Evaluation of the Efficiency of Solutions Used at Active Pedestrian Crossings

Abstract

This article examines the components of active pedestrian crossing systems, focusing on an innovative solution implemented at Jaworski Street in Kielce. The proposed method for evaluating the effectiveness of this system is based on observations and performance analysis under conditions of limited visibility and increased pedestrian traffic. The key elements of the methodology include assessing the accuracy of pedestrian detection using IR/MW and curtain sensors, analyzing system errors such as false activations or failures to detect pedestrians, and evaluating driver responses to warning signals, with an emphasis on optimizing signal activation timing. The innovation of the method lies in its flexibility and real-time adaptability to dynamic traffic conditions, significantly enhancing system performance and pedestrian safety at crossings. Through integrated detection and activation time management mechanisms, the system can respond to changing conditions in real time, ensuring optimal operational effectiveness.

1. Introduction

Among the solutions aimed at improving road safety, a growing number of them are focused on the most vulnerable group of road users—pedestrians. In the scope of vehicle design, the concepts of shaping the geometry of vehicle hoods so that they are free of sharp edges or installing external airbags for pedestrians are well known. Although these solutions aim to improve safety in the event of an accident, the key objective should be to reduce the number of road incidents.

According to [1], the basic problems of road traffic safety include dangerous behavior of road users, poor quality of road infrastructure, as well as the lack of an effective system for managing road traffic safety. In response to such problems, we observe rebuilding and modernization of existing pedestrian crossings, which improve the comfort of both pedestrians and drivers.

An extensive analysis of systematic and other studies of pedestrian behavior and pedestrian–driver relations in Poland and other countries is presented in [2]. In countries where analyses of pedestrian and pedestrian–driver behavior are carried out (including the United States, Australia, New Zealand, Switzerland, France, Great Britain, Ireland, Germany, Holland, Denmark, Spain, Sweden, Belgium, Italy, Finland, and Austria), it was found that they are not carried out systematically but only for the assessment of selected issues.

In 2017, the main reasons for pedestrian collisions were failure to yield the right of way to pedestrians at pedestrian crossings [3].

Table 1. Road accidents by place of occurrence—based on [3].

Place of Occurrence	Accidents	Injured	Fatalities
Roadway	24,198	2198	30,306
Pedestrian crossing *	4322	267	4309
Hard shoulder	967	150	1155
Abutment, ditch	695	100	865
Sidewalk, pedestrian way	589	22	628
Driveway, fields	283	12	328
Parking lot, square, rest and service area	257	7	270
Public transport stop	170	8	177
Tramway intersection, tracks	170	11	220
Bicycle lane, intersection, lock	803	8	821

* Concerns accidents with pedestrian participation as well as other accidents, e.g., accidents resulting from cyclists passing through pedestrian crossings.

presents road accidents and the consequences thereof by location of their occurrence. Pedestrian crossings are the second most common places for road incidents, right behind the roadway. It is therefore paramount to implement innovative systems, especially at the intersection of pedestrian and road traffic.

Designating a pedestrian crossing at crossroads gives pedestrians the conviction, unfortunately often false, of being visible to motorists, increasing the confidence of pedestrians in the use of the marked crossing [4]. The introduction of a constantly pulsating light at pedestrian crossings has become a popular solution. However, the use of constantly pulsating yellow lights can lead to a dangerous phenomenon—drivers become accustomed to this type of signaling, which reduces their concentration and reaction time to a pedestrian entering the road in the crossing area. Therefore, on the one hand, the false conviction of pedestrians about their safety and their willingness to use marked crossings and, on the other hand, the fact that drivers grow accustomed to constantly pulsating lights located at crossings can lead to dangerous situations and road incidents.

Pedestrian safety represents a significant challenge in the context of improving road traffic safety, as pedestrians account for over 20% of all fatal road accidents worldwide [5].In Poland, pedestrians constitute more than 22% of road accident participants, with nearly 27% being fatalities [6]. Unsignalized pedestrian crossings are particularly hazardous, especially where there is high pedestrian traffic, high vehicle speeds, and limited visibility, which collectively increase the risk of accidents. In response to these risks, road authorities have implemented measures to enhance safety, such as road narrowing, crossings with additional traffic-calming features, and so-called active crossings equipped with warning lights that respond to pedestrian presence.

