Study and Characterization of New KPIs for Measuring Efficiency in Urban Loading and Unloading Zones Using the OEE (Overall Equipment Effectiveness) Model

Featured Application

This study proposes the use of the OEE industrial model for the evaluation of the efficiency of loading and unloading zones (LUZs). For this purpose, manual observations were conducted in five specific urban zones of to characterize and establish the parameters and variables necessary for calculating OEE for the evaluation of the effectiveness of LUZs. A new Key Performance Indicator (KPI), the weighted time of occupation, was defined and integrated into the model, serving as a useful tool for identifying inefficiencies in the occupation of these zones. These contributions provide urban governance with new quantitative and precise metrics to evaluate and manage LUZs, addressing current limitations in the management of urban goods loading and unloading activities. Expected improvements include up to an 11% increase in efficiency through the elimination of unauthorised parking and a 36% improvement in compliance with allowed parking times when measures aligned with the OEE model are implemented.

Abstract

The use of LUZs in urban environments is a critical factor for ensuring efficient vehicle mobility in cities. Poor utilisation of these zones can generate negative externalities, such as double parking or illegal occupation of pedestrian crossings or garage doors. The purpose of the study is to provide city governance with a methodology based on the OEE model to evaluate the efficiency of individual zones or sets of zones and to inform decisions that improve their use without disrupting the coexistence with other city users. To validate the methodology, all deliveries made in selected areas of the city of Zaragoza over the course of one month were studied. The results of the study reveal a considerable loss of efficiency and some recommendations are proposed achieve a better use: only 51.44% of deliveries used the LUZs correctly, and the total OEE ratio was just 0.37. This low level of efficiency is due to the incorrect use by delivery drivers, who often use LUZs as parking spaces, and the illegal occupation of the zones by unauthorised private vehicles.

1. Introduction

The vast majority of cities were not designed to accommodate the distribution of goods by trucks or delivery vans. Managing vehicle mobility on urban streets, while balancing the needs of pedestrians, personal mobility vehicles (PMVs), buses, private cars, and delivery transporters, is a critical factor that municipal authorities must manage. Parking in dense urban areas is a major challenge for last-mile logistics. The scarcity of parking spaces and policies that do not address the needs of commercial vehicles often lead vehicles to park illegally, according to Ghizzawi et al. [1] The solution currently in place in cities for loading and unloading of goods to retail stores is to reserve LUZs on the streets, but in most cases, these LUZs are neither located in the right place nor are they of adequate size for the use required by transporters, nor do they have adequate reservation times. Alho et al. [2] argue that optimising the number, location, and use of LUZs can lead to substantial improvements in mobility indicators, particularly in reducing double parking, an externality that disproportionately affects traffic and the efficiency of the urban transport system. Incorrect location, incorrect sizing, and poor choice of reserved times, as well as the time allowed for each unloading, standardised for all areas, create very harmful externalities for urban mobility, such as double parking and illegal occupation of sidewalks, garage doors, crosswalks, or street intersection islands. As suggested by Amaya et al. [3], when commercial vehicle drivers cannot find legal parking nearby, they resort to double parking or park in illegal spaces to ensure on-time delivery. Alho et al. [4] indicate that there are no clear criteria for choosing the location, size, and timing of LUZ reservations.

This study proposes the application of the OEE model, typically used in industrial and production environments, as a KPI, considering a LUZ as if it were a production element in order to evaluate operational efficiency and identify the factors that reduce productivity in that area. This information can be very useful for city governance, which can draw valuable conclusions about the performance of established areas. Comi et al. [5] propose a methodology for verifying and designing the number of delivery bays needed based on demand but do not indicate how to measure their efficiency. Although Muñuzuri et al. [6] assert that, in the context of their research, they believe that decisions about the number, size, and location of loading zones should be made for individual streets rather than for sectors or areas, the application of the OEE model proposed in this article is very versatile and flexible, as it can be applied to a zone or a set of zones and over any time period that can be defined, whether in hours, days, weeks, or months.

There is an abundant body of literature on urban distribution of goods (UDG); however, studies focusing specifically on LUZs are scarce. The existing references tend to focus research on access routes to these zones [7,8,9] and the walking distance from the LUZ to the supply point [10,11]. Castrellon et al. [11] suggest that a maximum distance of 75 m provides a good level of accessibility service for cargo. Other authors, such as Dezi et al. [12] and Muñuzuri et al. [6], suggest a distance of 50 m, while McLeod et al. [13] and Ochoa-Olán [14] suggest up to 100 m. Other studies focus on booking spaces in LUZs [15,16,17,18] to avoid externalities and save on energy consumption when searching for parking. Several articles have linked the OEE model to UDG, providing the foundation for the present article.

Table 1. KPIs found in the literature.

KPIs Related to the Delivery Process	Source
Average parking duration	[1,2,10,11,15,18,19]
Vehicle arrival rate	[5,14,16,18]
Average delivery time	[1,4,12,13,14,20,21]
Average waiting time (cruising)	[1,3,5,10,16,17,19]
Number of deliveries made	[3,11,12,14,15,20,22,23]
% of deliveries during peak vs. off-peak hours	[12,20]
Deliveries not made on time	[24,25]
Occupancy rate of places s/total available	[2,5,11,15,16,17,18,19,20,21,24,26]
Type and capacity of vehicles	[1,2,3,10,12,13,17,18,26,27]
Number of vehicles entering the area	[1,2,3,10,12,13,17,18,26,27]
Day of the week	[18]
Time of arrival/departure	[1,16,18,25,26,28]
Professional activity (sector)	[1,4,17,25]
Type of cargo, weight, and volume	[5,18,25]
KPIs related to LUZ	Source
Number of delivery bays required	[2,4,5,14,15,16,20,27,29]
Size of loading bays	[2,4,11,14,15,16,17,20,27,28]
Location and availability of LUZ	[2,4,10,11,12,13,14,15,16,24,25,26,30]
Walking distance	[3,10,11,12,16,20,29,31]
Frequency of use of loading areas (rotation)	[5,18,20,26]
KPI auxiliaries	Source
kg per person consumed in an area	[32]
Average emissions per vehicle	[4,11,18]
Impact of external events	[24]
Parking costs	[3,4,17,24]
LUZ demand	[11,13,16,19,22]
LUZ parking space reservation	[13,16,17]
KPIs related to externalities	Source
Number of double-parked vehicles	[2,3,13,15,20,21,24]
Location of double-parked vehicles	[2,10,12,16,18,27,33]
Illegality rate by vehicle type and duration	[2,4,17,27]
Percentage of unauthorised vehicles in LUZ	[27]

below shows the KPIs identified in the literature for evaluating processes in LUZs, divided into four categories, according to the authors’ criteria:

