Fuzzy Logic-Based Expert Evaluation of Tram Driver’s Console Fidelity in a Universal Simulator

Abstract

Simulators are an effective tool for improving tram driver training. In urban rail transportation, the fidelity of reproducing the driver’s working environment is crucial due to the high diversity of vehicle models. This study presents a structured assessment model for evaluating the mapping of a tram driver’s console in a universal simulator. The model is based on expert judgment and utilizes fuzzy logic to evaluate four key criteria: perspective, button placement, functionality, and time required to locate safety buttons. A group of 30 experts, including experienced tram drivers and technical specialists, assessed the fidelity of the simulated consoles for three tram types: Solaris Tramino S105p, Moderus Gamma LF 06 AC, and Škoda 16T RK. The results enable the classification of console fidelity levels (low, moderate, high) and support the identification of design inconsistencies. The proposed model provides a standardized tool for assessing simulator realism, which can be applied by transport operators, manufacturers, and training centers to improve simulator configurations. Researchers may also use the model as a methodological framework for further evaluation studies involving human–machine interface fidelity.

1. Introduction

Annual accident reports indicate that the average annual number of road accidents in Poland between 2014 and 2023 caused by a tram or trolleybus was 46 (Figure 1). In 2014–2023, 19 people were killed and 797 were injured in road accidents involving a tram or trolleybus [1].

Simulators are a tool to support the training process and are used in the air [2,3,4], marine [5,6], road [7,8] and rail sectors [9,10]. Simulators are also used in urban rail transportation. In rail vehicle simulators, accurate reproduction of the control panel is essential for realistic training, which can help reduce accidents.

In an era of developing advanced simulation technologies, the accuracy of the simulators’ reproduction of the real working environment is crucial to the effectiveness and operational safety of training. The tram driver’s console is the primary tool for controlling the vehicle, determining the safety and efficiency of the operations performed. Between 2000 and 2021, there were 7535 collisions involving trams and pedestrians in Sweden, Germany, Switzerland, and Australia [11]. Incidents involving trams pose a direct threat to the lives of road users. Therefore, the training process for tram drivers is becoming increasingly important. A problem in the training process is that each vehicle is custom-designed for a particular city. For example, MPK Wroclaw has 10 different types of vehicles [12], Silesian Trams has 11 types [13], and Warsaw Trams has as many as 14 types [14]. During their work, tram drivers often change vehicle types. Currently, there are six tram simulators in Poland:

• NGT6 tram simulator, developed by the Institute of Rail Vehicles at the Faculty of Mechanical Engineering of Cracow University of Technology (the first Polish tram simulator), and the GT8S tram at the Museum of Urban Engineering;
• Tram simulator 105 Na of the company Polskie Simulatory in the Museum of Technology and Communication in Szczecin;
• 805N-ML tram simulator and Pesa Swing tram simulator by Polskie Simulatory in Lodz;
• Konstal 105N2k/2000 tram simulator by Lander in Warsaw;
• Moderus Beta MF 24 AC tram simulator from Lander in Wroclaw;
• Universal tram simulator at the Wroclaw University of Technology in Wroclaw, Poland.

The diversification of current models on the market underscores the need to develop universal training tools.

This paper aims to develop and validate a structured model for assessing the mapping fidelity of a tram driver’s console in a simulator by comparing it with the actual console of selected vehicles. The proposed model addresses a practical and currently unresolved problem: the lack of standardized procedures for evaluating the realism of tram simulator control panels. The model is based on expert evaluation and fuzzy logic, allowing a repeatable and transparent assessment applicable to different tram types. In total, 30 experts evaluated the consoles based on four critical criteria: perspective, button placement, functionality, and the time required to locate safety buttons. The study involved the selection of specific tram models, for which the corresponding driver’s consoles were recreated in the simulator. A two-stage verification was conducted. First, participants evaluated the real tram consoles using a structured questionnaire. Then, they assessed their simulated equivalents. This procedure enabled an evaluation of the degree of fidelity in console mapping. The model is intended to support simulator developers, operators, and training providers by identifying inconsistencies in control panel replication and improving interface realism. The approach is subject to certain limitations related to the expert-based nature of the evaluation. These primarily concern the context-specific character of the results and the potential variability in individual expert opinions. However, such risks were mitigated by selecting a balanced and experienced group of 30 domain experts, applying a uniform assessment procedure, and analyzing the consistency of responses using statistical indicators. As a result, the model provides a reliable and consistent structure that can be adapted to other urban rail simulation contexts. The identified research gap, confirmed through the literature review, concerns the lack of formalized methods or reference models for assessing the fidelity of tram simulator driver consoles. Existing studies typically focus on driving performance or ergonomic design, but they do not address how accurately the real-world control interface is replicated in a simulator. This gap highlights the need for a systematic and repeatable evaluation approach, which the proposed model aims to fulfill. Therefore, the contribution of this study is relevant for both practice and academic research.

The paper is structured as follows: Section 2 reviews the literature and available solutions, identifying the research gap. Section 3 is a description of the research process presented in the paper. Section 4 presents a detailed description of the assessment criteria on which the model described in Section 5 is based. Section 6 is a description of the case study. Section 7 presents conclusions from the work carried out within the article.

2. State of the Art

2.1. Literature Review

The use of simulators in training processes is associated with numerous challenges, primarily focused on maintaining a high-fidelity version of the real environment in the simulator [15,16]. Appropriately developed training scenarios are also important [17], influencing behavioral and psychological aspects. The literature includes works that focus on the training of tram drivers. Areas considered include risk assessment of training scenarios [18], design of visual attention training [19] and the use of simulation [20]. Analyzing the Scopus and Web of Science knowledge libraries, a few papers that deal with the tram driver’s cab and its components were identified.

Nathanael and Marmaras [21] analyzed a tram driver’s workstation to develop a concept for redesigning the workstation. The workstation analysis included the following elements: console surfaces with instruments and push-button switches, seat, traction and brake controller, and armrest. An analysis of the tram rolling stock of seven tram networks in France by Guesset et al. [22] found the following problems: accessibility to or actuation of pedals, fogging of side windows, and bouncing of the control console on the windshield. On the other hand, Naweed and Moody [23] identified that changing the vehicle generates a problem involving changes in button locations, which confuses the vehicle controller. Tokic et al. [24] stressed the need to consider anthropometric measures when designing a control panel in a tram cab. Sumpor et al. [25] pointed out the need to place frequently used buttons on the control panel within arm’s reach of the driver. Kierzkowski et al. [26] presented a concept of a universal tram driver’s console with interchangeable panels for a Polish tram simulator developed based on a review of available solutions offered by Polish transportation companies.

Callari et al. [27] used a virtual reality environment for research on the operation of a tram’s master controller. Using a virtual reality tram simulator (TramVR), the speed at which drivers accelerated and braked the tram was verified. The study verified that using tactile and visual feedback during the operation of the main controller helps drivers provide a smoother ride.

A study of thirty-eight drivers with varying levels of driving experience by Nabatilan [28] found that focus varies depending on experience level. Novice drivers focus more on the vehicle’s console than the view from the front of the vehicle. An inverse relationship was noted for experienced drivers. Similar results were obtained for driving in urban rail transportation. Warchoł-Jakubowska et al. [29] subjected the attention focus of experienced drivers and novices to verification in terms of the windshield and the driver’s control panel. The research was carried out by using simulated excerpts from videos depicting tram rides. Experienced drivers focused more on monitoring the driving environment than on the control panel. Tram cab designs should, therefore, consider the location of the various buttons. Beginners tend to focus more on the panel where the speedometer is located [30].