In the literature addressing driver–pedestrian behavior, studies have examined driver behaviors [7], pedestrian speeds when crossing roads [8], and the interactions between drivers and pedestrians at crossings [9,10,11]. Prior research on pedestrian crossings has primarily focused on evaluating the effectiveness of safety measures aimed at improving pedestrian safety [12]. Analytical findings indicate a positive influence of flashing warning signals and supplementary lighting on driver behavior [13]. Additionally, the use of LED curb lighting combined with flashing signals has proven effective in enhancing visibility and driver responsiveness [14].

The identified research gaps predominantly concern the long-term impact of interventions on road user behavior. Many studies analyze changes in driver and pedestrian behavior shortly after the introduction of a given solution but lack long-term analyses spanning several years. This perspective is crucial, as drivers may become accustomed to implemented measures over time, potentially reducing their effectiveness.

In this study, the authors focus on traffic safety as the primary evaluation criterion. Traffic safety is not a typical performance metric, as it is challenging to define and measure using a single, universal indicator. This article employs non-standard metrics, such as the efficiency of pedestrian detection by sensor systems, the number of false alarms, and the percentage of drivers yielding to pedestrians at activated crossings. While these are not conventional performance indicators, they allow for assessing the effectiveness of active pedestrian crossings from a traffic safety perspective.

The aim of this study is to evaluate the long-term effectiveness of implemented solutions using non-standard performance metrics. The traffic safety system was analyzed for its effectiveness in achieving its intended goal: a high rate of vehicle stops at pedestrian crossings when the warning signal system is activated.

Research incorporating long-term changes in road user behavior provides a reliable assessment of the effectiveness of applied measures. By examining these changes over time, this study contributes to a comprehensive understanding of the sustained impact of safety interventions on driver and pedestrian behavior.

The proposed method for evaluating the effectiveness of the active pedestrian crossing system is based on conducting observations and analyzing the system’s performance under conditions of limited visibility and increased pedestrian traffic. The aim of the study is to assess how effectively the system detects pedestrians and vehicles, how it responds to unintended activations (e.g., by pedestrians not intending to cross), and how the activation timing of light signals influences driver behavior.

The key elements of the methodology:

• Data Collection:

This study was conducted during afternoon and evening hours, encompassing 1000 vehicle trials and 300 pedestrian crossing activation attempts. This sample size ensures the representativeness of the results.

• Pedestrian Detection Assessment:

The system was evaluated for its effectiveness in detecting pedestrians within the detection zones of curtain sensors and IR/MW sensors, enabling an assessment of the system’s accuracy.

• System Error Analysis:

Cases where the system exhibited detection anomalies were identified. These included inappropriate crossing activations, such as false detections of pedestrians not intending to cross, and failures to detect pedestrians who intended to cross.

• Driver Reactions:

This study investigated how drivers responded to the activated warning signals, particularly in the context of yielding to pedestrians. Special attention was given to how the duration of warning signals influenced driver habituation.

• Signal Timing Optimization:

The efficiency of the system in managing the activation timing of signals was analyzed, with adjustments made for extended signal durations depending on detected pedestrian or vehicle activity.

The method aims to provide a comprehensive evaluation of detection accuracy, minimization of false activations, and driver responses to the operation of the active pedestrian crossing system. This allows for an assessment of how effectively the system enhances pedestrian safety at crossings, particularly under conditions of limited visibility.

2. An Example of the Components of an Active Pedestrian Crossing System

The components of an active pedestrian crossing (APC) and a brief description of their functions are presented below:

• Anti-Slip Surfaces enhance vehicle braking before the crossing, reducing stopping distance.
• Warning Lights (Pulsating and Steady) emit light to improve the visibility of the pedestrian crossing and draw drivers’ attention.
• Reflective Elements installed along the crossing enhance visibility at night and in adverse weather conditions.
• Pedestrian Detection Sensors detect the presence of pedestrians and activate the warning signaling system.
• The Warning Signaling System automatically activates when a pedestrian is near the crossing, triggering the warning lights.