The authors highlight that the KPIs most frequently found in the literature are those related to the delivery process, average parking time, average delivery time, the number of deliveries made, occupancy rate of available spaces, the type and capacity of vehicles, and the number of vehicles entering the area. Those related to LUZs include the number of delivery bays required, the size of the loading bays, and the location and availability of LUZs. In any case, no indicator highlights the numerical quantification of illegal parking in a LUZ, either by private vehicles or commercial vehicles that exceed the time allowed. These KPIs will be the components of the OEE calculation formula.

After analysing all the KPIs in the table, none were found that allow the efficiency of a LUZ to be measured from a city governance perspective. Therefore, the objective of this study is to define a new indicator based on the OEE model that enables the evaluation ofthe use of a LUZ or a set of LUZs over a specific time horizon—hourly and daily in this study. Its implementation will provide a methodology to support decision-making regarding whether the location, size, and reservation hours for loading and unloading operations are appropriate or not. Wilson et al. [20] assert that the same space will be less valuable during periods of lower activity. With the application of the proposed model, these periods can be identified. The analysis can also lead to actions that improve the use of the area, which will help city governance make urban coexistence among various users more compatible. Alho et al. [4] discuss the variable size of LUZs but highlight the difficulty of assessing whether their dimensions are appropriate. As demonstrated in the conclusion of this work, illegal occupation by unauthorised vehicles and the excessive time transporters spend unloading are the two factors that most hinder LUZ efficiency. The contribution of this study is to quantify the impact of these problematic practices.

The article is organised as follows: Section 2 describes the OEE model and its application to the DUM. It then addresses the characterisation of loading and unloading zones in the city of Zaragoza, details the five zones that have been chosen for study and the reasons behind their selection, and describes thefieldwork from which the data were obtained. Section 3 analyses the results of applying the model in all its aspects and possibilities for analysis, and Section 4 interprets the results and makes recommendations for city governance. Finally, Section 5 presents the conclusions and possible future lines of research.

2. Materials and Methods

Before explaining the study’s analysis methods, it is useful to explain how data collection was carried out in the field study. The data were collected directly by the authors of the study through direct observation during the 21 days of the month under study and during the total number of hours reserved for loading and unloading. The different classifications of vehicle types, double parking, illegal parking, and parking times were recorded and categorised directly by the authors at the time of data collection.

However, there is a bias in the identification of the distribution sector, as the observations only recorded the type of vehicles and the arrival and departure times, not the activity carried out by the delivery person, which could only have been ascertained through direct surveys. Associated with the sector, we could also consider as bias a lack of knowledge about the type of goods being handled and the ability to perform more objective analyses of operating parameters such as kg or m³.

2.1. The OEE Model

Moving into the objective of this article, we will now define the OEE model for evaluating the selected LUZs. OEE (Overall Equipment Effectiveness) is a concept commonly used in the industrial sector, introduced by Nakajima (1988) [34] as part of total productive maintenance (TPM) to measure the productivity and efficiency of equipment [35]. It is a productivity ratio between real production and what could ideally be produced [36], which in our case could be likened to the unloading carried out in the time reserved for loading and unloading tasks, compared to what could have been achieved. The objective of this model is to increase productivity and reduce losses in time, speed, and quality. Castrellon et al. [26] suggest that future research could be developed using analytical methods to examine the statistics collected and propose techniques for standardising performance measures given the multiple scales and metrics encountered. Alho et al. [2] recommend investigating how different levels of compliance with standards affect the efficiency of the system and what policies can encourage better compliance. Castrellon et al. [11] also suggest in another study that there is a lack of analysis based on objective data and highlight the inefficiency of public policies.

In the literature reviewed, Corrales et al. [36] conduct a systematic review of the literature and provide an overview of the different approaches to OEE. This study identifies the logistics and transport sector as one with potential for development, since all current approaches focus on industry, maintenance, and manufacturing and are related to terms such as lean manufacturing, improvement, implementation, reliability, design, and optimisation. García-Arca et al. [37] conducted a study of KPIs focused on transport routes to improve road transport efficiency. Muñoz-Villamizar et al. [34] investigated the use of the OEE model to evaluate the effectiveness of urban freight transport systems with a case study of the activity of two companies. Les et al. [23] applied the model to measuring a transport route, but no study has proposed measuring the effectiveness of a LUZ.

In the logistics sector, there is no defined methodology for calculating OEE. Studies that address the OEE model for transport [38] do not compare vehicles to industrial equipment in the traditional sense; rather they analyse the efficiency of individual routes [38] or the truck loading and unloading process [36]. These studies highlight that efficiency depends not only on the vehicle but also on the driver and external factors. Les et al. [23] introduced OEEM, a metric specifically designed for OEE analysis of delivery routes. In no case has the model been found to apply to LUZ. Teodorovic et al. [39] indicated that it would be necessary to study in more detail segmentation, occupancy rates, average parking duration, indirect parking problems, and the enforcement of parking violations. Marcucci et al. [40] suggested that future research could also investigate the spatial optimisation of loading bay locations in relation to demand and supply in a given sector. Hence, research such as that presented in this article provides an important foundation for developing evaluation mechanisms for urban loading and unloading areas.

2.1.1. Factors That Make up OEE and Their Application to LUZ

Availability

Drawing an analogy with the manufacturing sector, this factor reflects the time that a machine, in our case, a LUZ, is actively in use compared to the total time it could be producing. Considering a 24-hour day, only a fraction is reserved for loading and unloading. From those reserved hours, scheduled downtime must be deducted, such as reserving the LUZ for removals, diesel supplies to communities, or external city events. Therefore, this factor reflects the measurement of the time that the LUZ is in use, i.e., occupied within the reserved time, excluding scheduled downtimes.

