Condividi, , Google Plus, LinkedIn,


Posted in:

The setting up of Milan APMS

Hypothesis, trigger and acceptability levels and intervention types are presented together with examples of the APMS implementation for SEA Company assets

An airport pavement

  The setting up of Milan APMS – Per la versione in Italiano:

Pavements are critical assets for airports. Compared to those of road networks they must withstand heavier traffic and higher tires pressures; moreover, intelligent asset management is required to ensure high functional and structural characteristics while not impacting its operativity.

This paper highlights the steps toward the implementation of the Milan Airports’ Pavement Management System. Pavements of Malpensa are the most heavily loaded in Italy since the airport serves about the 70% of the cargo movements in the State. Linate, with single runway layout, is the city strategic Airport and serves mostly the small to medium range of connections towards Europe.

The company underwent a complete survey for determining the functional parameters (PCI, IRI, Rut Depth, ETD, skid resistance) using a Multifunctional Vehicle and a structural campaign with HWD for runways, taxiways and aprons.

Data were combined for the determination of homogeneous sections, having similar distress rates, bearing capacity and traffic, which form the basis for a multi-year maintenance planning.

For the analysis purpose, a specific combination of softwares, ELMOD and Airports, were used aiming at minimizing the life-cycle cost of the assets while preserving high efficiency standards in compliance with international regulations.

Hypothesis, trigger and acceptability levels, intervention types are presented together with examples of the APMS implementation on a 3 million m2 assets. The Airport Pavement Management System (APMS) is a comprehensive structure of plans and procedures which enable a sustainable management of airport pavements.

The function of an APMS is to develop cost-effective strategies and, thus, direct decision-makers to maintain the pavements functionally over a given period. APMS follows a systematic procedure for determining needs and priorities, planning and schedules for maintenance, and required resource allocation. It delivers specific pavement network maintenance recommendations at an acceptable level of service after analyzing the collected information and optimizing the expenditures [1].

An airport pavement
1. Pavements are critical assets for airports

The concepts of an APMS are well outlined in FAA AC150/5380-7B [2], which uses the term Airport Pavement Management Program (PMP) and in TRB Circular E-C127 [1]. In 2015, the italian national civil aviation authority (ENAC) published a guideline for airport authorities for the implementation of APMS [3].

An APMS must supply airport authorities with a detailed picture of the present and future structural and functional capabilities of their airside pavements and predicted Maintenance and Rehabilitation (M&R) measures over the total design life of each pavement section, including the required budgets.

In order to be really effective, an APMS should not be reactive in the sense that it will respond to visual distresses only, but should be pro-active when it comes to predicting M&R requirements under the local climatic conditions, based on the actual loading by aircraft traffic as well as environmental [4].

Deterioration models are the key factor for effective Pavement Management Systems, helping out road agencies to assess the actual pavement condition and forecast future performance of the asset [5]. Historically, most agencies responsible for managing airport pavements have made decisions regarding maintenance and rehabilitation (M&R) based on experience and best engineering practices [1].

In the last decade several factors have encouraged the interest for smart asset management techniques. The availability of accurate technologies for pavements’ surveys and evaluation, like spatial technologies (e.g. GIS) and the rise of material prices, especially crude oil derivates (which worsen after pandemic) has made the use of pavement management system particularly popular and recommended.

Covalt et al. (2019) reviewed the state airport pavement management practices and airport pavement conditions in the United States from 2003 through 2016 [6]; findings shows that the APMS has become more diffused over the years with an adoption rate over 80% among US Airports.

In 2005, the Shanghai airport pavement management system (SHAPMS) was introduced to develop Hongqiao and Pudong International Airports. Chen et al. (2012) conducted a study on the airport pavement management system to improve and develop the most appropriate maintenance and rehabilitation techniques in the airport pavement system [7].

Different approaches have been taken into consideration to forecast pavement performance in Australia. Australia’s currently available systems include ground-penetrating radar surveys, falling weight deflectometer (FWD) surveys and high-resolution surface scanning [8]. 

Airside pavements
2. Pavements of Malpensa are the most heavily loaded in Italy

Brief history of Milan airports needs

The implementation of Milan airports’ APMS effectively started in 2019, but ideas and best practices had been studied for years since before the publishing of ENAC guidelines to which SEA, the airport authority, gave its contribution SEA manage Linate (IATA code: LIN) and Malpensa (IATA code: MXP) which represent important Italian airports in terms of passengers served and goods movements, relying on the spread catchment area which covers a great part of the Northwest of Italy.