Each component of the active pedestrian crossing fulfills a specific function.

Anti-Slip Surfaces: the application of a specialized anti-slip surface aims to enhance pedestrian safety by reducing the risk of accidents caused by excessive vehicle speed near the crossing. The stopping distance of a vehicle depends on multiple factors, including its speed and weather conditions. A high-friction surface significantly shortens this distance, which is critical for ensuring pedestrian safety.

Warning Lights: illuminated elements play a vital role in capturing drivers’ attention. Pulsating yellow lights emitted from the crossing increase its visibility to approaching vehicles, thereby influencing the driver’s reaction dynamics. Additionally, depending on the variant used, white or yellow lights can be placed beyond the crossing, which may emit steady or pulsating light to enhance its effectiveness.

Reflective Elements: reflective components installed along the crossing (e.g., perpendicular to the direction of travel) improve visibility in low-visibility conditions. Their purpose is to increase recognition of the crossing from both short and long distances, enabling drivers to detect pedestrians earlier and assess the situation more quickly.

Pedestrian Detection Sensors: active pedestrian crossings are equipped with sensors that detect the presence of pedestrians in the crossing area. When a pedestrian is within the sensor’s range, the system automatically activates the warning signals.

Warning Signaling System: the warning signaling system is a core element of an active pedestrian crossing, automatically triggered upon pedestrian detection. It typically involves activating pulsating yellow lights to alert drivers to the presence of pedestrians at the crossing. With advanced signaling technologies, the system can also adjust the duration of warnings based on prevailing conditions.

Active pedestrian crossings (APCs) represent advanced technologies that integrate various infrastructural elements and detection systems to enhance pedestrian safety. The anti-slip surface, warning lights, reflective elements, pedestrian detection sensors, and warning signaling system work together to improve crossing visibility, attract driver attention, and increase reaction time to pedestrian presence. The effectiveness of these solutions depends on proper design, calibration, and integration with the existing road infrastructure.

The main elements of an active pedestrian crossing (APC) are presented in Figure 1. The presented pedestrian crossing is equipped with a special anti-skid surface supporting the driver’s braking directly before the crossing. The anti-skid effect can be achieved by replacing the existing bituminous surface with a new one with a rougher aggregate. The braking distance depends on a number of factors, in particular the speed of the incoming vehicle. According to [15], in the case of a passenger car driving on a dry and even surface, the entire braking distance from the moment the hazard is detected at the speed of 50 km/h is 28.8 m, while at 60 km/h, the distance increases to 37.4 m. A road surface with a sufficiently high friction can significantly reduce braking distance. The presented solution concerns a surface made of polyurethane resin and bauxite aggregate with a grain size of 1–3 mm.

It should be stressed that the occurrence of unfavorable weather conditions, rain, snow, or ice will reduce the vehicle’s grip and thus increase the braking distance.

The use of a roughening material ensures adequate anti-skid properties. The measure of the roughness of the marking is the roughness index—the SRT (Skid Resistance Tester), which refers to the marking made with paints and chemically cured masses. For highways and expressways, the value of the SRT roughness index must be at least 50 SRT units (class S2), and that for other roads must be 45 SRT units (class 1) [17,18].

Increased driver attention is focused on the elements emitting pulsating yellow light from the side of the oncoming vehicle and, depending on the variant used, white or continuous yellow (or pulsating) light behind the pedestrian crossing. The absence of provisions and harmonized standards for active spot retroreflectors placed perpendicularly to the direction of travel does not impose or specify the manner of installing particular elements. Therefore, some solutions found in the above combinations can cause chaos in the drivers’ intuitive interpretation of the road situation.

According to [19], the color of the light reflected by the spot retroreflectors should be

• White for permanent traffic organization, with the exception of lines indicating the right edge of the roadway;
• Red, indicating the right edge of the roadway;
• Yellow, for marking temporary changes in traffic organization, e.g., in case of road work.

This applies to reflections along the road lane.

Yellow light used as a warning can be interpreted by drivers as temporary, as yellow markings, signs, lights, or information boards are not intended for permanent solutions. Red elements indicate the right edge of the roadway, while white elements indicate the left edge.