In addition to the KPIs or variables currently used to evaluate loading and unloading areas, and in order to characterise this availability factor, it has been necessary to define a new KPI that will be called weighted occupancy time (t_p). It is calculated as follows for a vehicle i:

L_t = total length of the LUZ

L_i = length of the vehicle occupying the area

T_t = Total time reserved for loading and unloading in the area

t_j = time that is occupying the zone

L_i/L_j = % zone occupation

L_i/L_j × t_j = weighted occupation time (t_p)

$$\sum _{\mathrm{j}=1}^{\mathrm{N}}{\mathrm{t}}_{\mathrm{p}}=\mathrm{total}\mathrm{occupation}\mathrm{of}\mathrm{the}\mathrm{area}$$

The interpretation of this new weighted time indicator is that its sum reflects the spatial-temporal contribution of each vehicle to the LUZ, i.e., how much time the vehicle’s length occupies the area, and its sum is compared with the Total Time available to assess the availability of the area. To calculate the indicator, standard lengths have been taken for each type of vehicle: private car (4 m), small van (4 m), delivery van (4.5 m), large volume van (5.8 m), light truck with a maximum authorised mass (MAM) of 3.5 tons (6.2 m), and light truck with a MAM of 7.5 tons (7.7 m).

An availability indicator close to 1 means that it has been permanently occupied during that time, either legally or illegally. On the other hand, a low result would indicate a long period of non-occupation.

$$\mathrm{Availability}=\frac{{\displaystyle \sum _{\mathrm{j}=1}^{\mathrm{N}}}{\mathrm{t}}_{\mathrm{p}}}{{\mathrm{T}}_{\mathrm{t}}}$$

Efficiency

This factor measures the time that the area has actually been occupied by commercial DUM vehicles, i.e., the weighted occupancy time of non-commercial and unauthorised vehicles is discounted, leaving only goods delivery vehicles to be considered. The calculation of this indicator takes into account the reduction in the weighted time of use of unauthorised vehicles t_x over the total time T_t.

$$\mathrm{Efficiency}=1-\frac{{\mathrm{t}}_{\mathrm{x}}}{{\mathrm{T}}_{\mathrm{t}}}$$

A value close to 1 would indicate that there has been no illegal occupation in the area, while a low value would indicate abuse by unauthorised vehicles. The indicator shows the length of time the area has been occupied by commercial vehicles.

Quality

In cities, there are rules and regulations regarding the time allowed for loading and unloading operations. To correctly quantify the time exceeded, the first 30 min are considered correct use, as indicated in the Zaragoza City Council’s Urban Mobility Ordinance, and the excess time is considered illegal occupation of the LUZ. This factor quantifies the excess time that the area has been occupied by commercial vehicles that have exceeded the time allowed by the city authorities, i.e., the weighted occupancy time of commercial vehicles is subtracted, which, once the legal time has been exceeded, becomes illegal occupation (t_oi). The calculation of this indicator takes into account the reduction in the weighted time of illegal occupation due to exceeding the time allowed.

$$\mathrm{Quality}=1-\frac{{\mathrm{t}}_{\mathrm{oi}}}{{\mathrm{T}}_{\mathrm{t}}}$$

A value close to 1 would reveal that commercial vehicles comply with the time allowed for loading or unloading, while a low value would indicate abuse by commercial vehicles which use the LUZ as a free parking space. The indicator measures the quality of deliveries made correctly, in their corresponding zone, and within the time allowed.

Calculation of the OEE for the LUZ

With the three factors defined, the generic formula of the OEE model is applied As follows:

OEE = Availability (A) × Efficiency (E) × Quality (Q)

The OEE can be measured for a single area or a group of areas. It can also be measured over any time period one wishes to define. It should be noted that in this study, the OEE KPI proposed could be used when there is a need to evaluate an area from different points of view, for example, an OEE can be calculated by hours of occupancy to see if there are any times that suggest the need for improvement actions, or measured by day of the week to see how it behaves with weekly seasonality, or measured over a long period such as the entire month evaluated to obtain consolidated data. It will also be interesting to analyse the different behaviour of each area in the morning compared to the afternoon.

Figure 1 shows the OEE model applied to a LUZ in graphical form for better understanding:

2.2. Characterisation of Selected LUZs

The analysis of the loading and unloading zones in the city of Zaragoza is shown below. This city was chosen because it is a pioneer in mobility initiatives, such as the implementation of an Urban Consolidation Centre (CCU) in combination with PMVs for UDG [41], distribution pilots using autonomous robots [42], autonomous buses [43], and collaborations with the General Directorate of Traffic (GDT) [44,45], which has selected it as a test laboratory. Additionally, the city participates in national and European projects such as URBANDUM [46] and DISCO. Based on the 2018 Sustainable Urban Mobility Plan (PMUS) [47], in the section on UDG, and using open data from Zaragoza City Council, 779 LUZs have been identified. [48,49].

The average length of LUZs in the 14 areas defined in the PMUS in the urban area of the city is 19.4 m and the average time reserved is 7.7 h. Considering that this includes recently urbanised areas with very large spaces and generous times for loading and unloading, the central area, with 19.6 m and 6.5 h, seems to be a representative area for the study. In this area, 108 LUZ were identified, representing 13.86% of the 779 areas.

Article 93 of Zaragoza’s 2024 Urban Mobility Ordinance [50] establishes the maximum time allowed for loading and unloading goods as 30 min in general terms.

This article is based on field research with direct observation in five LUZ in the central area. These areas were chosen for their representativeness:

Zone 1:

One-way street with free parking, 9.2 m wide in total and 5.3 m between vehicles, allowing for comfortable double parking. In addition, opposite the LUZ there is a ‘dark store’ (an urban warehouse supplying stock to Glovo delivery drivers using PMVs), which generates high demand for supplies. The operating hours for this areaare from 9 a.m. to 12 p.m. and from 2 p.m. to 5 p.m. It is a 13-metre area, corresponding theoretically to space for threevans or trucks. The entire street, excluding garage entrances and pedestrian crossings, including both pavements, is 168 m long.

