI-25 (runway/taxiway/apron operations and ground traffic) reflect infrastructural, environmental, and human-factor interaction. Deviations in SPI-21 (apron maintenance) in 2020 and SPI-25 (ground traffic and vehicle driving) in 2016 and 2024 correspond with increased construction/maintenance activities, seasonal congestion, and adherence to airside driving rules.

Based on the results, it is evident that Split Airport places strong emphasis on safety and responds promptly to emerging issues, demonstrating exemplary practices in safety performance management.

The following

Table 5. Determining an acceptable level of safety performance at Split Airport through 25 SPIs in the period from 2015 to 2024.

SPIs	SP Levels	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024
SPI -1	SPT	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001
	Achieved	0.000000	0.000025	0.000000	0.000000	0.000000	0.000000	0.000000	0.000024	0.000000	0.000000
	Acceptable	YES *	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-2	SPT	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001
	Achieved	0.000019	0.000134	0.000130	0.000000	0.000030	0.000023	0.000000	0.000000	0.000018	0.000000
	Acceptable	YES	NO **	NO	YES	YES	YES	YES	YES	YES	YES
SPI-3	SPT	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001
	Achieved	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000020	0.000000	0.000018	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-4	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000000	0.000025	0.000023	0.000037	0.000000	0.000000	0.000000	0.000038	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-5	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000312	0.000360	0.001456	0.000423	0.000547	0.000445	0.000188	0.000445	0.000699	0.000596
	Acceptable	YES	YES	NO	YES	YES	YES	YES	YES	YES	YES
SPI-6	SPT	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001
	Achieved	0.000056	0.000061	0.000000	0.000000	0.000252	0.000203	0.000000	0.000000	0.000032	0.000134
	Acceptable	YES	YES	YES	YES	NO	NO	YES	YES	YES	NO
SPI-7	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000076	0.000271	0.000000	0.000283	0.000057	0.000000	0.000160	0.000103	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-8	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000886	0.001848	0.000666	0.000906	0.000031	0.000903	0.000106	0.000162	0.000000	0.000095
	Acceptable	YES	NO	YES	YES	YES	YES	YES	YES	YES	YES
SPI-9	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000028	0.000018	0.000142	0.000016	0.000000	0.000030	0.000041	0.000077	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-10	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000069	0.000264	0.000079	0.000088	0.000043	0.000091	0.000081	0.000017	0.000000	0.000027
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-11	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000020	0.000035	0.000000	0.000022	0.000000	0.000000	0.000000	0.000000	0.000148	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-12	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000000	0.000168	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-13	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000066	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-14	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000019	0.000096	0.000581	0.000021	0.000000	0.000116	0.000018	0.000000	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-15	SPT	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01
	Achieved	0.000019	0.000018	0.000142	0.000000	0.000000	0.000000	0.000000	0.000000	0.000043	0.000039
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-16	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000074	0.000025	0.000127	0.000016	0.000104	0.000000	0.000000	0.000000	0.000000	0.000019
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-17	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000000	0.000000	0.000207	0.000321	0.000000	0.000102	0.000000	0.000000	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-18	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000000	0.000000	0.000039	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-19	SPT	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001
	Achieved	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-20	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000000	0.000000	0.000000	0.000111	0.000049	0.000023	0.000000	0.000034	0.000044	0.000038
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-21	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000258	0.000044	0.000016	0.000000	0.000120	0.005208	0.000041	0.000118	0.000000	0.000103
	Acceptable	YES	YES	YES	YES	YES	NO	YES	YES	YES	YES
SPI-22	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000311	0.000000	0.000000	0.000160	0.000955	0.000000	0.000189	0.000000	0.000000	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-23	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000165	0.000000	0.000270	0.000429	0.000086	0.000000	0.000000	0.000395	0.000106	0.000000
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-24	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000168	0.000186	0.000032	0.000037	0.000000	0.000000	0.000000	0.000038	0.000126	0.000032
	Acceptable	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
SPI-25	SPT	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001	0.001
	Achieved	0.000686	0.001319	0.000213	0.000672	0.000837	0.000223	0.000218	0.000179	0.000123	0.000152
	Acceptable	YES	NO	YES	YES	YES	YES	YES	YES	YES	YES