In order to improve the visibility of pedestrians at crossings, innovative lighting systems are introduced for crossing areas and within the waiting area. Specialized directional lights allow for correct assessment of the situation at the crossings not only by the driver but also by pedestrians.

An equally interesting and innovative solution is a selective pedestrian recognition system. The lights are then equipped with video analysis devices. When a person enters the area, the algorithm determines the size of the object and is able to automatically increase the light intensity at the crossing.

There is a wide range of lighting dedicated for pedestrian crossings, with symmetrical and focused optical systems available on the market [20]. However, such a beam does not provide adequate illumination on the entire area of the pedestrian crossing. As a result, LED lighting emitting a beam with an asymmetrical light distribution curve has proven to be a success. They provide a high contrast between the luminance or chrominance of the pedestrian and the background. The luminance contrast determines the ratio of the brightness level of the illuminated object to the brightness of the background. The chrominance contrast, on the other hand, refers to the shade and saturation of the color of the pedestrian against the background on which they are seen [21].

Depending on the lighting conditions at the crossing and the geometry of the crossing, creating a positive contrast between the pedestrian and the background is key. The light is introduced horizontally in order to illuminate the pedestrian vertically while avoiding glare hindering the driver’s vision [22]. Optimizing beam distribution requires continuous research into optics and lighting systems that provide the proper lighting parameters.

An example of luminous intensity distribution for lighting with an asymmetrical light beam distribution is presented in Figure 2 in the vertical plane and in Figure 3 in the horizontal plane of a pedestrian crossing.

The simulated case presented here concerns a 9 × 4 m pedestrian crossing. The mounting height of the lighting fixtures in relation to the road surface was 4.5 m. Average vertical luminous intensity obtained in the pedestrian crossing axis of Ev = 70.13 lx as well as average horizontal luminous intensity at the pedestrian crossing and in the waiting area of Eh = 74.47 lx were confirmed after construction work and field research [20]. The principle of using two lighting fixtures, illuminating pedestrians from the direction of oncoming traffic, has been respected.

3. Analysis of the Smart Pedestrian Crossing Installed at the Jaworski Street in Kielce

A detection system at the pedestrian crossing at Jaworski street was built in a spot where a number of dangerous road behavior situations were observed, with a very high intensity of road traffic in both directions and very high pedestrian traffic. The pedestrian crossing is located on the arch of the road, where pedestrians had to cross two roadway lanes in each direction, with the use of the narrow (2 m) refuge island between the opposite direction lanes—Figure 4. Along with the construction of the detection and warning system at the crossing, changes were made to the marking, consisting of narrowing the roadway to one lane in each direction and widening the central refuge island to 3.5 m.

3.1. Parameters of the Components Used

In the analyzed case, in addition to typical lighting for pedestrian crossings, a system of warning lights mounted on lighting poles and active spot retroreflectors built into the road surface were also used.

The Smartpole Crossing composite lighting pole (Figure 5) is equipped with a dual detection system in the pedestrian crossing area: an IR curtain motion detector and an IR/MW motion detector.

Dual-arm lighting poles with a height of 10 m were equipped with additional booms to suspend the lighting to illuminate the crossing from a height of 6 m. A warning signal in the form of three yellow LED elements was integrated into the poles. Early warning lights with a diameter of 10 cm were placed 40 cm above the D-6 road sign. In addition, the poles were equipped with D-6 and T-27 road signs unilaterally illuminated along the perimeter of the sign. The signs are illuminated when the traffic lighting is active [23]. Inside the pole, there is also a control system for the active pedestrian crossing.

The components of the smart crossing system include active spot retroreflectors mounted in front of each conditional stop line. They are double-sided with yellow warning LEDs. The shape of the spot retroreflectors ensures smooth passage of vehicles and snowplows and guarantees high durability of the element. The active retroreflector insert is resistant to temperatures ranging from −35 °C to +70 °C, made of high-impact, clear polyethylene with IP68 protection [24]. The insert is equipped with three LEDs for each direction—from the oncoming side and from the horizontal marking side. The bodies of the retroreflectors are permanently placed in the abrasive layer of the roadway, and the active elements are connected by cables covered with roadway filler (Figure 6).