Zona 2:

A six-lane avenue, with three lanes in each direction, features regulated parking that also allows frequent double parking. The reservation time for loading and unloading is from 8 a.m. to 11 a.m., only in the morning. The avenue only has vehicles parked on one side, as the other side is dedicated to a bike lane, and has a total length of 173 m reserved for parking. The length of the loading area is 10 m, for two theoretical unloading spaces.

Zones 3, 4 y 5:

One-way street with free parking and no double parking, with cars on both sides of the street, with a total length of 165 m dedicated to parking. The three areas are: a first area of 18 m, with 4 theoretical spaces assigned and reservation times from 7 a.m. to 12 p.m. and from 2 p.m. to 5 p.m., a second 10 m zone for two theoretical spaces, from 9 a.m. to 12 p.m. and from 2 p.m. to 5 p.m., and a third 8-metre zone (reduced due to the allocation of a few metres to a bar terrace) for a single theoretical space, with the same hours as the previous zone. On this street, illegal parking occurs in front of garage doors, pedestrian crossings and at the intersection with the adjacent street. Figure 2 below shows a map of the area with the LUZ marked in red:

The five areas have been selected for their variety, both in terms of the width of the streets where they are located and the possibility of double parking as an alternative to LUZ, as well as in terms of LUZ sizes. It is also taken into account that one of the areas is in a regulated parking zone. The fact that they are so close together makes it easier to collect data simultaneously in all five areas. They are considered to offer a representative sample of the city’s LUZs.

Table 2. Characteristics of each street and each zone.

Zone	Lengh Zone	Street Width (Lanes)	Parking Type	Reservation Time	Reservation Hours
Zone 1	13	1 with double width	Free	9–12/14–17	6
Zone 2	10	6 lanes	Regulated	8–11	3
Zone 3	18	One line	Free	7–12/14–17	8
Zone 4	10	One line	Free	9–12/14–17	6
Zone 5	8	One line	Free	9–12/14–17	6

compares the zones, highlighting the differences in the characteristics of each street and each zone.

At this confluence of areas, 43 retail establishments have been identified as recipients and/or senders of goods, 27 on the avenue, 4 on the wide street, 7 on the narrow street, and 5 on the street that completes the set of zones. Of these, 8 are food retailers, 18 are general stores, 1 is a supermarket, 10 are Horeca channel, 2 are pharmacies, 1 is a ‘dark store’ and 3 are car repair shops, which offers a standard representation of a typical service area.

Ochoa-Olan et al. [14] proposed that the use of numerical methods to estimate the location, number, and size of truck parking spaces should be complemented by empirical studies in order to balance mathematical calculations. Therefore, this paper presents the field study carried out during the entire month of May 2025 through direct observation, recording all entries and exits from the indicated areas. The month of May was chosen as it is a representative month, with full activity during the school year due to its impact on the city’s mobility and the absence of holidays, and is therefore considered to be a time of coexistence between goods transporters and citizens in a normal situation. A total of 1582 downloads of all types and with all types of vehicles were collected in the five LUZs.

3. Results

3.1. Global Indicators

Before addressing the results to apply them to the OEE model, we will explain the overall indicators to give a general idea of the use of LUZs in the period analysed.

Table 3. Loading and unloading operations observed in the areas.

Day		2	5	6	7	8	9	12	13	14	15	16	19	20	21	22	23	26	27	28	29	30	Total
Zone 1	M	10	22	18	21	14	11	15	19	19	10	11	15	13	19	9	10	13	8	19	10	9	295
	A	7	10	14	10	9	5	11	6	10	7	5	10	11	10	9	7	6	6	10	6	9	178
Zone 2	M	10	18	13	20	13	15	14	15	13	19	15	14	14	13	13	10	11	12	13	19	13	297
	A
Zone 3	M	14	12	13	0	9	11	12	9	12	10	11	12	12	12	17	15	20	11	16	9	8	245
	A	8	7	12	0	8	9	7	7	8	10	9	7	9	8	8	8	11	5	8	10	10	169
Zone 4	M	3	4	7	9	5	7	4	5	3	2	6	4	7	3	5	3	4	5	3	2	5	96
	A	4	2	4	2	5	2	3	4	4	5	2	2	8	4	5	4	9	6	4	5	6	90
Zone 5	M	6	3	2	5	7	5	3	7	7	6	5	3	12	7	7	6	3	4	5	6	4	113
	A	4	6	6	6	5	6	7	3	4	2	6	6	7	4	5	3	5	5	4	2	3	99
Total	M	43	59	53	55	48	49	48	55	54	47	48	48	58	54	51	44	51	40	56	46	39	1.046
	A	23	25	36	18	27	22	28	20	26	24	22	25	35	26	27	22	31	22	26	23	28	536

provides a summary of the loading and unloading operations observed in the areas.

A total of 1582 loading and unloading operations were recorded across the entire area, of which 1092 were in the LUZ areas under analysis. The remaining 490 operations occurred outside these designated areas: 329 were carried out in double rows, and 161 were carried out illegally (in front of garage doors, on pedestrian crossings, or on intersection islands). This can be seen in

Table 4. Data observed in total operations.

Day	2	5	6	7	8	9	12	13	14	15	16	19	20	21	22	23	26	27	28	29	30	Total
Double parking	9	32	23	24	14	15	21	26	12	14	15	20	19	12	13	9	13	6	12	14	6	329	20.80%
Illegal Ocupationl	6	4	7	6	4	10	8	8	14	3	10	6	15	14	4	6	8	7	14	3	4	161	10.18%
Zone CyD	51	48	59	43	57	45	47	41	54	52	45	47	59	54	61	51	61	49	59	52	57	1092	69.03%
Total	66	84	89	73	75	70	76	75	80	69	70	73	93	80	78	66	82	62	85	69	67	1.582

.

As the article focuses on calculating OEE for LUZs, we will disregard double parking and illegal occupation operations, and only analyse data observed in LUZ operations during their authorised hours. Data referring to the number of operations can be found in

Table 5. Data observed in operations only in the loading and unloading areas.