Only Hankiewicz and Lasota [31], based on interviews with tram drivers, focused on the description of the driver’s panel in trams operated in Poznan, considering the evaluation of the ergonomics of the workstation. The authors of the work pointed out differences in the design of the control panels of Düwag, Konstal, Moderus, Simens Combino, and Solaris tram models. Accordingly, the literature review did not identify any method to evaluate the representation of the driver’s console in the simulator.

2.2. Research Gap

Despite the increasing use of simulator-based training systems, no standardized methodology exists for assessing the fidelity of a driver’s console. The literature review highlights various aspects related to driver workstations, including ergonomic design [21], visibility and accessibility issues [22], and the impact of control panel layouts on driver performance [23]. However, these studies focus primarily on design considerations rather than on a structured assessment method that quantitatively evaluates the accuracy of the console’s mapping in a simulator. Some research explores anthropometric factors in control panel design [24,25] and button placement optimization for ease of use [26]. Other studies investigate simulator-based training effectiveness [27,28], including how haptic and visual feedback improve driver response [29]. However, there is no systematic framework for assessing whether a simulator realistically replicates the layout, functionality, and usability of the console.

Despite the growing reliance on simulators for standardized and scalable driver training, there is no validated assessment framework to ensure their effectiveness. This study addresses this gap by developing a structured methodology for evaluating the fidelity of a simulator’s driver’s console through expert assessments, statistical validation, and fuzzy logic-based inference. This approach provides a repeatable and objective basis for identifying discrepancies, improving console mapping accuracy, and enhancing the realism of training environments.

2.3. Results Reliability

Ensuring the reliability of expert-based evaluations is essential for the credibility of the proposed model. To ensure the reliability of the results, a unified evaluation procedure was applied to all experts, and the same set of criteria and instructions was used throughout the experiment. The group of 30 experts involved in the study represents a relatively large sample for this specific domain, as access to qualified tram drivers and simulator specialists with sufficient experience is inherently limited. Their assessments were independent and carried out under controlled conditions, which minimized the risk of bias or mutual influence. The analysis of mean values and 95% confidence intervals allowed for assessing the central tendency and precision of the expert responses. Additionally, coefficients of variation were calculated to evaluate the internal consistency and variability of the assessments across different tram types and criteria. These statistical measures confirm that the results are not random or overly sensitive to individual differences, supporting the robustness and reproducibility of the findings.

2.4. Overview of Tram Consoles

2.4.1. Solaris Tramino S105p

Solaris Tramino S105p is a fully low-floor, articulated tram with five sections. The vehicle’s total length is 32.026 m, while the body’s width is 2.4 m. The console on the S105p tram consists of three sections—right, center, and left. In addition, three buttons and one switch are located on the right armrest. The tram is also equipped with a side panel. On the main panel in the central section, there is one operator panel with a 10″ diagonal screen size showing the current status of the vehicle’s components. The right panel contains switches related to the operation of wipers, exterior lighting, door control, driving direction, and safety. The left panel contains only the passenger information system. On the side console, some buttons are responsible for vehicle startup, heating, interior lighting, mirrors, and roller shutters. The console is not equipped with bottom panels. The cab controller is on the left side and is not equipped with an external bell. It allows the vehicle to start and brake, and has standby options. The driver’s seat has a functional armrest with the following buttons: turn signals, emergency braking, sandbox, and external bell.

2.4.2. Moderus Gamma LF 06 AC

The Moderus Gamma LF 06 AC is a 32,500 mm long five-member vehicle based on four bogies. The tram is 88% low-floor, with the high floor located only within the first and last bogie, and all passenger doors are accessible from the low-floor level. The driver’s panel on the LF 06 AC tram consists of three sections—right, center, and left, where the right and left panels have additional lower sections. The tram also has a side panel with buttons to start the tram. There is one operator panel with a diagonal screen size of 10″ on the main console in the central section. The central terminal displays only necessary information related to the operation of individual systems and general errors of individual systems. The right panel contains the passenger information system and switches related to wipers, exterior lighting, door control, driving direction, and safety. The left panel contains switches related to heating, interior lighting, and mirrors. The console is also equipped with bottom panels (bottom panel), switches related to door control, safety, windshield washers, and blinds. The cab controller is on the left side and has an external bell. It allows the vehicle to start and brake, and has standby options. The driver’s seat has a functional armrest with the following buttons: external bell, door control, turn signals, and high beam.

2.4.3. Skoda 16T RK

The Skoda 16T RK is a 65% low-floor, five-member single-circuit streetcar with two end members, two inset members, and a center member. The end members are belted on two-axle traction bogies, each driven by two asynchronous traction motors that always form a single motor unit. The center member fits on a two-axle rolling carriage. The driver’s panel on the 16T tram consists of three sections—right, center, and left. The tram also has a side panel, and the seat does not have a functional armrest. The main panel in the central section contains a touch-screen operator panel showing the current status of the vehicle’s components and buttons responsible for exterior lighting. The right panel contains the passenger information system, buttons responsible for wipers, doors, driving direction, and safety-related controls. The left panel contains switches for interior lighting, mirrors, and roller shutters, while the side console contains those related to vehicle startup and heating. The console is not equipped with bottom panels. The cab controller is on the left side and has an external bell. It allows the vehicle to start and brake, and has a standby option.

2.4.4. Universal Tram Driver Control Panel

The universal tram driver control panel was used to develop the assessment model.

Table 1. Arrangement of button groups in the universal tram driver control panel.

Group of Buttons	Actual Solaris Tramino S105p	Simulated Solaris Tramino S105p	Actual Moderus Gamma LF 06 AC	Simulated Moderus Gamma LF 06 AC	Actual Skoda 16T RK	Simulated Skoda 16T RK	Configuration Options of the Simulator
vehicle startup	SC	SC	SC	SC	SC	SC	SC
heating	SC	LP	LP	LP	SC	LP	RP or LP
interior lighting	SC	LP	LP	LP	SC and LP	LP	RP or LP
mirrors and roller shutter	SC	LP and BL	LP and BL	LP and BL	LP	LP and BL	RP and BR or LP and BL
wipers	RP	BL	RP and BL	BL	RP	BR	BR or BL
exterior lighting	RP	BR	RP and LP	BR	CP	BL	BR or BL
door control	RP	BR and A	BR and A	BR and A	RP	BR and A	BR or BL and A
driving direction	RP	RP and A	RP and A	RP and A	RP and CP	RP and A	RP or LP and A
safety	RP and A	RP and LP	RP and, BR and A	RP and LP	RP	RP and LP	RP and LP
cab controller	CC	CC	CC	CC	CC	CC	CC
passenger information computer	LP	RP	RP	RP	RP	RP	RP or LP

Green color denotes the compatibility of a given button group with the actual vehicle, orange color denotes partial compatibility, and red color denotes complete incompatibility.

summarizes the distribution of button groups used in the universal panel for each type of vehicle. Table 1 presents a summary prepared on the basis of the authors’ analysis of existing solutions available on the Polish market. Table 1 uses symbols to denote the areas of the simulated dashboard: SC—side panel, RP—right panel, LP—left panel, A—armrest, CC—cab controller, CP—central panel, BR—bottom right panel, and BL—bottom left panel.