The D-6/T-27 vertical road signs made of translucent foil are a noteworthy element of the entire system as well. Such technology allows for the background of the sign to be illuminated, while also maintaining all the reflectivity parameters on the face of the sign, illuminated by the car headlights. The use of such signs significantly increases the visibility of the pedestrian crossing for drivers, as it is visible even before the car headlights illuminate the vertical road sign. In the case of this pedestrian crossing, it is of great importance, as it is located on an arch of the road, so the signs become illuminated by the vehicle headlights later than in the case of crossings on straight road sections. Translucent foil enables even illumination of the entire surface of the sign (Figure 7).

3.2. Operation Principle of the Active Pedestrian Crossings

The operation of the warning signaling system depends on the pedestrian sensor system. Detection zones at the pedestrian crossing area detect people attempting to cross the road.

Figure 8 shows the detection field of the curtain sensor, and the side and top-down views. The algorithm of the digital signal analysis from PIR detectors causes the sensor not to react to animals. Moreover, the sensor is characterized by a very narrow detection filed, 0.5 m at a distance of 5 m. When motion is detected, the controller activates fixed spot retroreflectors and early warning lights to flash for 20 s, informing drivers about pedestrians entering the crossing area. Each subsequent time the sensors are tripped (caused by successive pedestrians entering the crossing as the flashing lights are active) causes the mode activation time to be extended by 7 s.

The location and detection fields of the sensors recording pedestrian traffic are presented in Figure 9. The areas located immediately before the pedestrian crossing curtain sensors are marked 1A and 1B. As soon as a pedestrian appears, the system is triggered to activate.

Sensors that detect the appearance of a pedestrian trigger active retroreflectors (yellow on the side of oncoming vehicles and on the side of pedestrians) and a warning light over the D-6 and T-27 vertical road signs. Then, the control of the signal from IR + MW sensors (2A and 2B) takes place. If a pedestrian is at the crossing while the system is active, sensor 2A, followed by sensor 2B, detects the movement and modifies the duration of active crossing. The order of 2A and 2B sensor activation allows for the direction of pedestrian traffic to be evaluated. There is also a case when, with the system active, a vehicle passes through the areas of 2A and/or 2B, causing abnormal detector action, in which case, the system remains triggered for 10 s.

The algorithm for the pedestrian detection system operation is outlined as follows (Algorithm 1):

Algorithm 1: Pedestrian detection algorithm

IF curtain sensor (1A or 1B) detects movement THEN

ACTIVATE reflective elements and warning lights for 20 s

WHILE curtain sensor (1A or 1B) detects movement OR IR/MW sensors (2A or 2B) detect movement

EXTEND system activation time by 7 s

IF IR/MW sensors (2A or 2B) detect vehicle movement THEN

EXTEND system activation time by 10 s

END IF

END WHILE

END IF

This pseudocode describes the operational logic of the active pedestrian detection system, illustrating its responsiveness to pedestrian and vehicle movement. It highlights the system’s ability to dynamically adjust activation times based on real-time input from sensors.

4. Experience in the Use of Active Pedestrian Crossings

The experience gained so far with smart pedestrian crossings is still growing in importance. Pilot systems are being implemented in many Polish cities: Gdańsk, Puck, Olsztyn, Warsaw, Kielce, Radom, Poznań, Ełk, Grudziądz, and others.

Efficiency studies show that the percentage of pedestrians being given the right of way has increased, but not on all the analyzed arteries. The introduction of active crossings did not bring any significant improvement to large traffic routes, three-way roads, and larger roads. For roads smaller than three lanes, the waiting time for pedestrians to cross the road has been reduced. The speed of vehicles approaching the crossing has also decreased.

For the purpose of the case under study, a smart crossing on a two-way, single roadway road with one lane in each direction, with a refuge island between the opposite direction lanes, was studied. The pre-calibration of the active pedestrian crossing system at Jaworski street in Kielce did not provide satisfactory results in detecting pedestrians. The system incorrectly registered vehicles as users subject to detection in the crossing area activating the system, thus sending erroneous messages to motorized road users. Moreover, the arrival of pedestrians intending to enter the crossing did not always result in a proper reaction of the system; not triggering the system was a common occurrence.