* SPT achieved (green color) = acceptable. ** SPT not achieved (red color) = unacceptable.

shows the comparison of set target levels vs. achieved levels of safety performance, i.e., determining an acceptable level of safety performance (ALoSP) at Split Airport. If the achieved level is below or equal to set SPT, it is considered to be an acceptable level of safety performance or ALoSP. Where the acceptable level has been achieved, it is marked in green letters with “YES”, and where it is not achieved, it is marked in red letters with “NO”.

Figure 5 shows the illustrated summary of achieved levels of safety performance at Split Airport, through all 25 monitored SPIs. If the achieved level of safety performance is acceptable it is marked in light green color. The breaches, i.e., “unacceptable” levels are marked in orange color.

4. Discussion and Conclusions

The findings of this study comply with the theoretical background and earlier research discussed in the paper. As highlighted in the ICAO Safety Management Manual (Doc 9859) [10], the effectiveness of SMS depends heavily on an organization’s ability to monitor its safety data, understand what that data represents, and take action when needed. This is clearly presented at Split Airport, where long-term SPI trends show both areas where operations are strong and areas where the system occasionally comes under pressure. Additionally, ICAO’s Safety Management Manual [10] points out that safety performance indicators and alert levels should serve as proactive warning signals, and not just numbers to be recorded. Split Airport’s repeated return of SPI values to acceptable levels after short-term deviations shows that these indicators function as intended in real operations. ICAO Annex 19 [1] also emphasizes that safety performance monitoring needs to combine both predictive and reactive elements. This principle is reflected in the way the airport’s SPIs behave over time, e.g., training-related indicators, human-factors issues, wildlife indicators, and others complement one another to give a full oversight of system safety performance. Similar guidance appears in EASA’s safety performance management principles [22,30,31,32,33,34], which highlight that safety objectives must remain flexible and responsive to real operational conditions. Split Airport’s data clearly shows this, as periods of construction activity, seasonal peaks in traffic, and temporary resource shortages all create noticeable shifts in some SPIs; they also confirm the link between changing operational environments and safety performance. Other analyses of airport safety occurrences [36,37] add that maneuvering area operations, interactions between staff and equipment, and adherence to procedures are among the most sensitive parts of airport systems. This aligns with the behavior of Split Airport’s SPIs for ground handling issues, apron maintenance, and vehicle-movement deviations, i.e., areas where even small disruptions can influence adverse safety outcomes. Further research cited in this paper shows that reporting culture and organizational maturity significantly affect SPI stability (e.g., [24,25,26,28,29]). This corresponds well with Split Airport’s dataset, which demonstrates strong data quality and consistent patterns over time. Other referenced studies, such as those highlighting the predictive value of leading indicators [26,54,55,56,57,58], the effect of seasonal traffic patterns on SPI variability [49,50], and the influence of infrastructure configuration on SPI trends [13,48], also support the interpretation of results in this research. Finally, the findings align with studies emphasizing organizational resilience (e.g., [27]), which define resilience as the ability of a system to withstand operational pressures and quickly return to desired performance levels.

The results of this study provide a comprehensive overview of how Split Airport systematically determines, monitors, and maintains its Acceptable Level of Safety Performance (ALoSP) through a structured Safety Management System (SMS). The results demonstrate a mature Safety Performance Management (SPM) framework where the organization not only meets regulatory compliance but also actively uses data-driven indicators to maintain proactive safety management. By analyzing 25 Safety Performance Indicators (SPIs) over a ten-year period (2015–2024), the study highlights the airport’s consistent achievement of Safety Performance Targets (SPTs), with only limited exceedances, thus evidencing a high level of operational safety and effective risk control.