Pavements of Malpensa are the most heavily loaded in Italy since the airport serves about the 70% of the cargo movements in the State; furthermore, the heavy movements have been steadily going up over the last years and the same is expected for the near future. Linate, with single runway, is the city strategic Airport and serves mostly the small to medium range of connections towards Europe.

In 2018, before the spreading of pandemic, Linate reached 115,300 movements serving 9.23 million passengers, while Malpensa reached 194,500 movements serving 24.7 million passengers.

The airside airports’ assets in terms of runways, taxiways, and aprons (a mix of flexible and rigid pavements) sum up to 3 million sqm; consequently in 2019 the executive board made the decision to introduce the APMS to optimize the management and put in place the optimal solution for maintenance and rehabilitation activities to minimize the life-cycle cost of the assets while preserving high efficiency standards in compliance with international regulations.

Besides the engineering enhancement of the pavements’ asset management, an important target of the implementation would be the compliance to the guidelines by the National Civil Aviation Authority (ENAC, Ente Nazionale per l’Aviazione Civile) about the adoption of APMS.

The task was assigned to a global airport consultancy group on a three-years contract, which entailed a detailed a survey for determining the functional parameters (PCI, IRI, Rut Depth, ETD, skid resistance) via Multifunctional Vehicle and a structural campaign with HWD for both runways, taxiways, and aprons.

The effort ended up in the development of a multi-year maintenance and rehabilitation plan, which is annually updated and revised by an Italian prestigious consultancy firm. 


The data collection program entailed two different categories: functional and structural parameters. Measurements of functional condition, consisted of:

  • International Roughness Index (IRI) as a measure of longitudinal evenness;
  • Rutting (Rd) as indication of transverse profile;
  • Detection of surface distress for the calculation of Pavement Condition Index (PCI);
  • Estimated texture depth (ETD).
Airports' view
3. Aerial view of the airport

For this purpose, a Dynatest multifunctional vehicle (MFV) was used, equipped with:

  • a high-resolution Road Surface Profiler (RSP) – Mark III system, meeting meet Class 1 precision and bias specifications as defined by ASTM E950, AASHTO R56-14, and AASHTO R48-10 [9, 10 e 11];
  • a 3D Pavement Imaging – Laser Crack Measurement System (LCMS). The LCMS, acquires 3D profiles and 2D images of the pavement surface using two high speed cameras, laser illumination, and advanced optics;
  • a front camera positioned inside the vehicle;
  • a Trimble AgGPS 132 antenna with a less than 1 m accuracy, positioned on the car top;
  • a distance measurement instrument (DMI) odometer, mounted at the rear left wheel.

The survey scheme was defined to cover the entire infrastructures area, through parallel test alignments of 3.75 m width and an overlap of 0.25 m to avoid the loss of information between two adjoining alignments. According to ASTM D5340, for the purpose of PCI calculation, areas with flexible pavements were divided in sample units of around 450 m2 (60 m length x 7.5 m width), while area with rigid pavements in sample units comprising 20±8 slabs [12].

The structural survey was carried out using a heavy weight deflectometer (HWD) and Ground Penetrating Radar (GPR) for the evaluation of the bearing capacity (expressed in terms of stiffness moduli), thicknesses and residual life of layers and subgrade, besides ACN and PCN values of each asset, according to the FAA AC 5370/11B [13].

The HWD was set to best simulate the movement of a wide-body aircraft: stresses of 3400 kPa were imparted by a series of masses weighing 700 kg dropped on buffers placed on a 300 mm diameter segmented circular plate; thus, the deflection basin of the pavement under the applied loads is measured using nine geophones, with an absolute accuracy better than 2% ±2 μm and a resolution of 0.1 μm.

The surveys with HWD were conducted, for each infrastructure, along parallel alignments in the longitudinal direction at distances of ± 3 m, ± 6 m and 12 m from the center line. The ±3 m and ±6 m alignments are tested to characterize the central and loaded portion of the infrastructure, affected respectively by narrowbody and widebody aircraft movements.

The 12 m alignment refers instead to the “unloaded” area. The longitudinal spacing of the test points has generally been assumed to be 60 m; each alignment is longitudinally offset by 30 m to the adjacent one, according to a quincunx arrangement.

Surveys were carried out following international standards (ASTM D 4694 – ASTM D 4695) and the subsequent calculation method relies on specific international standards (ASTM D 5858) with the aid of the software ELMOD6 [14, 15 e 16]. The entire procedure is framed into the FAA Circular AC 150/5370 – 11B [13].