The pedestrian detection zone has therefore been calibrated in such a way as to minimize the risk of false tripping of the sensor beam (e.g., by passing vehicles) and to cover 100% of the width of the sidewalk leading to the crossing. The motion detector is equipped with beam height and radius adjustments. After many attempts to set the correct mounting angle of the sensors and the sensitivity of the detectors, a satisfactory level of detection was achieved. After the sensor adjustment, the effectiveness of the smart pedestrian crossing was reassessed.

5. Methodology of Efficiency Assessment of Pedestrian Detection

In order to assess the efficiency of the solutions used at the active pedestrian crossing, it was necessary to observe the correctness of operation—the triggering of the active crossing.

False-positive triggering of an active pedestrian crossing may result from several factors. These include, first of all, triggering of the crossing system by oncoming vehicles or by pedestrians moving in parallel on the sidewalk without any intention of crossing the street. Another incorrect operation is the system not triggering when a pedestrian appears. The studies were carried out in the afternoon and evening in a period of peak road traffic and increased pedestrian traffic. The analyzed cases relate to the operation of the system in adverse external conditions.

This article includes applied formulas to enhance the precision of the description and facilitate a better understanding of the system’s operation.

This study presents the calculation of two key performance indicators:

System Response Efficiency Indicator, calculated as the ratio of the number of correct activations by pedestrians to the total number of pedestrian activation attempts;

Vehicle Stopping Indicator, defined as the percentage of vehicles that stopped at the pedestrian crossing after system activation, relative to the total number of vehicles approaching the crossing during the experiment.

To ensure that the method adopted reflects the representative efficiency of the solution, the experiment was carried out on 1000 samples of vehicles and 300 samples of pedestrian crossings.

Table 2. Results of the observation of the active pedestrian crossing [own elaboration].

Users	System Status	Number of System Response Tests	Efficiency Ratio [%]
Vehicles	correct operation (not triggered)	993	99.3
	incorrect operation (triggered)	7	0.7
Pedestrians	correct operation (triggered)	274	91.1
	incorrect operation (not triggered)	21	7.1
	incorrect operation no intention of crossing the street(triggered)	5	1.8

shows the results of the experiment conducted.

With regard to pedestrian safety, in unfavorable weather conditions in particular, it is required that the system provides the best possible performance at dusk and at night time, as it operates in a state of limited visibility.

Vehicle stopping rates were also recorded, with the system triggered and a clear expectation of pedestrians to cross the road—

Table 3. Vehicle stopping rates at the active pedestrian crossings [own elaboration].

Users	Driver Response	Number of Vehicles in the Experimental Test	Vehicle Stopping Rate [%]
Vehicles	Yielding the right of way (active crossing)	139	27.8
	Not yielding the right of way (active crossing)	361	72.2

. In order to obtain reliable data, the sample size was set at 500 observations.

5.1. Analysis of the Obtained Results

High efficiency of the system was observed in the detection of pedestrians in the detection field of the curtain sensors. The efficiency was over 91%. As for pedestrians causing incorrect operation of the system—triggering the system without the intention of crossing the street—the percentage was very low and amounted to 1.8%. The share of the system functioning incorrectly and not being triggered despite pedestrians intending to cross the street amounted to 7.1%.

Correct operation of the system was also evidenced by the high efficiency in eliminating vehicle triggers; that is, in 99.3% of cases, vehicles did not trigger the system. The rate of incorrect operations for vehicles triggering the active pedestrian crossing was reported to be negligible, at 0.7%.

All unwarranted triggers are unfavorable because with frequent activation of the crossing without pedestrian presence, the effect of the light signals will depreciate as the drivers grow accustomed to activation of the signal without pedestrian presence, and after a certain period of time, they stop paying attention to the warning lights. The system was set up so that it remained in an active state after detecting movement in the area of the curtain sensors for 20 s. When motion is detected once the IR/MW sensors have been triggered, the activation time is further extended by 7 s.

Moreover, if there is a vehicle in the detection field that incorrectly prolongs the time the crossing is activated after being previously triggered by pedestrians, the time will be extended by another 10 s. This situation causes drivers to become accustomed to a triggered active pedestrian crossing in which there are no pedestrians. As a result, even when the crossing system is triggered, motorized road users may not yield the right of way to pedestrians.