The last column of Table 1 summarizes the reconfiguration capabilities of the developed simulator. Individual groups of buttons can be swapped with each other between slots of the same size according to the concept shown in Figure 2, with Figure 3 presenting a real photograph of the control panel. This allows several tram models to be mirrored using real buttons rather than controllers displayed on touch screens. This method greatly increases the realism of the training experience. The group associated with the vehicle startup can only be on the side panel. Heating, interior lighting, and passenger information computer are on the right or left panel, as well as mirrors and roller shutters on the left or right panel, including the bottom panels, due to the separation of this group between the two slots. Wiper and exterior lighting are on the right or left bottom panels. Door control and driving direction are also on the lower left or right panels. Driving directions are duplicated additionally in the armrest of the driver’s seat. Safety-related buttons are on the right and left panels. The cab controller is on the left side and has an external bell.

The tram console area is divided into eight elements:

• Side panel—equipped with switches used infrequently, such as those required for vehicle startup or to take action in case of tram damage;
• Central panel—equipped with the onboard computer;
• Armrest—often equipped with turn signals or buttons to open/close the doors, the armrest is the element in which the buttons most often used by the driver are located;
• Cab controller—equipped with buttons for starting, unloading, and braking, including emergency braking; in many cases, it is equipped with an external bell switch;
• Left panel—equipped with buttons such as current collector, battery, riding on battery power, driving direction, air conditioning and compartment heating, internal lighting, disabled person, fan speed of heater, heating mirrors, STOP (emergency braking), heating of the cabin, adjustment of mirrors;
• Bottom left panel—equipped with buttons such as dimming of the controls, wiper operating mode, sprinkler, wiper impulse, emergency lights, fog lamps, cab lighting, roller shutter;
• Right panel—equipped with elements such as turn signals, bell, crossover control, passenger information computer;
• Bottom right panel—equipped with buttons such as positioning lights, passing lights, passing lights, activation of the passenger buttons, door opening, door closing, and front door.

The first of the simulated consoles was the Solaris Tramino S105p. Its console shell is not equipped with two lower panels, and many buttons are located on the side console, which entails specific compromises in button placement, as described further below. Buttons related to vehicle startup were located on the side console, consistent with their placement in the actual vehicle. Heating and interior lighting were placed on the left panel, which is inconsistent with the actual vehicle but is closest to the side console. Mirrors and roller shutters were located on the left panel and the left bottom panel, and are actually on the side console. Wipers were located on the left side due to the lack of space on the right side, caused by the placement of other partially compliant button groups—this does not coincide with the actual tram. Exterior lighting is located on the right panel, and in the simulator, it has also been placed on the right part, but on the side panel. The same applies to door control, which is duplicated in the armrest simulator. Buttons related to driving direction are on the same panel in both cabins, but the simulator duplicates them in the armrest. Safety-related buttons are partially consistent within the right panel. The cab controller is in the same place, and the passenger information computer has been placed on the opposite side from the actual tram due to the left console’s space being occupied by other conforming button groups.

The second simulated tram was the Moderus Gamma LF 06 AC. The console shell in this vehicle also has two lower consoles, just like the simulator. This is currently a solution used by most manufacturers of modern trams. In both cases, the vehicle’s starting buttons are located on the side console. The heating and interior lighting buttons are placed in the same place as in the real vehicle, i.e., on the left panel. The mirrors and roller shutter control are also in a similar location—on the left panel and the lower left console. The wipers buttons in the real tram are divided between the right panel and the lower left panel, while in the simulator, they are located only on the left panel, which is partially consistent. Exterior lighting is located in the real tram on the right and left panels and in the simulator on the lower right panel, which is partially consistent. The door control and driving direction button groups are consistent; the first is placed on the lower right panel and armrest, and the second is on the right panel and armrest. The safety buttons are additionally duplicated in the real vehicle in the armrest, while in the simulator, they are not duplicated in this place, which is partially consistent. The cab controller is in the same place, and the passenger information computer is on the right side.

The third simulated tram was the Skoda 16T RK. This tram does not have a functional armrest in the driver’s cabin and is not equipped with two lower panels on the console. Therefore, the lower panels in the simulator were used to reflect similar button positions on the right and left consoles in reality as a partial match. The buttons related to vehicle startup are located in the same place—on the side console. The heating control in the real tram is located on the side console, and in the simulator, it is closest to this location, i.e., on the left panel. The interior lighting in the tram is divided into the side console and the left panel, and in the simulator, it is only on the left panel—there is a partial match. The mirrors and roller shutters in the real tram are on the left panel, and in the simulator, they are also on the left side, but they have been divided between the left panel and the lower left panel. Wipers in the tram are located on the right panel, and in the simulator, they are also on the right side, but on the bottom panel due to slot configuration limitations. Exterior lighting is located on the central panel—the simulator does not allow such a configuration, so it is placed on the lower left panel. The door control is located on the lower right panel and armrest; in reality, it is on the right panel. The choice of travel direction is located in the tram on the right panel and central panel, and in the simulator on the right panel and armrest. Safety-related buttons are partially consistent within the right panel. The cab controller is located in the same place, and the passenger information computer is on the right side.

3. Methodology for Assessing the Mapping of the Tram Driver’s Console in a Tram Simulator

The proposed methodology assessment of the mapping of the driver’s console in a tram simulator consists of three main phases: data collection and analysis, the model, and the assessment.

The process starts with the selection of an expert for the study, categorized into two groups: experienced tram drivers and technical specialists. Including these two perspectives ensures a comprehensive evaluation that accounts for operational usability and technical accuracy. To ensure the credibility of the assessments, the selection of experts should be diversified in terms of professional experience and practical and technical perspectives in the presented case of human–machine interface validation. After selecting the experts, the next step is the development of the expert survey. This stage focuses on creating a structured evaluation tool to assess the fidelity of the driver’s console representation systematically. The survey is structured around four predefined assessment criteria derived from the literature review. The first criterion, perspective, considers the indicators’ visibility and the driving position’s ergonomics. The second criterion, button arrangement, evaluates the intuitiveness and consistency of control elements compared with real tram consoles. The third criterion, functionality, measures the correctness of the operation of control elements in the simulator. The fourth criterion, time required to locate safety buttons, assesses the response time in emergency scenarios. All assessment criteria are described in detail in Section 4—Assessment criteria for the tram driver’s panel.

Once the survey is prepared, the next step involves collecting input data from experts. They conduct practical evaluations of the simulator, assessing its fidelity based on the predefined criteria. Each expert completes the survey by providing subjective ratings for all assessment aspects. In this manuscript, a motion platform was used to evaluate the simulator, allowing for the simulation of tram movement dynamics. The simulator was equipped with five monitors. Additionally, the simulator included a Universal Tram Driver Control Panel, described in Section 2.4.4. Each participant was given approximately 25 min to evaluate the simulator configuration for the analyzed tram model.

The second stage is the model. The evaluation model building process includes three main stages: the fuzzification block, the inference block, and the defuzzification block. In the fuzzification block, membership functions are developed for the variables considered in the model, allowing the degree of membership in the fuzzy set to be determined. The inference block consists of developing fuzzy inference rules to assess the fidelity of the simulator. The inference rules were also developed in cooperation with a group of experts selected in the first stage. The last block, the defuzzification block, focuses on aggregating the results from the four evaluation criteria into one final value. The model was developed in the Matlab environment. The detailed description of the assessment model is provided in Section 5—Assessment model for the representation of the tram driver’s console in the simulator.