Day	2	5	6	7	8	9	12	13	14	15	16	19	20	21	22	23	26	27	28	29	30	Total
Correct use of the LUZ	26	26	24	31	33	21	25	22	30	25	21	25	31	29	29	26	29	24	35	25	31	562	51.44%
Unauthorized vehicle	14	11	26	7	14	13	10	8	15	17	13	11	19	15	21	14	24	14	15	17	16	314	28.77%
Excess time	12	12	10	5	11	12	13	11	10	11	12	12	9	10	11	12	8	11	10	11	10	216	19.79%
Total	51	48	59	43	57	45	47	41	54	52	45	47	59	54	61	51	61	49	59	52	57	1.092

.

From this table, we can draw an initial conclusion: only 51.44% of operations were carried out correctly within the time allowed. This information is in line with the findings of Dezi et al. [12], whose research carried out at delivery points in the study area of the city of Bologna showed that more than 50% of delivery points were occupied illegally. In this case, the correct use of the LUZ only considers operations that have been carried out within the time allowed, although subsequently, when calculating availability, the first 30 min are counted as correct, with any time after the 31st minute being considered illegal.

For a better understanding, Figure 3 shows the daily distribution of the use of loading and unloading zones during the month analysed, differentiating between the three types of behaviour observed: operations carried out correctly within the permitted time, operations exceeding the maximum time allowed (30 min) and occupancy by unauthorised vehicles.

Combining Table 3 and Table 4, the Figure 4 reflects the proportion of operations carried out in the LUZ, classified into five categories: correct use, excessive time, unauthorised vehicles, double parking, and illegal occupation outside the authorised zone. This visualisation reinforces the urgency of adopting corrective and control measures to improve efficiency and legality in urban distribution.

Another interesting fact, whose impact on the OEE calculation is referenced in the weighted time calculation formula, is the type of vehicle used for unloading. It has been observed that only in very few cases are LUZs used by vehicles exceeding 3.5 tonnes of maximum authorised mass MMA, with double parking normally being used to unload such large vehicles. In the case of unloading at the supermarket or ‘dark store’ in the area, the unloading operation includes lowering the pallets from the truck with a tail lift and then walking to the delivery point by dragging the pallet with an electric pallet truck along the road and pavement, with the consequent risk to the delivery driver, other road users and pedestrians.

Table 6. Total operations recorded by vehicle type.

Vehicle type	Operations	%
Light Truck 7.5 Tn MMA	21	1.33%
Chassis Cab 3.5 Tn MMA	148	9.36%
Large Van	94	5.94%
Small Van	474	29.96%
Delivery Van	524	33.12%
Passenger Car	321	20.29%
Total	1.582

shows the total operations recorded by type of vehicle, including those carried out in the LUZ, double parking, and illegal parking.

The Figure 5 shows the number of operations recorded according to the type of vehicle used. Delivery vans (33.1%) and small vans (30%) account for the majority of operations, highlighting the predominance of light vehicles in urban goods distribution.

In contrast, heavier trucks (3.5 and 7.5 tonnes) account for a very small percentage of operations, suggesting that these vehicles tend to make their deliveries outside the LUZ or by double parking. This information may be useful for redesigning areas according to the type of fleet actually in operation.

3.2. Availability

The results obtained from applying the defined formula can be seen in the following tables, which analyse the factor according to zone, time slot and day of the week.

Table 7. Availability factor by zone.

	Minutes Reserved		Days of Operation		Sum of tp		Tt Available		Availability		Availability
	M	A	M	A	M	A	M	A	M	A	Daily
Zone 1	180	180	21	21	2.246	2.697	3.780	3.780	0.59	0.71	0.65
Zone 2	180		21	21	2.023		3.780		0.54		0.54
Zone 3	300	180	20	20	3.181	2.987	6.000	3.600	0.53	0.83	0.64
Zone 4	180	180	21	21	2.322	2.405	3.780	3.780	0.61	0.64	0.63
Zone 5	180	180	21	21	1.464	1.417	3.780	3.780	0.39	0.37	0.38
Total	1.020	720	104	104	11.237	9.506	21.120	14.940	0.53	0.64	0.58

shows the availability factor by zone.

It can be observed that there is a considerable difference between zones and also between the morning and afternoon periods.

Table 8. Availability by time slot.

Hour	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	T Total	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	Availability
			738			1.260			0.61			0.61
8		662	617			1.260		0.53	0.51			0.52
9	1.198	922	640	1.447	1.070	1.260	0.95	0.73	0.53	1.15	0.85	0.84
10	609	381	775	618	262	1.260	0.48	0.30	0.65	0.49	0.21	0.43
11	447	58	332	257	133	1.260	0.35	0.05	0.28	0.20	0.11	0.20
14	1.803		1.988	1.534	1.101	1.260	1.43		1.66	1.22	0.87	1.29
15	614		712	635	172	1.260	0.49		0.59	0.50	0.14	0.43
16	273		367	235	144	1.260	0.22		0.31	0.19	0.11	0.21

shows availability by time slot.

By analysing behaviour by hour, it can be deduced that the first hours are those with the highest activity, while the last hours are those with the fewest operations. It should be noted that the arrival time is taken as a reference.

Table 9. Availability by day of the week.

	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	T Total	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	Availability
Monday	1.018	385	1.380	920	457	4	1.440	720	1.920	1.440	1.440	0.60
Tuesday	1.341	468	1.317	1.056	642	4	1.440	720	1.920	1.440	1.440	0.69
Wednesday	732	241	613	337	208	3	1.080	540	1.440	1.080	1.080	0.41
Thursday	822	364	1.747	1.686	739	4	1.440	720	1.920	1.440	1.440	0.77
Friday	1.029	566	1.111	727	836	5	1.800	900	2.400	1.800	1.800	0.49

shows availability by day of the week.

In terms of weekly seasonality, Fridays are the days when the areas are used the least. There are external factors that affect activity, such as the fact that Friday 2nd was a bank holiday and activity was noticeably lower. The data obtained for Wednesdays is also noteworthy, but there was one day (7/05) when the area was not used for the LUZ as it was reserved for a removal, during what had been defined as scheduled time when the LUZ was not to be used.

In short, we can take the value of 0.58 as an indicator of the availability of the set of LUZs under study. The reading is that 58% of the time LUZs are occupied.

3.3. Efficiency

At this point, the evaluation proceeds by discounting the factor that penalises efficiency, which is the illegal occupation of the LUZ by private non-commercial vehicles that make use of free parking, preventing the proper use of the LUZ. Let us also see how this indicator behaves with the level of analysis in the previous section. The efficiency by zone can be seen in

Table 10. Efficiency by zone.