During the measurements, the efficiency of the warning signal in relation to the number of vehicles reacting to it and stopping before the pedestrians was also determined. Among all the vehicles observed during the experiment, only 27.8% yielded the right of way to pedestrians and 72.3% did not. The lengthy exposure to the flashing warning lights accustomed the drivers, not only of passenger cars but also city buses, to the situation where no pedestrians are present at the crossing and the lights are still in an active state. Therefore, our findings of the pedestrian–driver relationship shows that yielding of the right of way occurred mainly when the pedestrian was on the refuge island separating the lanes.

5.2. Discussion

The results obtained in this study on the effectiveness of pedestrian crossing detection systems provide valuable insights into their functionality, reliability, and areas requiring further improvement. The analysis highlights several key aspects:

• Pedestrian Detection Effectiveness

One of the most significant findings is the high pedestrian detection accuracy, reaching 91.3% (274 correct activations out of 300 attempts). In the context of road traffic safety systems, this level of precision demonstrates the system’s reliability, which is crucial for ensuring pedestrian safety. However, despite this favorable result, a small number of detection errors were still observed.

• False Activations by Pedestrians

The minimal rate of false activations by pedestrians (5 out of 300 attempts) indicates a high-quality detection mechanism. This result confirms the system’s ability to distinguish between pedestrians intending to cross and those merely in the vicinity of the crossing without crossing intentions.

• Elimination of Vehicle-Related Errors

The system’s performance in eliminating errors associated with vehicles is particularly promising. Out of 1000 detection attempts, only 7 cases (0.7%) involved false activations triggered by vehicles. This low error rate suggests precise system operation, though additional testing under varying traffic conditions is necessary to validate these results further.

• Missed Activations

In 21 out of 300 attempts (7%), the system failed to activate despite the pedestrian showing an intention to cross the road. Such cases represent a significant safety concern, as a lack of system response could expose pedestrians to collision risks. The underlying causes of these missed activations need to be analyzed to enhance the detection algorithm or system configuration, minimizing the risk of non-activation events.

• Driver Behavior in Yielding to Pedestrians

The findings related to driver behavior at active pedestrian crossings reveal significant discrepancies in expected responses. Only 139 out of 500 vehicles (27.8%) yielded to pedestrians, indicating a persistent problem with driver compliance to warning signals. Over 70% of vehicles ignored the active crossing, reflecting low driver awareness or adherence to traffic regulations. These results suggest that while the detection system performs well, addressing driver behavior remains a critical challenge beyond the technical scope of the system.

It is also important to highlight certain study limitations:

• Typical Road Conditions and Context Variability

The findings are based on specific road conditions, which may not fully represent all scenarios. The system’s effectiveness might vary in different cities, depending on infrastructure, traffic volume, or crossing characteristics.

• Sample Size and Variety of Tests

Although the sample size was substantial, it focused on a single location and a specific type of crossing. Expanding this study to include various crossing types, such as those in high-traffic areas or accident-prone locations, could provide a more comprehensive evaluation of the system’s performance.

• Driver Behavior

Despite the high detection accuracy, this study revealed a significant number of instances where vehicles failed to yield to pedestrians. This suggests that pedestrian detection systems alone cannot fully address issues related to driver compliance with traffic regulations. Further research into integrating detection systems with other technologies, such as speed monitoring systems or mechanisms encouraging drivers to yield, may be necessary.

The study results demonstrate the pedestrian crossing detection system’s high effectiveness in identifying pedestrians and vehicles. However, they also reveal areas requiring further observation and refinement, particularly in eliminating detection errors and addressing driver behavior. Integrating detection systems with other technologies could represent a step toward achieving even higher safety levels at pedestrian crossings.

6. Conclusions

The conflict between pedestrian and motorist traffic is a source of risks that requires continuous research and solutions improving the safety and comfort of road users. The implementation of active pedestrian crossings and continuous analysis of the effects of such solutions serve to improve the functioning of the triggered crossings. The efficiency assessment indicates that the crossing is characterized by high pedestrian detection efficiency, with correctly positioned sensor detection fields. In addition, vehicles do not cause unwarranted triggers.

Worrying results were obtained for the pedestrian–driver relation, where despite a busy pedestrian crossing, motorized users did not react to the pedestrians waiting at the crossing point. It is advisable to observe the drivers and their reactions as they near pedestrian crossings and then to take appropriate measures to improve safety. Therefore, the improvement of the situation may be affected by conducting systematic research within the scope presented above.

The most significant quantitative results of the pedestrian crossing efficiency evaluation include the following:

• High pedestrian detection accuracy: Out of 300 system activation attempts, 274 activations demonstrated correct functioning.
• Minimal rate of erroneous activations by pedestrians: The activation of the system by pedestrians not intending to cross was extremely low, occurring in only five instances.
• Effectiveness in eliminating vehicle-related errors: The system showed high effectiveness in preventing activations caused by vehicles, with 993 out of 1000 vehicle interaction attempts not triggering the system. Only seven instances resulted in incorrect activations.
• Rate of missed activations: Among 300 pedestrian activation attempts, 21 instances were observed as incorrect (no activation occurred).
• Large sample size: This study included 1000 vehicle interaction attempts and 300 pedestrian activation attempts, ensuring the representativeness of the data.
• Yielding behavior of vehicles: Among 500 observations, 139 vehicles yielded the right of way to pedestrians at the active crossing.
• Non-yielding behavior of vehicles: Among 500 observations, 361 vehicles failed to yield the right of way to pedestrians at the active crossing.

Given the obtained results, further research should focus on solutions aimed at improving the effectiveness of the active pedestrian crossing system. Although the system demonstrated high efficiency in pedestrian detection and minimizing errors related to vehicle interactions, maintaining its effectiveness in the context of driver habituation to the system remains a challenge. Additionally, ensuring that false activations do not reduce the effectiveness of the warning light signals is of critical importance.

A particularly concerning and noteworthy issue requiring further observation is the low level of priority yielded by drivers to pedestrians. In the examined experiment, only 27.8% of vehicles yielded to pedestrians, highlighting a problem with drivers’ responses to warning signals, particularly after prolonged exposure to their operation.

Future work may be directed towards optimizing the system’s activation timing. While prolonging the activation duration upon detecting movement theoretically enhances safety, it may lead to situations where drivers cease to respond to the warning lights, ultimately reducing the system’s overall effectiveness in ensuring safety.

Reducing the number of false activations remains a significant challenge. Unintended system activations can undermine the effectiveness of the entire system, particularly if the displayed warning signals fail to elicit compliance from drivers.

The low yielding observed could be attributed to several interrelated factors. One possible explanation is insufficient awareness of the system among drivers, which may stem from a lack of exposure to or understanding of how the system operates. Drivers who are unfamiliar with the system may fail to recognize or appropriately respond to the warning signals. Another factor could be potential issues with system calibration, such as misaligned sensors or inaccurate timing of signal activation, which might reduce effectiveness in capturing driver attention. Additionally, the effectiveness of the signals and alerts in influencing driver behavior plays a critical role. The design, visibility, or intensity of these signals may not be sufficiently compelling to prompt drivers to yield consistently.

To address these issues, further field tests should be conducted under varying traffic and environmental conditions to comprehensively evaluate driver responses. Ensuring proper calibration of the system is equally important, as well-aligned sensors and optimally timed signals are critical for effective communication with drivers.

Moreover, incorporating targeted driver education or awareness campaigns could significantly enhance the system’s impact. These campaigns might include informational materials, road signage, or public service announcements that explain the purpose and functionality of the system, encouraging drivers to adapt their behavior accordingly. By combining technical refinements with behavioral interventions, these efforts would collectively improve yielding behavior, ultimately enhancing the overall safety and effectiveness of the system.

The findings of this study hold particular value in identifying the developmental potential of such solutions. They enable further exploration of the feasibility of implementing similar technologies in other locations, especially in areas with a high risk of pedestrian-related traffic accidents. Furthermore, the analysis facilitates the identification of directions for improving the system under study to enable its adaptation to other contexts, taking into account the specific characteristics of different environments and traffic conditions. In this regard, this study provides a foundation for further research and development efforts in innovative road safety solutions.