The longitudinal analysis revealed that, across most safety areas, the achieved levels of safety performance remained below or equal to the predefined SPTs throughout the observed period. This outcome signifies that the airport has been operating within its defined ALoSP boundaries, maintaining an acceptable level of safety despite fluctuations in operational activity and external pressures such as the COVID-19 pandemic. Only six out of the 25 SPIs experienced exceedances of their set targets: LIRF and loadsheet crosscheck (SPI-2), Wildlife (SPI-5), Dangerous goods (SPI-6), Passenger handling at the gate (SPI-8), Apron maintenance (SPI-21), and Ground traffic (GSE) and vehicle driving (SPI-25). These deviations occurred sporadically in 2016, 2017, 2019, 2020, and 2024 and were promptly addressed by the airport’s safety management team. The limited number and temporal distribution of breaches reflect an organization capable of timely identifying emerging risks and implementing corrective actions before they escalate into systemic safety issues.

Deviations, i.e., minor increases, in SPI-2 (LIRF and loadsheet crosscheck) during 2016–2017 corresponded with intensive summer-season operations, indicating process saturation and the need for refresher training and strengthened loadsheet crosscheck mechanisms. SPI-5 (wildlife occurrences) showed episodic spikes (e.g., 2017) correlated with migratory periods, vegetation cycles, and construction-related disturbances, highlighting the dynamic nature of environmental hazards and reinforcing the need for adaptive wildlife management strategies. SPI-6 (dangerous goods incidents) deviated in 2019, 2020, and 2024 due to increased cargo diversification, temporary staffing shortages during COVID-19, and procedural inconsistencies among ground handlers; these findings prompted enhanced DG training and standardized handling procedures. The increase in SPI-8 (passenger handling at the gate) in 2016 aligned with operational strain prior to terminal expansion, as documented in airport operational records. Deviations in SPI-21 (apron maintenance) in 2020 and SPI-25 (GSE & vehicle driving) in 2016 and 2024 were associated with construction activities, increased congestion, and varying airside driving movements.

Detailed analysis of SPI-8 (Passenger handling at the gate), presented as a representative example, shows that the achieved safety levels were generally below the set SPT (≤0.001) in almost all years except 2016. This deviation was influenced by increased operational demand prior to the 2019 terminal expansion, introducing temporary process strains. However, the improvement trend observed from 2017 onward confirms that corrective and preventive measures, such as enhanced ground handling procedures, personnel training, and process standardization, were effective. Similar cyclical or one-off exceedances were observed in wildlife and ground traffic–related indicators (SPI-5 and SPI-25), which are often linked to environmental and infrastructural factors rather than organizational deficiencies. Their quick return to acceptable limits points to the adaptability of Split Airport’s safety management system and the responsiveness of its safety culture.

The results also show Split Airport’s balanced use of both leading and lagging indicators. Leading indicators (e.g., training deficiencies, equipment maintenance, and apron inspections) capture the preventive side of safety management, whereas lagging indicators (e.g., incidents, load sheet errors, and wildlife occurrences) measure the consequences of safety risks that have materialized. By combining these complementary perspectives, the airport achieves a holistic understanding of its operational safety status. This aligns with ICAO and EASA recommendations, which advocate a shift from purely reactive monitoring of accidents toward proactive, process-based measurement of system resilience. The fact that the majority of leading indicators remained below alert thresholds over the decade suggests a stable internal safety culture that anticipates and mitigates potential hazards before they result in adverse occurrences.

A key strength of the approach applied in this study lies in the statistical methodology used to calculate mean SPI values and standard deviations for three alert levels. The introduction of three progressive alert thresholds allows for early detection of unusual trends and supports evidence-based decision-making. The alert system operates as an internal “safety thermostat,” distinguishing between normal operational variability and significant deviations requiring management intervention. The observed low frequency of alert activations indicates that the airport’s SPM system successfully maintains operations within its expected control limits. Moreover, the consistent adherence to SPTs across multiple operational domains demonstrates the effectiveness of using historical data as a basis for target setting and continuous improvement.

The period analyzed also covers substantial operational fluctuations caused by the COVID-19 pandemic. The 79% drop in traffic in 2020 and the subsequent recovery through 2025 presented unique safety management challenges, including personnel turnover, procedural adaptations, and fluctuating workload levels. Despite these pressures, the airport maintained acceptable safety levels across all SPIs, indicating system robustness and adaptive capacity. This resilience is an essential characteristic of an effective SMS, as it ensures that safety performance does not deteriorate during periods of change or operational stress.

An important insight from the results is the demonstration that the ALoSP concept, originally developed for State-level implementation under ICAO Annex 19, can be successfully operationalized at the service-provider level. Split Airport’s experience exemplifies how ALoSP can be translated into concrete organizational practices through internally defined SPIs, SPTs, and alert thresholds. By aligning its performance framework with the national State Safety Program (SSP), the airport ensures vertical consistency between State safety objectives and organizational execution. This integration fosters accountability, transparency, and measurable safety outcomes at the service-provider level, strengthening the overall aviation safety system.

The stable results across the ten-year monitoring period also illustrate the maturity of Split Airport’s safety culture. Continuous monitoring, systematic data analysis, and proactive safety assurance activities demonstrate a learning organization that values improvement and prevention over mere compliance. The airport’s use of Galiot Aero SMS software further enhances its capability to manage safety information, conduct trend analysis, and maintain traceable safety records. These practices align with international standards and contribute to a data-rich environment where management decisions are evidence-based, and safety management is both timely and transparent.

Although Split Airport’s safety performance is commendable, the few instances of target breaches highlight opportunities for refinement. The occurrences related to, e.g., wildlife, ground traffic, and dangerous goods, suggest the need for enhanced coordination between operational departments, more sophisticated monitoring technologies (e.g., wildlife radar systems or vehicle tracking), and continuous staff training. Moreover, while the dataset spans ten years, expanding the monitoring framework to include predictive modeling, could strengthen the airport’s capacity to forecast potential safety degradations. Another challenge is the dependence on sufficient and high-quality safety data; i.e., airports with limited reporting culture may find it difficult to replicate such detailed analyses without robust data management systems.

While the ICAO-prescribed three-alert system provides structure and statistical rigor, it has some shortcomings. It assumes normal distribution of SPI values and steady operational conditions, which is not always representative of airport environments. SPIs with rare events (e.g., SPI-19 Runway incursions) may produce unrealistic means and standard deviations, reducing sensitivity to meaningful changes. However, high-volume SPIs may generate alerts driven by short-term operational variability rather than real safety degradation. The method also does not incorporate seasonal patterns, traffic peaks, construction activity, etc., which make it necessary for safety managers to apply expert judgment alongside alert triggers.

This case study contributes to both industry practice and academic understanding by providing a replicable framework for assessing and maintaining organizational ALoSP. Practically, it offers a model for airports and other aviation service providers to integrate safety performance monitoring into everyday operations. Theoretically, it demonstrates how statistical methods and multi-level indicators can operationalize abstract safety objectives into measurable, actionable performance metrics. The approach bridges the gap between regulatory expectations and practical implementation, serving as an example of how safety performance management can mature into a dynamic, feedback-driven process rather than a static compliance exercise.

In summary, this research confirms that Split Airport has successfully established and maintained its ALoSP over a decade of monitoring through a robust and data-driven SMS framework. The achieved results validate the airport’s proactive approach to safety performance management, its capability to detect and address deviations promptly, and its alignment with international aviation safety standards. The methodology presented in this study, i.e., integrating SPIs, SPTs, alert levels, and achieved performance analysis, provides an example for other service providers seeking to enhance their safety performance and demonstrate compliance with ICAO’s performance-based safety standards.