The first step of the last stage is presented in Figure 4, and the assessment phase. This stage includes two main activities: the assessment of the representation of the driver’s console in the simulator and the interpretation of the results. Based on the results obtained in the defuzzification phase, fidelity is classified into three categories. If the indicator is in the high fidelity range, the simulator closely corresponds to the real driver’s console and is considered well-adapted for training. If the indicator indicates moderate fidelity, the simulator contains some discrepancies. If the indicator is in the low fidelity range, the simulator’s representation differs significantly from reality, which requires modification to increase the effectiveness of the training process. Interpretation of the evaluation results allows obtaining information on the assessment of the fidelity of the control console representation in the simulator and identifying areas requiring improvement.

The structured nature of the proposed methodology ensures that the assessment process can be applied to different tram models as well as other vehicle types through expert assessment while maintaining a standardized approach for determining simulator fidelity. It is stated based on the literature review that such an approach does not currently exist, yet it is highly significant due to the ongoing development of simulator-based training systems. Moreover, such a methodology is particularly important due to the diversity of tram models used in transport companies and the high cost of full-scale simulators. As a result, enterprises will increasingly rely on universal simulators rather than those dedicated to a specific tram model, making it essential to have a standardized approach for assessing their fidelity.

4. Assessment Criteria for the Tram Driver’s Panel

Trams are tailor-made vehicles, which means that each Ordering Party defines all vehicle parameters in the description of the subject of the order, including interior layout, number of units, bogies, equipment, etc. A similar situation is related to the arrangement of buttons and switches on the driver’s panel. In most cases, the final arrangement of individual consoles, sections, and buttons is carried out as part of a dialogue between the Ordering Party and the Contractor.

4.1. Perspective

The first criterion for assessing the control panel is the perspective, assessed on a scale from 1 to 5. It considers four critical visual aspects: floor-to-ceiling view range, view range from left to right side of the cabin, blind spot area, and the possibility of adjusting the seat height. Assessment of the perspective based on these parameters allows for determining how realistically the simulator reproduces the actual view from the driver’s cabin. The characteristics of the perspective criterion elements are presented in

Table 2. Assessment criterion—perspective.

Scale	Description
floor-to-ceiling view range
1	Very limited range of view. Visibility is very limited, and the driver has difficulty seeing the cabin’s floor and ceiling. Many areas are obscured, making it difficult to monitor the interior space fully.
2	Limited range of view. Floor-to-ceiling visibility is limited. The driver has difficulty seeing certain areas of the cabin, which can lead to problems monitoring the entire space and requires extra caution.
3	Moderate range of view. The visibility from floor to ceiling is moderate, allowing the driver to see most cabin areas. Although the view is relatively good, minor restrictions may require additional attention.
4	Wide range of view. The floor-to-ceiling visibility is broad and covers most of the cabin area well. The driver has a good view of the area with minimal restrictions.
5	Full range of view, faithfully reproduced as in the actual cabin. The driver has complete, unobstructed visibility from the cabin floor to the ceiling.
view range from the left to the right side of the cabin
1	Very limited range of view. The driver has minimal visibility from both sides of the cabin. Much of the surroundings are not visible, which makes it difficult to assess the situation on the road and increases the risk of danger.
2	Limited range of view. Visibility is limited, making it difficult for the driver to monitor the surroundings fully. There are visible blind spots, which can delay identifying obstacles and require additional attention.
3	Moderate range of view. The driver has relatively good visibility from both sides of the cab, but some restrictions may exist. The view allows monitoring of most traffic situations.
4	Wide range of view. Visibility is excellent, covering a large part of the surroundings on both sides of the cabin. However, there may be minor restrictions.
5	Full range of view, faithfully reproduced as in the real cabin. The driver has perfect visibility from both sides of the cabin, covering all areas without blind spots. The view is optimal and provides complete control and immediate reaction to situations on the road.
blind spot area
1	Large blind spots. The driver has significant difficulty seeing objects and vehicles in certain areas of the visible area. The blind spots are extensive and can cover significant areas in front of the tram, leading to the invisibility of other vehicles or pedestrians in certain areas.
2	Significant blind spots. The driver has difficulty seeing in certain areas around the tram cabin, although these areas are smaller than in the case of large blind spots. There may be frequent situations where the tram driver fails to see other road users.
3	Moderate blind spots. The driver has limitations in visibility, but they are relatively minor compared to previous levels. They do not significantly affect the driver’s ability to monitor their surroundings. However, awareness of these areas is required.
4	Small blind spots. The driver has a relatively small amount of obstructed visibility. These minimal restrictions do not significantly affect the driver’s ability to assess the traffic situation.
5	Minimal or no blind spots. The driver has complete control over what is happening around the vehicle, and the appropriate placement of mirrors and visibility technologies minimizes any potential blind spots.
seat height
1	Very low or very high. The seat height is far from optimal. The driver’s position is not optimal, which makes visibility and access to the console difficult.
2	Low or high. The seat height is not optimal, causing significant differences in the vision perspective. The driver sits too low or too high, which affects visibility and work ergonomics.
3	Moderate. The seat height is relatively in line with the optimum, with some differences. The viewing perspective is optimal, but a slight deviation may affect the comfort and efficiency of the driver’s work.
4	Seat height is close to optimal. Seat height is close to optimal, with minor differences. The viewing perspective is almost identical to the optimal one, which provides almost ergonomic working conditions.
5	Seat height is optimal. The seat height is perfectly matched and in line with the optimal position of the driver. The viewing perspective is fully ergonomic, ensuring the highest comfort and work efficiency.

. Table 2 was developed by the authors based on empirical analysis of key elements of the evaluation of the perspective.

4.2. Button Arrangement

The second criterion is the arrangement of buttons, assessed on a scale from 1 to 5. The assessment takes into account the convenience and intuitiveness of the arrangement of buttons, which are divided into the following groups: vehicle startup, heating, interior lighting, mirrors and roller shutter, wipers, exterior lighting, door control, driving direction, safety, cab controller, and passenger information computer. A detailed list of buttons in the given groups is provided in

Table 3. Characteristics of button groups.

Button Group	Button Function
vehicle startup	current collector
	battery
	riding on battery power
	driving direction
heating	air conditioning and compartment heating
	the fan speed of the heater
	cabin heating
interior lighting	interior lighting
	dimming the lights
	cabin lighting
mirrors and roller shutter	heating mirrors
	mirror adjustment
	roller shutter
wipers	wipers work mode
	sprinkler
	impulse wipers
exterior lighting	emergency lights
	fog lights
	positioning lights
	low beam
	light impulse
door control	disabled
	activation of passenger buttons
	opening the door
	closing the door
	front door
driving direction	turn signals
	turnout control
safety	STOP (emergency braking)
	bell
cab controller	cab controller
passenger information computer	passenger information computer

. Table 3 was developed by the authors based on empirical analysis of button groups.

Each group of buttons was assessed on a scale from 1 to 5. This assessment allows for a precise determination of how the arrangement of buttons in the tram cabin meets ergonomic requirements and ensures intuitive operation, which is crucial for the comfort and efficiency of the tram driver’s work. The characteristics of the elements of the button arrangement criterion are presented in

Table 4. Assessment criterion—button arrangement.