Number and thicknesses of layers are inferred from previous design projects and the use of a Ground Penetrating Radar (GPR), provided with two antennas at different frequencies; the first antenna, operating at 600 MHz, was able to detect thicknesses up to a depth of about 1 m while the second, characterized by a higher frequency (1600 MHz), allowed a more detailed survey even if for only 50 cm depth.

Implementation plan 

Input Data

At first, the ENAC guideline on the implementation of APMS was thoroughly studied. This document was published in 2015 and entailed the collaboration of Italian’s main airport representatives and University Professors coordinated by ENAC Technical Committee.

APMS implementation flowchart
4. APMS implementation flowchart

It defines in detail the main aspects and expected contents of an APMS, referring but not limited to the role of the Manager, the surveys’ schedule, the requested parameters, the traffic data format, the database content, and the expected output. Accordingly, the Company built the APMS to be compliance to this framework; once the needs were understood it was possible to set up the implementation plan, which can be summarized in the flow chart in Figure 4.

The first step was the data collection as described in the previous section. Malpensa airport has two parallel twin runways: 17L35R and 17R35L (3,920×60 m) both made of asphalt concrete and heads in cement concrete (except 35L head). A series of taxiways, mostly in flexible pavements, spread over an area of around 820,000 m2 connecting the runways to two terminals made of cement concrete which extend respectively to 680,000 and 320,000 m2.

Data were collected with surveys in several night sessions, divided in three major groups: runways, taxiways, and aprons. ETD and Rd were calculated with a rate of 10 m, IRI with a rate of 100 m, and PCI for sample units.

In order to comply with ENAC guidelines, SEA plans to repeat data acquisition according to the daily number of aircraft movements; considering the actual traffic, MFV data are planned to be reacquired at minimum after 36 months while HWD tests will be repeated after 48 months; actually, these represent minimum frequencies as more tests are foreseen for mid-low lasting sections to preserve operation safety and reduce lifecycle costs.

Traffic was carefully considered both in terms of volume and spectrum, as its knowledge is essential not only for predicting the evolution of degradation, and therefore the performance of the pavements, but also for defining the M&R priorities.

Accordingly, the spectrum was first categorized into the six ICAO Annex 14 categories based on wingspan [14], then cataloged in ten different classes, each of which represented by the aircraft with the highest MTOW who was assigned the whole number of passages of all aircraft belonging to the relative class.

Each reference aircraft is then considered to damage distinctively the pavements, based on its axle configuration and loads, tyre pressures, number of movements and pass to coverage ratio on the infrastructure section. For residual life evaluations, the software used for the backcalculation considers all these aspects for calculating the cumulative damage as the sum of different contributions. 

Homogeneous sections

An important step towards the implementation was definition of homogenous sections; they represent areas having similar properties or functional/structural conditions. This step was carried out longitudinally according to deflections, subgrade and bound layers conditions, assets layout, traffic volumes, and pavement condition indexes.

  • Rating and triggers for IRI
    5A Rating and triggers for IRI
    5A. Rating and triggers for IRI [18 and 19]
  • Rating and triggers for PCI
    5B Rating and triggers for PCI
    5B. Rating and triggers for PCI [6]
  • Rating and triggers for Rd
    5C Rating and triggers for Rd
    5C. Rating and triggers for Rd parameter [6]
  • Triggers for Elastic Moduli
    5D Triggers for Elastic Moduli
    5D. Rating and triggers for Elastic Moduli [20]

Transversely, assets were divided into a 30 m width central loaded section, 15 m width lateral unloaded sections and shoulders. Beside Aprons, assets were divided into 75 and 400 sections for respectively Linate and Malpensa. 

Software computation and strategies

This section refers to the core part of the APMS. The backcalculation of Elastic Moduli was performed with ELMOD6, by using the method of equivalent thickness (MET) which is based on Odemark’s assumption: deflections of a multilayered pavement system with various moduli and layer thicknesses can be obtained using a single layer of thickness H and modulus E.

Elmod makes use of Odemark-Boussinesq MET and the radius of curvature method: at first, the subgrade material properties, stiffness and nonlinearity are calculated using the deflections from the outer sensors; then the radius of curvature from the central sensors are used to assess the stiffness of the upper pavement layer.

The stiffness of the remaining layers is then calculated based on the overall pavement response to the applied load. The software was set to alternate two times between the absolute and percentage RMS to better fit both inner and outer geophones. The backcalculation relies on layers thickness data, defined after a detailed GPR and cores survey.

The backcalculated pavement structure in terms of thicknesses and Moduli was then directly imported in Airports for the purpose of calculating the M&R plan; the critical alignment (i.e. the one showing worst bearing capacity) was chosen as representative for the entire section. Before running the calculation, a fundamental step was the definition of the criteria which trigger the rehabilitation; accordingly, threshold values for the different functional and bearing capacity parameters were established: attention (trigger) and critical thresholds.