	Minutes Reserved		Days of Operation		Sum of tp		Tt Available		No Efficiency		No Efficiency
	M	A	M	A	M	A	M	A	M	A	Dayily
Zone 1	180	180	21	21	481	789	3.780	3.780	0.87	0.79	0.83
Zone 2	180		21	21	445		3.780		0.88		0.88
Zone 3	300	180	20	20	673	939	6.000	3.600	0.89	0.74	0.83
Zone 4	180	180	21	21	166	553	3.780	3.780	0.96	0.85	0.90
Zone 5	180	180	21	21	168	146	3.780	3.780	0.96	0.96	0.96
Total	1.020	720	104	104	1.932	2.427	21.120	14.940	0.91	0.84	0.88

.

In this case, more regular behaviour was observed, both by area and in the morning and afternoon. Only two viable solutions were proposed to try to resolve this reduction in performance. On the one hand, awareness campaigns for the public and, on the other, increased police control. It should be noted that during the 21 days of analysis in which data was collected in the field, only once the police issued a fine in a LUZ, to a private vehicle that had been parked for 10 min, while right next to it there was a large van that had been parked in the area for two days and was not fined, as it is supposed to be an authorised vehicle and the time it spends in the area is not monitored.

Table 11. Efficiency by time slot.

	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	T Total	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	No Efficiency
7			143			1.260			0.12			0.88
8		162	33			1.260		0.13	0.03			0.92
9	242	96	182	112	25	1.260	0.19	0.08	0.15	0.09	0.02	0.89
10	110	167	267	29	137	1.260	0.09	0.13	0.22	0.02	0.11	0.89
11	129	19	47	26	6	1.260	0.10	0.02	0.04	0.02	0.00	0.96
14	275		571	349	54	1.260	0.22	0,00	0.48	0.28	0.04	0.80
15	396		300	157	78	1.260	0.31	0.00	0.25	0.12	0.06	0.85
16	118		68	46	14	1.260	0.09	0.00	0.06	0.04	0.01	0.96

shows the performance by time slot.

No significant variation in this alteration was detected depending on the time of day or the area. The day of non-use in zone 3 was taken into account.

Table 12. Efficiency by day of the week.

	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	T Total	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	No Efficiency
Monday	138	116	440	37	159	4	1.440	720	1.920	1.440	1.440	0.87
Tuesday	262	55	351	292	77	4	1.440	720	1.920	1.440	1.440	0.85
Wednesday	375	33	210	30		3	1.080	540	1.440	1.080	1.080	0.88
Thursday	291	59	302	217	78	4	1.440	720	1.920	1.440	1.440	0.86
Friday	204	181	308	143		5	1.800	900	2.400	1.800	1.800	0.90

shows the performance per day of the week.

The table shows that there are no significant differences between days of the week. The day on which Zone 3 was not used has also been taken into account.

To calculate OEE, we will use the value 1 (no efficiency). This value quantifies the proportion of time that the LUZ is occupied by private vehicles and provides the city government with tools to improve loading and unloading operations without increasing reserved times. In this case, the value for all areas is 0.88, which means that 12% of the time, the LUZ is occupied by illegally parked private vehicles.

3.4. Quality

This indicator identifies the time that LUZ spaces are illegally occupied by commercial vehicles that abuse the areas, in some cases using them directly as free parking spaces. During the analysis period, several vans from nearby businesses were detected repeatedly making this improper use of the spaces. On several occasions, vehicles have even been detected parked for several days in a LUZ. As in the previous sections, these will be analysed separately. It should be specified that times exceeding the 30 min allowed are considered t_oi, with the first 30 min of the operation being reported as normal.

Table 13. Quality by zone.

	Minutes Reserved		Days of Operation		Sum of tp		Tt Available		No Quality		No Quality
	M	A	M	A	M	A	M	A	M	A	Dayily
Zone 1	180	180	21	21	807	1.317	3.780	3.780	0.79	0.65	0.72
Zone 2	180		21	21	573		3.780		0.85		0.85
Zone 3	300	180	20	20	1.277	1.284	6.000	3.600	0.79	0.64	0.73
Zone 4	180	180	21	21	1.359	1.191	3.780	3.780	0.64	0.69	0.66
Zone 5	180	180	21	21	785	821	3.780	3.780	0.79	0.78	0.79
Total	1.020	720	104	104	4.800	4.612	21.12	14.940	0.77	0.69	0.74

shows the quality factor by zone.

The table shows that there is greater abuse of excess time in the afternoon sections, as well as greater abuse in Zone 4.

Table 14. Quality by time slot.

	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	T Total	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	No Quality
7			402			1.260			0.34			0.66
8		108	336			1.260		0.09	0.28			0.82
9	502	381	186	927	736	1.260	0.40	0.30	0.16	0.74	0.58	0.56
10	184	68	189	254	23	1.260	0.15	0.05	0.10	0.20	0.02	0.88
11	120	16	83	178	26	1.260	0.10	0.01	0.07	0,14	0.02	0.93
14	1.118		1.007	822	760	1.260	0.89	0.00	0.84	0.65	0.60	0.40
15	84		179	240	3	1.260	0.07	0.00	0.15	0.19	0.00	0.92
16	115		176	129	58	1.260	0.09	0.00	0.15	0.10	0.05	0.92

shows the quality factor by time section.

It has been observed that there is greater abuse by delivery drivers at the end of the reserved time slot, probably because they take advantage of the reserved time to find a space and park their vehicles. The quality factor per day of the week is shown in

Table 15. Quality by day of the week.

	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	T Total	Zone 1	Zone 2	Zone 3	Zone 4	Zone 5	No Quality
Monday	615	78	535	639	126	4	1.440	720	1.920	1.440	1.440	0.71
Tuesday	707	209	672	399	341	4	1.440	720	1.920	1.440	1.440	0.67
Wednesday	50	32	225	139	126	3	1.080	540	1.440	1.080	1.080	0.89
Thursday	356	108	673	1.002	359	4	1.440	720	1.920	1.440	1.440	0.64
Friday	395	146	456	370	655	5	1.800	900	2.400	1.800	1.800	0.77

.

Wednesdays receive the best response, while Tuesdays and Thursdays receive the worst results.

To calculate OEE, the value of Non-Quality = 1 (weighted time of illegal use due to excess time) will be used. A very low indicator would denote little solidarity among transporters, who do not make an effort to comply with the established times. In this case, the result is 0.74, indicating that 26% of the time, the LUZs are illegally occupied by vans or trucks that exceed the permitted time.

If we exclude operations that exceeded 30 min and focus solely on successful deliveries, there were only 305 (27.9% of those made in the LUZ), and the average delivery time was 17.9 min, which suggests to the city authorities that the 30 min currently established by the ordinance could perhaps be reduced. This figure is similar to that of the city of Bologna, which has set the time for completing the operation at 14 min. Obviously, if the time allowed is reduced, this non-quality indicator would increase considerably.

One possible solution to try to alleviate this excess, apart from public awareness and police control, would be to equip each parking space with a light and a presence sensor that, after 30 min, begins to flash to indicate that the vehicle is illegally parked due to exceeding the time limit.

3.5. OEE Calculation

Finally, with all factors analysed separately, the OEE will be evaluated, also broken down by the types analysed above. The OEE formula is applied as shown in

Table 16. OEE by zone calculation.

	A	E	Q	OEE
Zone 1	0.65	0.83	0.72	0.39
Zone 2	0.54	0.88	0.85	0.40
Zone 3	0.64	0.83	0.73	0.39
Zone 4	0.63	0.90	0.66	0.38
Zone 5	0.38	0.96	0.79	0.29
Total	0.58	0.88	0.74	0.37

.

OEE = A × E × Q

It can be seen that Zone 5 has a significantly lower OEE than the others, and in reference to the overall OEE. In general, a value of 0.37 is very low. The result suggests that there is significant margin for improvement in the area. One interpretation could be that the areas analysed as a whole are only achieving 37% of the targets set by the city’s governance and that improvement measures should be established to address the 63% of non-compliance.

Figure 6 provides an integrated and visual overview of the performance of each of the loading and unloading areas (LUZ) analysed. The three factors that make up the OEE model, Availability, Efficiency and Quality, are shown alongside the total OEE value.

It can be seen that, although some areas have high levels of efficiency or quality of use, others, such as area 5, show very low levels of availability, which penalises their overall performance. The direct comparative between areas allows specific opportunities for improvement to be identified in each dimension, which is key to guiding differentiated management measures by the urban administration.

OEE by time slot, in

Table 17. OEE by time slot.

	A	E	Q	OEE
7	0.61	0.88	0.66	0.36
8	0.52	0.92	0.82	0.39
9	0.84	0.89	0.56	0.43
10	0.43	0.89	0.88	0.33
11	0.20	0.96	0.93	0.18
14	1.29	0.80	0.40	0.42
15	0.43	0.85	0.92	0.34
16	0.21	0,96	0.92	0.18

.

These results suggest that there are time slots that are more favourable for logistics operations and others in which the use of LUZ is less efficient, either due to improper occupancy or excessive time. This information can be key to redesigning usage schedules or implementing specific control measures. It can also be seen that the OEE is much lower in the final slots of the reserved time, which is probably influenced by the fact that the start time is taken as a reference.

Figure 7 shows how the OEE indicator value varies throughout the day in the areas analysed. A peak in efficiency can be seen around 9 h, followed by a gradual decline towards midday. In the afternoon, the OEE indicator rises slightly again around 14 h, before falling again.

OEE per day of the week, in

Table 18. OEE by day of the week.

	A	E	Q	OEE
Monday	0.60	0.87	0.71	0.37
Tuesday	0.69	0.85	0.67	0.39
Wednesday	0.41	0.88	0.89	0.32
Thursday	0.77	0.86	0.64	0.43
Friday	0.49	0.90	0.77	0.34

.

Finally, the OEE per day of the week shows that Wednesdays and Fridays are when the indicator is lowest, which could lead to a reconsideration of the schedules assigned based on the day of the week.

3.6. Possible Improvement of the OEE

During data collection, it was observed that small vans often do not use LUZ zones to unload goods but rather as parking spaces while they carry out their services at homes (plumbing, carpentry, cleaning, etc.). A simulation was carried out in which the use of LUZs was prohibited for small vans and these same activities were assigned to delivery vans exclusively in LUZs. OEE in the area would only increase by 3%. A comparative t-test was therefore carried out between the occupancy times of small vans and large vans, taking into account unloading in the LUZ, double parking and illegal parking, i.e., all unloading recorded during the time slot.

The results of the t-test showed a t-statistic of −3.44, indicating that, on average, the occupancy times of delivery vans are lower than those of small vans. The p-value obtained was 0.00059, which is considerably lower than the significance threshold of 0.05, allowing us to conclude that there is a statistically significant difference between the two groups. These findings suggest that delivery vans tend to have lower occupancy times in loading and unloading areas than small vans, which may have implications for the management and optimisation of urban logistics operations. This data reinforces what was observed in the data capture in the sense that small vans are mainly used for general services (carpentry, plumbing, cleaning, etc.) and not for unloading per se, and they could have restricted access to the LUZ.

4. Discussion

It is no surprise that the illegal occupation of LUZs by private non-commercial vehicles and the excessive use by transport operators are the main causes of poor LUZ performance. One of the contributions of this work is the proposal of a methodology to quantitatively evaluate and enable the comparison of the behaviour of an area or set of areas using the OEE indicator. To do this, the key is the introduction of the new weighted occupancy time KPI, in which each vehicle contributes its length over the period of time it occupies the area.

Evaluating separately and then jointly the availability of each zone, its occupation efficiency, and the quality of deliveries in terms of completion within the allowed time provides very valuable information for decision-making. Similarly, being able to evaluate each zone separately, as well as the performance of the zones based on booking times and days of the week, also provides relevant information for assessing the performance of the zone. Finally, temporal versatility is also an important factor. Each zone can be evaluated during a morning or afternoon slot, a full day, a week, a month, or any other relevant time frame.

Focusing on illegal occupation, it should be noted that in Zone 1, only 7.7% of the area is dedicated to LUZ (13 m of the 168 m dedicated to free street parking); in Zone 2, this figure is 5.7% (10 m out of 173 m of regulated parking on the avenue); and in Zones 3, 4, and 5 combined, it is 21.8% (18 + 10 + 8 m out of the 165 m of street space). It is a significant finding that of the 472 instances of illegal parking observed in Zone 1 (a one-way street with the possibility of double parking), 38.77% (183) involved double parking, while only 3.81% (18) involved illegal parking in front of garage doors, on pavements, or on pedestrian crossings In contrast, in the streets correspondi ng to Zones 3, 4, and 5, where double parking is not possible, illegal parking accounted for 17.22% (140 out of 813). The street with the three zones and alarger total LUZ allocation still had three times more illegal parking than Zone 1, indicating that the impossibility of double parking leads to a significant increase in illegal parking in unauthorised areas, regardless of the size of the space reserved for LUZ on the street.

The field study carried out shows that the use of free parking by private vehicles and inconsiderate transport operators leads to poor management of LUZ and generates externalities such as double parking and illegal occupation of pedestrian crossings, garage doors, and intersection islands. However, this study has detected deliveries by delivery drivers, especially from the HORECA sector, who double park even when there are spaces available in the LUZ in order to minimise walking distances and due to the heavy weight of the goods they have to deliver.

The proposed KPIs are applicable in urban environments in any area of the city and allow for the evaluation of the suitability of the location and size of the area. They can be used for a single area or to compare a set of areas. These parameters and the proposed methodology respond to the limitations that other authors have expressed in the literature discussed throughout the article. This method of calculating the OEE indicator is valid for any city, for an area or for a set of areas, and is therefore considered a very useful tool for city governance. As a general rule, both the location and size of the space reserved for the LUZ, as well as the time allocated for loading and unloading, are generally applied in cities in a generalised manner and without objective and quantifiable criteria for measuring them. This proposed methodology allows city authorities to make informed decisions regarding urban mobility regulations and could be particularly valuable for improving the management of low-emission zones to ensure better utilization.

The main limitation of this study lies in data collection, which was manual, tedious, and labour-intensive. This process could be standardised and automated with artificial vision cameras capable of identifying vehicle types based on dimensions or by reading number plates in connection with the Direction General of Traffic (DGT). In this case, OEE could be calculated automatically and evaluated in real time by the city’s governance system.

Another important limitation is the geographical scope, as the study is based on a single city and a single area of that city. While the operational dynamics are likely the same across different urban areas, other factors, such as pedestrian areas or low-emission zones governed by different legislation, may affect the generalizability of the findings.

5. Conclusions

One of the main contributions of this article is the proposal and characterisation of a new KPI, weighted occupancy time, which serves as the basis for calculating the values of the availability, performance, and quality factors that are used to evaluate OEE. This is a very important contribution, as it quantifies the proportion of time that the length of the transport element occupies the area relative to the total occupation time, forming the basis for calculating OEE. Only a 51.44% of the deliveries in the LUZ were correct, and the total OEE ratio was 0.37, which is a result that could be greatly improved.

The literature review identified the need for measurements of these occupancy levels, but no study indicated how to take these measurements. The studies analysed also pointed to the need to standardise these measurements, as well as segment occupancy ratesaverage parking durations, and indirect parking problems. All these needs can be addressed through OEE analysis based on the proposed weighted occupation time indicator.

Proper standardisation of data collection and analysis of the data obtained using the proposed methodology would provide city authorities with a decision-making tool to improve each of the OEE factors and thus achieve better use of street lighting, with the consequent reduction in negative externalities such as double parking and illegal occupation of spaces. Decisions could also be made regarding changes to the location, size, or reservation times of LUZs.

At the beginning of this study, the illegal occupation of LUZs by private vehicles and the excessive time spent by authorised commercial vehicles were identified as the main causes of LUZ underutilisation. A key contribution of this article is that it provides KPIs that enable these issues to be quantified. Specifically, analysing each of the factors used to calculate OEE, the low figure of 0.58 suggests that the current location of the zones may not be optimal. A gravitational analysis of consumption points in each zone could help identify better locations and reduce walking distances. If awareness-raising and, above all, coercive measures were taken to eliminate the illegal occupation by private vehicles, the overall OEE could rise to 0.43, i.e., an increase of 14.85%. And if this were achieved through sensorisation, time control discs, artificial vision cameras, or other coercive methods, eliminating excess time, the OEE could potentially rise to 0.51, representing an increase of 36.57%.

The OEE model makes it possible to identifydifferences in efficiency in the use of loading and unloading areas according to vehicle type, thanks to the introduction of a new KPI on weighted occupancy time. The results provide useful quantitative criteria for urban planning and the redesign of logistics spaces to better meet the growing demand for urban logistics. Increasing the OEE indicator for an area or a group of areas has a direct impact on reducing double parking and the illegal occupation of public space. City authorities could take action in areas with lower OEE scores and try to apply specific solutions related to each OEE calculation factor, such as those mentioned above.

In the future, research could be conducted on the impact of illegal parking, such as double parking or the illegal occupation of spaces reserved for public use, not only during the time reserved for loading and unloading but throughout the entire day, in order to assess the effect of expanding or reducing both the space reserved for loading and unloading and the time reserved for this purpose. Supporting what other authors have said, this study shows that flexibility is needed to make better use of urban space and time allocated.

The development of a real-time advance booking app would be an interesting proposal that could contribute to better use of spaces and reduce time wasted by users, as they would be able to move directly to an available space. The data generated by this application would be extremely useful for predicting LUZ usage patterns. Ideally, these ideas could be implemented in practice; however, this is not easily achievable due to the need for control systems for reserved spaces, especially in areas that are not currently subject to regulated parking.

Another possible future research direction could involve exploring ways to optimise LUZ efficiency without extending the weighted occupancy time in an area, thereby avoiding additionalinconvenience to citizens. This could include reassessing the length of the areas and the times reserved for loading and unloading. Additionally, it might be feasible to designatedifferent sections within a single zone, each withdifferent operating times. For example, a longer area could be reserved with a reduced overall time allocation, and a portion of that area could be dedicated to shorter operations, such as ten-minute stops. However, such measures would only be truly effective if there is strict enforcement of compliance with the designatedspaces and times reserved.