Scale	Description
1	Very uncomfortable and unintuitive arrangement of buttons. The buttons lack a logical layout, are hard to find, and reaching for them is awkward, making them difficult to use while working.
2	Inconvenient and unintuitive arrangement of buttons. The button layout is chaotic, with significant accessibility and ergonomics issues, negatively affecting the comfort of use.
3	Moderately comfortable and intuitive arrangement of buttons. The buttons can be assigned some logic in their layout, although there may be minor difficulties in finding or using them.
4	Convenient and intuitive arrangement of buttons. The button layout is well thought out and logical, with minor ergonomic issues, allowing for comfortable and efficient use.
5	Very convenient and intuitive arrangement of buttons. The buttons have been assigned an optimal and logical layout, allowing easy, quick, and convenient use without any difficulties.

. Table 4 was developed by the authors based on empirical analysis of key elements of the evaluation of the button arrangement.

4.3. Functionality

The third criterion is the functionality of the button groups, assessed on a scale from 1 to 5, where each assessment reflects different levels of effectiveness, ease of use, and intuitiveness of the buttons. In the case of an assessment at level 1, the buttons often fail, are difficult to use, and have unintuitive functions. In the case of an assessment of 2, which means low functionality, the buttons work insufficiently and are also difficult to use, which leads to delays and complications in operation. Moderate functionality, assessed at level 3, means that the buttons work reasonably well and are relatively easy to use. In the case of an assessment of 4, i.e., high functionality, the buttons are practical, easy to use, and their functions are intuitive, which facilitates efficient operation. At level 5, the buttons work optimally and are entirely intuitive and error-free, ensuring the highest comfort and efficiency of the tram driver’s work. The characteristics of the functionality criterion are presented in

Table 5. Assessment criterion—button functionality.

Scale	Description
1	Very low functionality. The buttons lack proper action, are challenging to use, and often fail. Their functions are unintuitive and require significant effort to use effectively.
2	Low functionality. The buttons lack some degree of functionality and are difficult to use. The functions are not very intuitive, which causes delays and difficulties in use.
3	Moderate functionality. The buttons are rated for moderate action and ease of use. The functions are relatively intuitive, although they may require a little effort and getting used to.
4	High functionality. The buttons are assigned high performance and ease of use. The functions are intuitive and work as expected, allowing for efficient operation.
5	Very high functionality. The buttons have been assigned optimal operation and ease of use. The intuitive functions work flawlessly and effortlessly, ensuring the highest work efficiency.

. Table 5 was developed by the authors based on empirical analysis of key elements of the evaluation of the button functionality.

4.4. Time to Find the Safety Buttons

The last parameter is the time needed to find safety buttons, such as the bell and emergency braking (STOP button, cab controller). The time is intended to determine how quickly the tram driver can locate and use these buttons in emergencies. The assessment simulated a situation where the tram driver must quickly find and activate the appropriate safety button in the tram cabin. This time was measured from when the tram driver was informed about the need to use the button until it was pressed. A longer time needed to find the button may indicate problems with the location or marking of the buttons, which may affect safety and the speed of reaction in emergencies.

5. Assessment Model for the Representation of the Tram Driver’s Console in the Simulator

Assessment Model

The assessment of the mapping of the tram driver’s control panel in the simulator is based on the theory of fuzzy logic.

The criteria taken into account in the assessment process are presented in Section 4, Assessment criteria for the tram driver’s panel.

Table 6. Description of the control console mapping indicators included in the model.

Assessment Criterion	Parameter	Tram	Simulator	Fidelity Index
perspective (${\mathit{x}}_{\mathit{p}}$)	floor-to-ceiling view range	$x_{p_{t_1}}$	$x_{p_{s_1}}$	$x_p={\displaystyle \frac{\sum _{i=1}^4{\overline{x}}_{p_{s_i}}}{\sum _{i=1}^4{\overline{x}}_{p_{t_j}}}}$
	view range from left to right side of the cabin	$x_{p_{t_2}}$	$x_{p_{s_2}}$
	blind spot area	$x_{p_3}$	$x_{s_3}$
	adjustable seat height	$x_{p_{t_4}}$	$x_{p_{s_4}}$
arrangement of the buttons (${\mathit{x}}_{\mathit{a}}$)	vehicle startup buttons	$x_{a_{t_1}}$	$x_{a_{s_1}}$	$x_a={\displaystyle \frac{\sum _{i=1}^{11}{\overline{x}}_{a_{s_i}}}{\sum _{i=1}^{11}{\overline{x}}_{a_{t_j}}}}$
	buttons from the heating group	$x_{a_{t_2}}$	$x_{a_{s_2}}$
	buttons from the interior lighting group	$x_{a_{t_3}}$	$x_{a_{s_3}}$
	buttons from the mirror and roller shutter group	$x_{a_{t_4}}$	$x_{a_{s_4}}$
	wiper buttons	$x_{a_{t_5}}$	$x_{a_{s_5}}$
	buttons from the exterior lighting group	$x_{a_{t_6}}$	$x_{a_{s_6}}$
	door control buttons	$x_{a_{t_7}}$	$x_{a_{s_7}}$
	driving direction buttons	$x_{a_{t_8}}$	$x_{a_{s_8}}$
	safety group buttons	$x_{a_{t_9}}$	$x_{a_{s_9}}$
	buttons from the cab controller group	$x_{a_{t_{10}}}$	$x_{a_{s_{10}}}$
	buttons from the passenger information computer group	$x_{a_{t_{11}}}$	$x_{a_{s_{11}}}$
funcionality (${\mathit{x}}_{\mathit{f}}$)	vehicle startup buttons	$x_{f_{t_1}}$	$x_{f_{s_1}}$	$x_f={\displaystyle \frac{\sum _{i=1}^{11}{\overline{x}}_{f_{s_i}}}{\sum _{i=1}^{11}{\overline{x}}_{f_{t_j}}}}$
	buttons from the heating group	$x_{f_{t_2}}$	$x_{f_{S_2}}$
	buttons from the interior lighting group	$x_{f_{t_3}}$	$x_{f_{s_3}}$
	buttons from the mirror and roller shutter group	$x_{f_{t_4}}$	$x_{f_{s_4}}$
	wiper buttons	$x_{f_{t_5}}$	$x_{f_{s_5}}$
	buttons from the exterior lighting group	$x_{f_{t_6}}$	$x_{f_{s_6}}$
	door control buttons	$x_{f_{t_7}}$	$x_{f_{s_7}}$
	driving direction buttons	$x_{f_{t_8}}$	$x_{f_{s_8}}$
	safety group buttons	$x_{f_{t_9}}$	$x_{f_{s_9}}$
	buttons from the cab controller group	$x_{f_{t_{10}}}$	$x_{f_{s_{10}}}$
	buttons from the passenger information computer group	$x_{f_{t_{11}}}$	$x_{f_{s_{11}}}$
button finding times (${\mathit{x}}_{\mathit{t}}$)	bell	$x_{t_{t_1}}$	$x_{t_{s_1}}$	$x_t={\displaystyle \frac{\sum _{i=1}^3{\overline{x}}_{t_{t_i}}}{\sum _{i=1}^3{\overline{x}}_{s_{t_j}}}}$
	STOP (emergency braking)	$x_{t_{t_2}}$	$x_{t_s}$
	emergency braking—cab controller	$x_{t_{t_3}}$	$x_{t_{s_3}}$

presents the methodologies for determining the mapping indicators which constitute the basis for the assessment model proposed in this work. Table 6 was developed by the authors based on empirical analysis of key elements of the evaluation of the mapping indicators.