Once an attention threshold is exceeded, the software chooses to apply the intervention, anticipating it as far as it deems economically convenient. It follows that a section having parameters higher than triggers will never be assigned an intervention.

The critical threshold is the one beyond which the software certainly applies the intervention. It follows that a section will never see any of its characteristic parameters go beyond the critical threshold under defined budget constraints. The thresholds were identified with reference to international regulations.

As for IRI a correlation between the Riding Comfort Index and IRI itself was used [18 and 19]. The main parameters for which to foresee the thresholds are IRI and Rd (not considered for rigid pavements), PCI and Elastic Moduli. To determine the structural pavement response, Airports uses the damage predictions of the incremental-recursive rational methods, in terms of calculated stresses and strains for each layer.

In other words, the method uses a mechanistic model to calculate the initial stress-strain condition, it derives the variation of this configuration (“incremental”) due to calculated stress-strain and applies it as a new input, to then repeat the calculation (“recursive”).

The setting up of Milan APMS
6. Parameters used in the incremental-recursive rational method

The process is repeated until reaching the critical thresholds. The calculation interval was set to one year, considering an average season with a pavement temperature of 19 °C.

An important feature is therefore considering that parameters decay is a function of the previous performance stage rather than a linear function of the load repetition number, hence more realistic.

Accordingly, the damage was hence calculated in relation to the effective reduction of the Elastic Moduli as follow:

The setting up of Milan APMS


  • A, R_ref, E_ref, k_1, k_2, k_3 = constants;
  • R_T (Response Type) = type of response to loads considered in the calculation (strain was selected);
  • R_ref = reference value for R_T;
  • E_ref = reference module.

Figure 6 shows the parameters adopted in the analysis, deriving from an adaptation to the case of decay models obtained from the literature.

The decay of the pavement functional parameters was implemented in the form of equations function of the service life. It is referred to groups of homogeneous sections by type of use (e.g. loaded/unloaded).

It must be noted that empirical models were used alongside the rational ones, according to the “use if worst” criterion: the software determines the progression over time of performance indicators both through the rational and empirical approaches and selects the more precautionary between the two.

The setting up of Milan APMS
7. Parameters used in the empirical model

This represents another “safety factor” which allows considering factors not contemplated by the rational models, like aging and climatic effects.

The empirical models are in the form:

The setting up of Milan APMS


  • a = parameter;
  • a_0 = initial value;
  • age = time in years;
  • A and B = constant.

Specifically, it was assumed as shown in Figure 7. The laws previously discussed are defined for each material type considered which constitute the pavements. At last, maintenance alternatives are defined as composition of the different materials and relative thicknesses; they are associated with cost per square meter.

Particularly significant are the improvements: any simulated maintenance alternative is associated with a regain of values proportional to the rehabilitation depth and the material types, thus closely linked to layers performance included in the work technical specifications, allowing to simulate also the effects of innovative materials with enhanced performances.

This allows Airports in defining which is the best M&R to put in places to be compliance with set boundaries and in accordance with the selected strategy. By way of example, any resurfacing alternative would raise the PCI to 100 and Rd to 1 mm, while IRI would be progressively improved ranging from a surface course mill and inlay to a full asphalt layer rehabilitation. 

Output and mapping

Among the five different optimization strategies available in Airports, the cheapest solution was selected for the analysis; this enables to obtain the minimum cost to the Agency for each section of the network, whilst maintaining the parameters in compliance with their respective critical triggers. The analysis period considered was ten years.

Figure 8 shows by way of example the plot of optimization for a taxiway of Malpensa. The output suggests two mill and inlay interventions of the surface course in 2026 and 2030.

The setting up of Milan APMS
8. Plot of optimization results for a taxiway of Malpensa

In Figure 8, on the top left corner, the predicted decrease in bearing capacity is plotted against time in years, i.e., the loss in equivalent thickness compared to the initial level (for example a loss of 20 mm means it takes 20 mm of overlay to regain the initial bearing capacity); this is calculated instantly based on structural information and traffic, together with the materials deterioration models as defined in the parameter setup.

On Figure 8, on the top right corner the plot shows the equivalent Elastic Moduli of the asphalt layers which obviously increases at any intervention following the substitution of old and fatigued layer with a new asphalt concrete. IRI, Rut Depth as well as PCI regain their best values after resurfacing as it allegedly removes irregularities and distresses. Finally, remaining WC life indicates an expected terminal year of the surface layer, in case no other condition type will result in a shorter lifetime.