The model uses trapezoidal membership functions. The main reasons for using trapezoidal functions are simplicity of structure, ease of interpretation of results by non-experts, and the possibility of obtaining qualitatively good results [32]. The membership function was developed using an expert-based approach that has already been discussed in the literature [33,34]. The membership function boundary determination process was based on the collaboration of 30 experts. The experts assigned point values ranging from 1 to 100 for the four separate evaluation criteria, taking into account three categories. Ryczyński et al. in their work indicated the use of statistical parameter values: mean and standard deviation in the construction of trapezoidal membership functions [35]. The mean and standard deviation values were used to construct the membership function in the work. The fidelity indices presented in Table 6 represent the ratio of parameters in the simulator to their real-world counterparts. Perspective ${(x}_p$) allows us to evaluate the compliance of the simulator mainly in the field of view. Arrangement of the buttons (${\mathit{x}}_{\mathit{a}}$) refers to the evaluation of the arrangement of groups of buttons and their arrangements. Functionality (${\mathit{x}}_{\mathit{f}}$) measures the degree to which the functionality of individual buttons is reproduced. Button finding time (${\mathit{x}}_{\mathit{t}}$) reflects the user’s interaction time with the control interface responsible for realizing critical functions.

The numerical values assigned in

Table 7. Membership functions in the evaluation model.

Input Variable	Fuzzy Set	Delimiters
perspective ($x_p$)	incompatible	[0 0 0.3 0.4]
	partially_compatible	[0.3 0.4 0.6 0.7]
	compatible	[0.6 0.7 1 1]
arrangement of the buttons $\left(x_a\right)$	incompatible	[0 0 0.3 0.4]
	partially_compatible	[0.3 0.4 0.6 0.7]
	compatible	[0.6 0.7 1 1]
functionality $\left(x_f\right)$	incompatible	[0 0 0.3 0.4]
	partially_compatible	[0.3 0.4 0.6 0.7]
	compatible	[0.6 0.7 1 1]
button finding times $\left(x_t\right)$	incompatible	[0 0 0.3 0.4]
	partially_compatible	[0.3 0.4 0.6 0.7]
	compatible	[0.6 0.7 1 1]

define a trapezoidal membership function, which is described by four parameters (1).

$$\mu \left(x_{p,a,f,t};a,b,c,d\right)=\left\{\begin{array}{c}0 if x_{p,a,f,t}\le a\\ {\displaystyle \frac{x_{p,a,f,t}-a}{b-x_{p,a,f,t}}} if a<x_{p,a,f,t}\le b\\ {\displaystyle \frac{{d-x}_{p,a,f,t}}{d-c}} if c<x_{p,a,f,t}\le d\\ 0 if x_{p,a,f,t}>d \end{array}\right.$$

(1)

The values a and d determine the function values for which the membership degree is 0, while b and c determine the function values for which the membership degree is 1. The overlap of the intervals indicated in Table 7 reflects the features of complex logic, since values can belong to more than one set at the same time. The linguistic terms for the variables presented in Table 7 were determined based on collaboration with 30 experts. Membership functions for input variables are presented in Table 7.

The reasoning in the model for evaluating the driver’s control panel mapping uses Mamdani implications and is based on the MIN and MAX operators. The model uses a set of 81 rules, an example of which is presented in

Table 8. Examples of rules used in the model.

No.	xp	xa	xf	xt	Assessment
1	incompatible	incompatible	incompatible	incompatible	low fidelity
2	incompatible	incompatible	incompatible	partially compatible	low fidelity
….	….	….	….	….	….
37	partially compatible	partially compatible	partially compatible	partially compatible	moderate fidelity
….	….	….	….	….	….
80	compatible	compatible	compatible	partially compatible	high fidelity
81	compatible	compatible	compatible	compatible	high fidelity

. The rules were developed with the participation of the experts mentioned in the work.

In the process of constructing the rules in the model, expert knowledge was used as a basis for formulating the inference rules. This approach has already been used in the literature [36,37]. The inference rules used in the model were developed based on consultations with experts experienced in the use and ergonomics of the control panel, including tram drivers and technical support specialists. The assessment of the mapping of the driver’s control panel considers the four criteria described in the paper. The formal result of fuzzy inference is obtained based on (2) [38].

$$z^{*}={\displaystyle \frac{\int {\mu }_A\left(z\right)·zdz}{\int {\mu }_A\left(z\right)dz}}$$

(2)

where $z^{*}$ represents the final crisp output value for the variable z, and ${\mu }_A\left(z\right)$ is the aggregated output of MF (membership function).

It includes a three-level scale: low, moderate, and high fidelity. The membership function is presented in

Table 9. Membership function for assessing the degree of mapping of the tram driver’s panel.

Input Variable	Fuzzy Set	Delimiters
assessment fidelity	low fidelity	[0 0 0.2 0.4]
	moderate fidelity	[0.2 0.4 0.6 0.8]
	high fidelity	[0.6 0.8 1 1]

.

6. Case Study

6.1. Collection Data

The assessment of the driver’s console was conducted by a group of 30 experts, consisting of 24 tram drivers and six technical support specialists, ensuring a balanced combination of operational experience and technical expertise. The selection of experts was designed to provide a comprehensive evaluation of the simulator’s fidelity, taking into account both practical usability and technical accuracy.

Among the selected tram drivers, a diverse range of professional experience was represented to capture different perspectives on the console’s usability and realism. The distribution of experience levels within this group included six drivers with 0–5 years of experience, six drivers with 5–10 years, six drivers with 10–20 years, and six drivers with over 20 years of experience. This structure ensured that the assessment incorporated the viewpoints of both less-experienced drivers, who rely more on intuitive interactions, and highly experienced drivers, who possess extensive knowledge of real operational conditions. The increased number of participants strengthens the reliability and statistical robustness of the assessment by reducing the impact of individual biases. The same expert group was involved in the assessment of both the real tram and the simulator, enabling a direct comparative analysis. This approach minimizes subjective bias and ensures that the evaluation of the simulator is conducted within the same cognitive framework as the real vehicle assessment.

Additionally, the inclusion of six technical support specialists provided an engineering perspective that complemented the operational feedback from tram drivers. These specialists played a crucial role in identifying discrepancies in the arrangement, functionality, and responsiveness of control elements, ensuring that the evaluation was not solely reliant on subjective assessments but also considered technical consistency with the real system.

The cabins subject to evaluation were evaluated by 30 experts, including tram drivers and technical support specialists. The test subjects were the trams Solaris Tramino S105p, Moderus Gamma LF 06 AC, and Skoda 16T RK.

6.2. Data Analysis

The analysis of expert evaluations presented in

Table 10. Statistical analysis of expert evaluations for the Škoda 16T RK.

Assessment Criterion	Parameter	Tram				Simulator
		$\stackrel{\mathbf{-}}{\mathbf{x}}$	Lower 95% CI	Upper 95% CI	CV	$\stackrel{\mathbf{-}}{\mathbf{x}}$	Lower 95% CI	Upper 95% CI	CV
perspective ($x_p$)	floor-to-ceiling view range	4.47	4.16	4.77	0.18	2.23	2.07	2.39	0.19
	view range from left to right side of the cabin	3.80	3.50	4.10	0.21	1.87	1.74	2.00	0.19
	blind spot area	3.57	3.38	3.75	0.14	1.87	1.74	2.00	0.19
	adjustable seat height	4.00	3.80	4.20	0.13	2.43	2.25	2.62	0.21
arrangement of the buttons $\left(x_a\right)$	vehicle startup buttons	4.63	4.37	4.90	0.16	3.47	3.19	3.74	0.21
	buttons from the heating group	4.90	4.79	5.01	0.06	2.23	2.07	2.39	0.19
	buttons from the interior lighting group	4.80	4.65	4.95	0.08	3.20	3.02	3.38	0.15
	buttons from the mirror and roller shutter group	4.57	4.38	4.75	0.11	3.13	2.94	3.32	0.16
	wiper buttons	4.90	4.79	5.01	0.06	3.00	2.78	3.22	0.20
	buttons from the exterior lighting group	4.27	4.03	4.51	0.15	2.33	2.15	2.51	0.21
	door control buttons	4.23	3.94	4.52	0.18	2.97	2.74	3.20	0.21
	driving direction buttons	4.53	4.34	4.72	0.11	2.63	2.45	2.82	0.19
	safety group buttons	4.73	4.54	4.93	0.11	3.17	3.03	3.31	0.12
	buttons from the cab controller group	4.67	4.42	4.91	0.14	4.20	3.88	4.52	0.20
	buttons from the passenger information computer group	4.77	4.61	4.93	0.09	4.20	3.99	4.41	0.13
functionality $\left(x_f\right)$	vehicle startup buttons	4.80	4.65	4.95	0.08	4.00	3.78	4.22	0.15
	buttons from the heating group	4.90	4.79	5.01	0.06	2.50	2.29	2.71	0.23
	buttons from the interior lighting group	4.93	4.84	5.03	0.05	2.20	2.05	2.35	0.18
	buttons from the mirror and roller shutter group	4.83	4.69	4.97	0.08	2.83	2.69	2.97	0.13
	wiper buttons	4.87	4.74	5.00	0.07	2.30	2.13	2.47	0.20
	buttons from the exterior lighting group	4.73	4.57	4.90	0.10	2.17	2.03	2.31	0.17
	door control buttons	4.87	4.74	5.00	0.07	2.87	2.70	3.03	0.15
	driving direction buttons	4.80	4.65	4.95	0.08	2.13	2.00	2.26	0.16
	safety group buttons	4.87	4.74	5.00	0.07	3.10	2.99	3.21	0.10
	buttons from the cab controller group	4.83	4.69	4.97	0.08	4.23	4.07	4.39	0.10
	buttons from the passenger information computer group	4.83	4.69	4.97	0.08	4.60	4.31	4.89	0.17
button finding times $\left(x_t\right)$	bell	3.63	3.39	3.87	0.18	4.53	4.29	4.77	0.14
	STOP (emergency braking)	2.51	2.34	2.68	0.19	3.39	3.18	3.61	0.17
	emergency braking—cab controller	2.51	2.37	2.64	0.14	3.44	3.18	3.70	0.20

, which includes results for a single analyzed vehicle, the Škoda 16T RK, allowed for the assessment of their statistical reliability. The study considered statistical indicators such as the arithmetic mean, coefficient of variation (CV), and the 95% confidence interval (CI). The same analysis was conducted for the remaining two vehicles. For all evaluation categories, the CV values are below 25%, indicating low variability in responses and high consistency of expert assessments. For each analyzed parameter, the calculated mean rating falls within the designated 95% confidence intervals, confirming the reliability and stability of the obtained results. The narrow width of the intervals indicates high precision of the results and limited estimation uncertainty. Based on the conducted analyses, it can be concluded that the obtained data exhibit homogeneity and stability, which enables the formulation of reliable conclusions regarding the evaluated system. The statistical analysis presented in Table 10 was based on the responses of the experts participating in the study.

To verify the stability of the proposed evaluation framework, the study was conducted in two stages. In the first stage, surveys were completed by 10 experts, while in the second stage, the sample was expanded to 30 experts, which constituted the final dataset used for the analysis. This approach enabled verification of whether increasing the number of respondents would substantially affect the evaluation results.

Table 11. Comparison of expert evaluation results for the Skoda 16T RK tram obtained from 10 and 30 experts.

Parameter	σ (10 Experts)	σ (30 Experts)	Relative Percentage Difference	p (Welch’s t-Test)	p (Mann–Whitney U Test)
floor-to-ceiling view range	0.70	0.82	−2.90	0.62	0.75
view range from left to right side of the cabin	0.82	0.81	2.70	0.74	0.74
blind spot area	0.53	0.50	1.90	0.73	0.73
adjustable seat height	0.57	0.53	−2.40	0.63	0.62
vehicle startup buttons	0.68	0.72	−1.40	0.79	0.83
buttons from the heating group	0.32	0.31	0.00	1.00	1.00
buttons from the interior lighting group	0.42	0.41	0.00	1.00	1.00
buttons from the mirror and roller shutter group	0.53	0.50	1.50	0.73	0.73
wiper buttons	0.32	0.31	0.00	1.00	1.00
buttons from the exterior lighting group	0.68	0.64	−0.80	0.89	0.89
door control buttons	0.82	0.77	5.80	0.44	0.42
driving direction buttons	0.53	0.51	0.70	0.86	0.87
safety group buttons	0.68	0.52	0.70	0.89	0.93
buttons from the cab controller group	0.68	0.66	−0.70	0.89	0.86
buttons from the passenger information computer group	0.48	0.43	1.40	0.70	0.69
vehicle startup buttons	0.00	0.41	−4.00	0.01	0.14
buttons from the heating group	0.00	0.31	−2.00	0.08	0.32
buttons from the interior lighting group	0.00	0.25	−1.30	0.16	0.43
buttons from the mirror and roller shutter group	0.00	0.38	−3.30	0.02	0.18
wiper buttons	0.00	0.35	−2.70	0.04	0.24
buttons from the exterior lighting group	0.00	0.45	−5.30	0.00	0.08
door control buttons	0.00	0.35	−2.70	0.04	0.24
driving direction buttons	0.00	0.41	−4.00	0.01	0.14
safety group buttons	0.00	0.35	−2.70	0.04	0.24
buttons from the cab controller group	0.00	0.38	−3.30	0.02	0.18
buttons from the passenger information computer group	0.00	0.38	−3.30	0.02	0.18
bell	0.93	0.64	3.10	0.73	0.98
STOP (emergency braking)	0.61	0.47	−4.90	0.55	0.69
emergency braking—cab controller	0.50	0.36	−4.30	0.52	0.67

presents a comparison of the two groups, including standard deviation (σ), relative percentage difference, and significance levels from Welch’s t-test for independent samples and the Mann–Whitney U test.

A comparison of the results of both evaluations allows us to determine whether the simulator faithfully reproduces the real tram driver’s cabin for a given type of vehicle. The values of the indicators of the tram driver’s control panel reproduction in the simulator for each of the analyzed criteria are presented in Table 12.

The results show that most criteria exhibit only minor changes in mean values, with differences in the range of 5%. For the majority of variables, the p-values are well above the 0.05 threshold in both tests, confirming that the observed differences are not statistically significant. This demonstrates a high degree of consistency and stability in the evaluations, regardless of the sample size. In some criteria, p-values below 0.05 were obtained in the t-test, indicating statistically significant differences. These cases are primarily associated with criteria that had no variance in the 10-expert group, where all respondents gave identical maximum ratings.

A general limitation of the proposed method is its dependence on the diligence and reliability of survey responses. However, this underscores the importance of performing a statistical analysis of agreement, as presented in Table 10, which confirms the consistency and reliability of the collected data.

The limited change in results after increasing the number of experts is explained by the homogeneity of the expert group, as all evaluations were conducted within a single urban transport company, involving experienced drivers and technical specialists. Conducting the assessment in other companies is not feasible, as trams are custom-made and no two operators use identical tram driver control panels. The number of such qualified experts is limited, making it difficult to recruit more than 30 participants. Including a larger group would require the participation of many less-experienced drivers, which could alter the results but also lead to a misinterpretation of the evaluations due to insufficient familiarity with the technical aspects of the assessed trams.

6.3. Assessment

The conducted research involved the analysis of three objects: Solaris Tramino S105p, Moderus Gamma LF 06 AC, and Škoda 16T RK. The results presented in Table 12 were implemented into the model developed in the MATLAB environment. The calculation results—fidelity index values—are presented in Table 12. Table 12 was created by the authors and is based on results obtained in MATLAB software, version R2024a.

Table 12. Assessment of the driver’s control panel mapping.

	Criteria	$\mathbf{Perspective} \mathbf{\left(}{\mathit{x}}_{\mathit{p}}\mathbf{\right)}$	$\mathbf{Arrangement} \mathbf{of} \mathbf{the} \mathbf{Buttons} \left({\mathit{x}}_{\mathit{a}}\right)$	$\mathbf{Functionality} \left({\mathit{x}}_{\mathit{f}}\right)$	$\mathbf{Button} \mathbf{Finding} \mathbf{Times} \left({\mathit{x}}_{\mathit{t}}\right)$	Fidelity Index
Type of Tram
Solaris Tramino S105p		0.8	0.5432	0.545	0.576	0.5
Moderus Gamma LF 06 AC		0.945	0.954	0.977	0.927	0.847
Skoda 16T RK		0.53	0.677	0.618	0.765	0.727

Table 12. Assessment of the driver’s control panel mapping.

	Criteria	$\mathbf{Perspective} \mathbf{\left(}{\mathit{x}}_{\mathit{p}}\mathbf{\right)}$	$\mathbf{Arrangement} \mathbf{of} \mathbf{the} \mathbf{Buttons} \left({\mathit{x}}_{\mathit{a}}\right)$	$\mathbf{Functionality} \left({\mathit{x}}_{\mathit{f}}\right)$	$\mathbf{Button} \mathbf{Finding} \mathbf{Times} \left({\mathit{x}}_{\mathit{t}}\right)$	Fidelity Index
Type of Tram
Solaris Tramino S105p		0.8	0.5432	0.545	0.576	0.5
Moderus Gamma LF 06 AC		0.945	0.954	0.977	0.927	0.847
Skoda 16T RK		0.53	0.677	0.618	0.765	0.727

The assessment results of the representation of the driver’s control panel for the Solaris Tramino S105p tram indicate that, according to the fidelity index (0.5), this model was classified as moderate fidelity. The simulated panel differed from the real object, primarily due to the modular nature of the actual tram, which features control screens on both the left and right sides. Given the universal character of the simulator, a fully faithful reproduction of this configuration was not feasible. As a result, certain buttons had to be relocated to the side console. Additional discrepancies involved the switch representation, the cab controller (notably missing the bell), and the design of the central screen interface. The altered positions of safety-critical buttons (e.g., bell, STOP—emergency braking, and the emergency brake assigned to the cab controller) led to a lower score in button finding time (0.576), likely due to increased search time.

In contrast, the Moderus Gamma LF 06 AC tram received the highest fidelity index (0.847), classifying it as high fidelity. The high degree of similarity between the simulator and the real tram was confirmed by very high scores across all individual criteria, including perspective (0.945), arrangement of the buttons (0.954), functionality (0.977), and button finding times (0.927). These results indicate that the simulator offers a highly realistic and operationally consistent experience for this tram model.

The Škoda 16T RK tram was classified within the region of the moderate fidelity and high fidelity fuzzy sets, with an overall fidelity index of 0.727, indicating partial membership in both categories. Although this value places the model near the lower boundary of the highest fidelity category, the individual assessment scores remained moderate. Specifically, button finding time (0.765) and arrangement of the buttons (0.677) were rated relatively high, while perspective (0.53) and functionality (0.618) received lower evaluations. The low rating in perspective was primarily due to a difference in seating position—in the real Škoda tram, the driver sits significantly higher than in the simulator, which alters the visual perception of the control panel and surrounding environment. This suggests that while the overall representation ensures acceptable consistency and usability, certain visual and ergonomic aspects of the real tram were only partially replicated in the simulator.

7. Conclusions

In the literature and on the market, there is a growing demand for tools supporting training processes. One such tool is a simulator. Reproducing real working conditions is essential for reliable and accurate training. Simulators are a training tool that must be constantly improved and adapted to changing technologies and standards. Adaptability to changing working conditions of tram drivers in urban transport is essential. Tram drivers perform transport tasks, often changing the vehicle model, which is challenging in training. So far, no models have been identified in the literature that would allow for assessing the reproduction of the working environment in a simulator. This article fills this gap.

The study demonstrated that meticulous representation of the driver’s console in the simulator is crucial for the realism and effectiveness of training. Participants drew attention to the need to standardize the arrangement of buttons between different vehicle models. Despite this, thanks to the use of interchangeable slots and different panel configurations, employees could locate the simulated vehicle in various configurations and easily identify its functionalities.

The proposed evaluation model enables the identification of fidelity levels for driver control panel representation. The model is an evaluation tool that can be used by urban transport operators, manufacturers, and training centers to ensure high-level vehicle representation. Urban transport operators can use the proposed approach to evaluate and analyze existing simulator configurations. Manufacturers can also use the model to support the design of ergonomic and realistic human–machine interfaces.

One of the study’s key conclusions is that the use of physical buttons in the simulator significantly contributed to increasing the degree of representation of the dashboard. Using real elements improved realism, which is essential for training effectiveness. Physical buttons allow participants to better understand the console’s layout and functioning, which translates into faster finding and operating of the safety buttons. Currently, existing universal simulators are equipped with touch screens imitating real buttons. Building a console with physical buttons reflecting every tram model is impossible. Nevertheless, planned training on the universal console with quality assessment will reduce the number of errors made by tram drivers. The simulator used in this study was designed to address these challenges. It is mounted on a motion platform that faithfully reproduces the dynamics of tram motion, including acceleration and braking, which significantly enhances the realism of the training experience. The visual system, based on five high-resolution monitors, ensures a wide field of view and provides high visual immersion. A key element is the physical driver’s console, which features real buttons and switches instead of touchscreen interfaces typically used in standard solutions. The console is built in a modular form, enabling the replacement of individual control panels to match different tram models. These features distinguish the simulator from other simulators available on the market.