For sections with relative low traffic the modeled conditions may not result in any critical states during the analysis period, but due to other influences from climate and surface weathering there will be a limit to the service life, indicated by the value in this data field.

Data were then exported and managed with a Excel spreadsheet and a Geographical Information System for a map representation. Figure 9 shows PCI values on a portion of Malpensa airport; values are presented for each homogeneous section the assets were divided. 


SEA APMS is then a complete system, allowing for strategic asset management. It:

  • relies on a comprehensive database composed of assets’ geometric features, traffic, structural condition and typical functional parameters (IRI, Rd, ETD), distresses map, PCI, and previous maintenance activities (and related parameters);
  • allows a monitoring of the condition and deterioration making use of historical information and a prediction of performances through the calibrated models used;
  • permits the evaluation of the effects of different M&R strategies at various stages in the life of the pavements, different scenarios in relation to budget constraints, and a plan of future M&R activities; the software takes also into account interest rates, user and agency costs;
  • includes the possibility to monitor defects and archive routine maintenance activities even if not modelled.
The setting up of Milan APMS
9. PCI representation on a portion of Malpensa airport

The continuous update of the pavements inventory and the systematic monitoring of performance make it possible to prioritize and plan the most appropriate maintenance and rehabilitation activities, also on the basis of a continuous comparison with available budgets and expected performance.

The continuous implementation of new data will allow SEA to calibrate, for each type of asset, material and pavement type, specific performance decay laws, to improve accuracy and reliability.


[1]. S.L. Tighe, M. Covalt – “Implementation of an airport pavement management system”, Transp. Res. Circular (E-C127), 2008.

[2]. FAA AC 150/5380-7B, 2014 – Airport Pavement Management Program (PMP).

[3]. ENAC – Airport Pavement Management System – “Linee Guida sull’implementazione del sistema di gestione della manutenzione delle pavimentazioni”, ENAC, Roma, Italy, 2015.

[4]. A. Marradi, M. Tamarozzi – “Pavement management system implementation at Rome Fiumicino international airport”, Airport Review, 2010.

[5]. S. Alberti, M. Crispino, F. Giustozzi, E. Toraldo – “Deterioration trends of asphalt pavement friction and roughness from medium-term surveys on major italian roads”, International Journal of Pavement Research and Technology, volume 10, Issue 5, 2017.

[6]. M. Covalt, L. Raczkowski, M. Fisher – “A review of state airport pavement management practices and airport pavement conditions in the United States from 2003 through 2016”, Pavement and Asset Management” Proceedings of the world conference on pavement and asset management, June 12-16, Baveno, Italy, 29-34, 2019.

[7]. W. Chen, J. Yuan, M. Li – “Application of GIS/GPS in Shanghai Airport pavement management system, Proc. Eng. 29 2322-2326, 2012.

[8]. Md T. Miah, E. Oh, G. Chai, P. Bell – “An overview of the airport pavement management systems (APMS)”, International Journal of Pavement

Research and Technology, 13.581-590.10.1007/s42947-020-6011-8, 2020.

[9]. AASHTO R 48, 2010 – Standard Practice for Determining Rut Depth in Pavements.

[10]. AASHTO R 56, 2014 – Standard Practice for Certification of Inertial Profiling Systems.

[11]. ASTM E950/E950M-09, 2018 – Standard test method for measuring the longitudinal profile of traveled surfaces with an accelerometer established inertial profiling reference.

[12]. ASTM D5340-20, 2020 – Standard Test Method for Airport Pavement Condition Index Surveys.

[13]. FAA AC 150/5370-11B, 2011 – Use of nondestructive testing in the evaluation of airport pavements.

[14]. ASTM D4694-09, 2020 – Standard test method for deflections with a falling-weight-type impulse load device.

[15]. ASTM D4695-03, 2020 – Standard guide for general pavement deflection measurements.

[16]. ASTM D5858-96, 2020 – Standard guide for calculating in situ equivalent elastic moduli of pavement materials using layered elastic theory.

[17]. Icao Annex 14, 2022 – Aerodromes, volume I, Aerodromes Design and Operations.

[18]. CROW 06-01“Survey and Airport Pavement Roughness Assessment”, State of the Art Study 2006.

[19]. “Guidelines respecting the measurement and evaluation of airfield pavement surface roughness”, Trasport Canada, International Aviation and Technical Programs Branch, 2005.

[20]. “Design Manual for Roads and Bridges”, CD 227, Design for pavement maintenance, March 2020.

  Per la versione in Italiano: