Pedestrian Fall Research Articles

Pedestrian Fall Detection Methods for Public Traffic Areas: A Literature Review

Pedestrian Fall Detection Methods for Public Traffic Areas: A Literature Review

An evacuation model considering pedestrian fall behavior in an inclined passenger ship

Passenger ships are likely to be inclined during and after accidents, which is a vital factor that affects pedestrian behaviors and crowd dynamics in the evacuation process. On a significantly inclined ship, pedestrian fall behavior becomes more complex, increasing the risk of stampedes. Nevertheless, this type of evacuation has rarely been investigated. In this study, an extended cellular automata (CA) model that meets the inherent characteristics of passenger ship evacuations is constructed. The inclination force and self-adjusting force are incorporated into the model to evaluate the balance mechanism of pedestrians, while a method to calculate the probability of pedestrians falling down is established by considering the ship heeling angle and pedestrian heterogeneity. Further, the state of the pedestrians after falling to the ground is analyzed. A hazard field is introduced to describe the interaction between normal pedestrians and fallen pedestrians. The simulation results show that deaths occur when the ship heeling angle is greater than 20° and the crowd is evacuated successfully and fastest at the ship heeling angle of 5°, even better than on the flat ground. The greater the ship heeling angle, the more likely pedestrians are to fall over, and the earlier the first death occurs, thereby resulting in longer evacuation time and increased fatalities. The heterogeneity of passengers has considerable effects on the effectiveness of the ship evacuation, with a greater proportion of strong-resistance pedestrians reducing the crowd death and fall risks, even though the influence of crowd constitution on crowd evacuation time is insignificant. The death toll and evacuation time increased with the initial number of pedestrians aboard the ship. This study is beneficial in terms of understanding the evolution mechanism of evacuation dynamics and developing effective strategies for crowd emergency management on large passenger vessels.

Disturbance Propagation Model of Pedestrian Fall Behavior in a Pedestrian Crowd and Elimination Mechanism Analysis

Disturbance Propagation Model of Pedestrian Fall Behavior in a Pedestrian Crowd and Elimination Mechanism Analysis

Pedestrian Fall Detection Algorithm Based on Improved YOLOv8

Pedestrian Fall Detection Algorithm Based on Improved YOLOv8

Finite Element Analysis of Energy-Absorbing Floors for Reducing Head Injury Risk during Fall Accidents

Energy-absorbing floor (EAF) has been proposed as one of several biomechanically effective strategies to mitigate the risk of fall-related injuries by decreasing peak loads and enhancing system energy absorption. This study aims to compare the protective capacity of four commercially available EAF products (Igelkott Floor, Kradal, SmartCells, and OmniSports) in terms of head impacts using the finite element (FE) method. The stress–strain curves acquired from mechanical tests were applied to material models in LS-Dyna. The established FE models were then validated using Hybrid III or hemispheric drop tests to compare the acceleration–time curves between experiments and simulations. Finally, the validated FE models were utilized to simulate a typical pedestrian fall accident scenario. It was demonstrated that EAFs can substantially reduce the peak forces, acceleration, and velocity changes during fall-related head impacts. Specifically, in the accident reconstruction scenario, SmartCells provided the largest reduction in peak linear acceleration and skull fracture risk, while Igelkott Floor provided the largest reduction in peak angular velocity and concussion risk. This performance was caused by different energy absorption mechanisms. Consequently, the results can contribute to supporting the implementation of EAFs and determine the effectiveness of various protective strategies for fall-related head injury prevention.

Fall and collision related injuries among pedestrians, sickness absence and associations with accident type and occupation

Objectives: This study explores pedestrian fall accidents and collisions with other road users in the Swedish road transport system, and sickness absence (SA) in relation to accident type, injury, and occupation. Further, it studies the associations between accident type, occupation, and duration of SA. Methods: Data from several national registers were used that included 15,359 working age pedestrians (20–64 years) receiving healthcare after a fall or collision throughout 2014–2016. Individual characteristics, accident type, injury, and occupation were presented and related to SA. Logistic regression was used to estimate odds ratios (OR), with 95% confidence intervals, for associations between accident type, occupation, and SA duration. Results: About 11,000 pedestrians (72%) were involved in fall accidents in the road traffic environment and well over 4,000 in collisions with another road user; 22% of all injured pedestrians had a new SA. The population had a higher proportion of women and individuals in older age groups (≥45). Of the falls, 31% were due to snow or ice, and these were associated with a higher OR for both short SA (<90 days) 1.76 (95% CI 1.56–1.98) and long SA (≥90 days) 1.81 (95% CI 1.51–2.18), compared to the group slipping, tripping, and stumbling. The working sectors health & social care, and construction had the highest ORs for SA. A higher OR was found for health & social care, short SA 1.58 (95% CI 1.38–1.81), long SA 1.79 (95% CI 1.45–2.20) and for construction, short SA 1.56 (95% CI 1.24–1.96), long SA 1.75 (95% CI 1.26–2.44), compared to the sector finance, communication, & cultural service. Conclusions: The OR for having short and long SA was higher in falls due to snow or ice and differed between occupational sectors. Practical implications: This information contributes to the knowledge base for planning a safe road transport system for pedestrians.

Measurement of traction properties of ceramic tiles and its attention for preventing pedestrian falls

Ceramic tiles are generally located in various community and industry venues because of their durability and ability to withstand blemishes, smells, and microorganisms. Thus, they are ideal for bathrooms, kitchens, entryways, laundry rooms, and utility areas. However, they can be extremely slippery in lubricated and even dusty environments. Regardless of the significance of this issue, the care aspects of ceramic tiles’ traction properties should have been systematically studied to avoid slip and fall incidence. The present study evaluates the slip resistance performance of ceramic tiles with different textures and identifies their controls on pedestrian fall prevention. For the ceramic tiles, the three most popular types of surface textures: smooth, moderately rough, and very rough, were chosen. Dynamic friction tests measured traction properties between the ceramic tiles and three shoes in four environments. The measurement results identified that the ceramic tiles exhibited high traction functioning under clean and dry conditions. Whilst the textures of ceramic tiles played a considerable role in soapsuds-covered moderately risky and oil-covered highly hazardous conditions. The current study also proposed a concept of an operational range of surface roughness and identified ideal boundaries of surface textures against diverse levels of slippery circumstances. The outcomes of the present study may be applied to enhance the surface design of ceramic tiles to preclude fall hazards in our community and workplaces.

Object detection and activity recognition in video surveillance using neural networks

PurposeThis paper aims to implement and extend the You Only Live Once (YOLO) algorithm for detection of objects and activities. The advantage of YOLO is that it only runs a neural network once to detect the objects in an image, which is why it is powerful and fast. Cameras are found at many different crossroads and locations, but video processing of the feed through an object detection algorithm allows determining and tracking what is captured. Video Surveillance has many applications such as Car Tracking and tracking of people related to crime prevention. This paper provides exhaustive comparison between the existing methods and proposed method. Proposed method is found to have highest object detection accuracy.Design/methodology/approachThe goal of this research is to develop a deep learning framework to automate the task of analyzing video footage through object detection in images. This framework processes video feed or image frames from CCTV, webcam or a DroidCam, which allows the camera in a mobile phone to be used as a webcam for a laptop. The object detection algorithm, with its model trained on a large data set of images, is able to load in each image given as an input, process the image and determine the categories of the matching objects that it finds. As a proof of concept, this research demonstrates the algorithm on images of several different objects. This research implements and extends the YOLO algorithm for detection of objects and activities. The advantage of YOLO is that it only runs a neural network once to detect the objects in an image, which is why it is powerful and fast. Cameras are found at many different crossroads and locations, but video processing of the feed through an object detection algorithm allows determining and tracking what is captured. For video surveillance of traffic cameras, this has many applications, such as car tracking and person tracking for crime prevention. In this research, the implemented algorithm with the proposed methodology is compared against several different prior existing methods in literature. The proposed method was found to have the highest object detection accuracy for object detection and activity recognition, better than other existing methods.FindingsThe results indicate that the proposed deep learning–based model can be implemented in real-time for object detection and activity recognition. The added features of car crash detection, fall detection and social distancing detection can be used to implement a real-time video surveillance system that can help save lives and protect people. Such a real-time video surveillance system could be installed at street and traffic cameras and in CCTV systems. When this system would detect a car crash or a fatal human or pedestrian fall with injury, it can be programmed to send automatic messages to the nearest local police, emergency and fire stations. When this system would detect a social distancing violation, it can be programmed to inform the local authorities or sound an alarm with a warning message to alert the public to maintain their distance and avoid spreading their aerosol particles that may cause the spread of viruses, including the COVID-19 virus.Originality/valueThis paper proposes an improved and augmented version of the YOLOv3 model that has been extended to perform activity recognition, such as car crash detection, human fall detection and social distancing detection. The proposed model is based on a deep learning convolutional neural network model used to detect objects in images. The model is trained using the widely used and publicly available Common Objects in Context data set. The proposed model, being an extension of YOLO, can be implemented for real-time object and activity recognition. The proposed model had higher accuracies for both large-scale and all-scale object detection. This proposed model also exceeded all the other previous methods that were compared in extending and augmenting the object detection to activity recognition. The proposed model resulted in the highest accuracy for car crash detection, fall detection and social distancing detection.

Safety Assessment of the Community Facilities for the Prevention of Pedestrian Fall Incidents: Concrete Footpaths and Walkways

Safety Assessment of the Community Facilities for the Prevention of Pedestrian Fall Incidents: Concrete Footpaths and Walkways

Tribology model to characterise shoe friction and wear behaviours and its application to evaluating traction properties

Wear and tear growths of shoes are unavoidable, but this issue has hardly been investigated in the literature from the viewpoint of slip resistance to prevent pedestrian falls. This study suggests a shoe tribology model to explore friction and friction-induced wear developments of shoe heel/sole surfaces and identify their effects on traction controls. To verify the model, dynamic friction tests were performed between two popular shoe kinds: Polyurethane and Polyvinyl Chloride against a vinyl sheet under dry conditions. Their results were investigated quantitatively and qualitatively. Results clearly show that traction properties are not constant and primarily depend on material types. Findings from this study may enhance our understanding of shoe surface friction and wear behaviours and provide helpful information to develop better designs for heels/soles to prevent pedestrian fall incidents.

The association between tree canopy cover over streets and elderly pedestrian falls: A health disparity study in urban areas

Older pedestrians are vulnerable to outdoor falling while walking on streets/sidewalks, but few studies have examined the role of the street environment and tree canopy cover over streets in relation to pedestrian falls among the elderly. We used spatial analysis to examine the association between tree canopy cover over streets and pedestrian falls reported to Emergency Medical Service (EMS) providers from March 2013 to February 2020 among adults aged 65 and older living in urban areas of Marin County, CA. Tree canopy cover over streets was measured using 1-m resolution of tree canopy within street polygons. After controlling for socioeconomic status and built environments, we found an inverse association between tree canopy cover over streets and elderly pedestrian fall rates at the census block level. Specifically, with a 10-percentage point increase in tree canopy cover over streets of a block, we expected to see about an 11.2% decrease in the elderly pedestrian fall rate. We found that the inverse relationship between tree canopy cover over streets and elderly pedestrian falls was only significant during the leaf-on season in the spring and summer. Finally, sub-analysis found that the relationship between tree canopy cover over streets and elderly pedestrian falls was stronger in low-income areas, compared to high-income areas. Planting street trees is a potential evidence-based intervention to prevent pedestrian falls. However, special attention must go beyond the quantity of tree canopy cover over streets to consider biophysical factors and social conditions.

Pedestrian Fall Event Detection in Complex Scenes Based on Attention-Guided Neural Network

To address automatic detection of pedestrian fall events and provide feedback in emergency situations, this paper proposes an attention-guided real-time and robust method for pedestrian detection in complex scenes. First, the YOLOv3 network is used to effectively detect pedestrians in the videos. Then, an improved DeepSort algorithm is used to track by detection. After tracking, the authors extract effective features from the tracked bounding box, use the output of the last convolutional layer, and introduce the attention weight factor into the tracking module for final fall event prediction. Finally, the authors use the sliding window for storing feature maps and SVM classifier to redetect fall events. The experimental results on the CityPersons dataset, Montreal fall dataset, and self-built dataset indicate that this approach has good performance in complex scenes. The pedestrian detection rate is 87.05%, the accuracy of fall event detection reaches 98.55%, and the delay is within 120 ms.

Fall- and collision-related injuries among pedestrians in road traffic environment – A Swedish national register-based study

Objective: To investigate the burden of pedestrian injuries, including pedestrian fall injuries (PFI), compared to other transport-related injuries in Sweden and document their characteristics in terms of demographics, causes, type of injuries, and severity level with a focus on long-term consequences. Methods: Data were retrieved from the national Swedish Traffic Accident Data Acquisition register. A total of 361,531 fatalities and injuries were reported by emergency hospitals during 2010–2019, of which 127,804 were pedestrians (35%). We assessed the magnitude of PFIs and conducted comparative analyses to assess differences compared to other types of road users regarding sex, age, severity level, injury circumstances, hospital care, causes of accidents, and type of injuries. Results: Pedestrians were the second largest group of traffic-related deaths in Sweden after car occupants and accounted for just over a quarter of all fatal accidents in the road traffic environment. Of the total number of pedestrian fatalities, three out of four have been in collision accidents and the others in fall-related accidents. In terms of injuries, pedestrians were the largest group among all road users, regardless of the type of accident. PFIs accounted for a third of all injuries in the road traffic environment and nearly half of all injuries resulting in permanent medical impairment (i.e., 2.2 times more long-term consequences among PFIs compared to injured car occupants). Females (particularly middle-aged and older) and older adults were overrepresented, and most PFIs occurred on urban and municipal roads. The causes were often related to maintenance (e.g., slippery surfaces such as ice, snow, leaves or gravel together with uneven pavements and roads are the cause three out of four of PFIs). Among collision injuries, the representation was almost equal for sex and age. Conclusions: Injuries and fatalities among pedestrians are a considerable issue in the road traffic environment in Sweden. Contrary to other traffic groups, the incidence has not decreased over time, meaning that this issue must be met with specific measures and address the specific risk factors they are associated with. Practical Application: Including fall accidents in the definition of traffic accidents increases the chances of getting better information about the accidents and taking preventive measures.

Methods, upper limit and reason for reducing pedestrian ground contact injury by controlling vehicle braking

In order to access the maximum benefits of controlled vehicle braking in reducing ground related injury, based on a basic simulation test sample, which include 24 simulations with three velocities, four pedestrian sizes and two pedestrian gaits, 10176 simulations were conducted by Madymo. After comparative study, we found that the vehicle caused injury will not increase if the vehicle braking is starting to reduce at the time the pedestrian head first contact the vehicle and the ground related WIC/HIC can be maximally reduced 88.9% and 75.6% respectively by controlling the vehicle braking efficiently. Through a further analysis we found that the optimal time at which the vehicle was to fully brake again (t2) is not hard to obtain because the optimal t2 is not an isolated point and hence it is easy to obtain a simple controlling method for a specific car shape but it may be impossible to find one way to control the braking of vehicles with different shapes. And the possible reason that the ground related injury can be reduced is the change of the landing mechanism. Not all ground related injury can be reduced because in some fully braked cases the WIC/HIC is already equal or close to 0; and two danger situations for controlling the vehicle braking were found, which are the pedestrian fall to the ground from the vehicle side and the controlled vehicle cannot contact the pedestrian’s main body in the controlling braking process. Research results will enhance our confidence.

Tribological approach for the safety assessment of flooring/walkway surfaces: Application for the prevention of pedestrian fall incidence

This study investigated the role of surface finishes from different material types of floors/walkways on traction properties and identified their effects on slipperiness controls under diverse walking environments to prevent fall incidence. Measurements and analyses of surface textures and traction properties of floors/walkways and interpretations were based on a tribological concept. Dynamic friction tests were conducted amongst nine indoor and outdoor floor specimens and three shoes under four different risk levels of environments. Test results showed that traction performances were considerably influenced by surface finishes of the floor specimens under the moderately (soapy) and highly (oily) risky conditions. This study also uncovered that rough floors did not mechanically ensure higher slip resistance whilst some smooth floors exhibited better slip resistance than the rougher ones. Nevertheless, outcomes from this study provided fundamental, but important information on the walkway traction properties that showed multifaceted features of tribological phenomena. Based on the overall findings, this study proposed an insight to identify the impact of floor/walkway textures on traction performances and consequently exploit sensible design concepts for the floor surfaces to prevent fall incidents.

Understanding Friction and Wear Behaviours of Smooth Resilient Surfaces: Application for Pedestrian Fall Safety Improvements

The recent literature identified that certain types of smooth floors and flooring materials showed good slip‐resistance performance under moderately contaminated situations such as water wet and soapsuds‐covered conditions. However, topographic structures of such flat floor/walkway surfaces are significantly changed by wear and tear developments with continuous ambulation. A number of reasons seem to be involved in flooring wear developments, but such changes may result in substantial losses of slip resistance (or traction) functions. Even though the importance of this issue is wide, it is scarce to find any systematic investigations on tribophysical characteristics of smooth floors (&lt;10 µm in Ra roughness) and their wear impacts on traction performance. Moreover, it is difficult to find studies on wear progress of floors and its impacts on pedestrian fall safety assessments. This study focuses on exploring wear activities of smooth resilient floors and disclosing wear consequences on traction properties. To measure slip‐resistance features and investigate wear advances of smooth resilient floors, dynamic friction tests were carried out amongst purposely arranged 4 flat specimens (polymethyl methacrylate: PMMA) and 3 shoes. Wear formations and progressions of the PMMA surfaces were methodically evaluated by observing surface profiles during the tests. The test outcomes evidently presented that initial surface features of the PMMA specimens were significantly modified by primary‐, secondary‐, and shared‐wear mechanisms and significantly influenced traction performance. This study may have design applications for the safety enhancements of smooth floors/footpaths to preclude pedestrian fall incidents.

Watch Out! Or do we? A Review of Literature on Human Locomotion, Visual Gaze, and Expectation of Trip Hazards

Pedestrian trip and fall events, a major source of accidental injury and death, are often the result of a vertical perturbation that goes undetected by a pedestrian. Inattention is frequently blamed as the underlying root cause of the fall. However, research suggests that a pedestrian’s visual gaze is typically focused out towards the horizon and not towards the ground in the immediate vicinity of their feet, leaving them with poor visual acuity to detect subtle details on the walking surface. To the extent a vertical perturbation is unexpected, the likelihood of a trip event is increased. Relevant research on gait cycles, toe clearance, visual gaze, and expectation is examined to determine why pedestrians may fail to detect walkway hazards that are physically possible to see. Synthesis and future research are also discussed.

Investigation of Floor Surface Finishes for Optimal Slip Resistance Performance.

BackgroundIncreasing the slip resistance of floor surfaces would be desirable, but there is a lack of evidence on whether traction properties are linearly correlated with the topographic features of the floor surfaces or what scales of surface roughness are required to effectively control the slipperiness of floors.ObjectiveThis study expands on earlier findings on the effects of floor surface finishes against slip resistance performance and determines the operative ranges of floor surface roughness for optimal slip resistance controls under different risk levels of walking environments.MethodsDynamic friction tests were conducted among three shoes and nine floor specimens under wet and oily environments and compared with a soapy environment.ResultsThe test results showed the significant effects of floor surface roughness on slip resistance performance against all the lubricated environments. Compared with the floor-type effect, the shoe-type effect on slip resistance performance was insignificant against the highly polluted environments. The study outcomes also indicated that the oily environment required rougher surface finishes than the wet and soapy ones in their lower boundary ranges of floor surface roughness.ConclusionThe results of this study with previous findings confirm that floor surface finishes require different levels of surface coarseness for different types of environmental conditions to effectively manage slippery walking environments. Collected data on operative ranges of floor surface roughness seem to be a valuable tool to develop practical design information and standards for floor surface finishes to efficiently prevent pedestrian fall incidents.

Public health consequences of road traffic injuries – Estimation of seriously injured persons based on risk for permanent medical impairment

Despite the fact, that Sweden has one of the safest traffic environments in the world, a large number of people are still injured in road traffic accidents in Sweden. The core concept of the Vision Zero that was adopted by the Swedish parliament in 1997 is to decrease the number of deaths and serious injuries caused by traffic. The vision has been followed-up in terms of decreased mortality, but there was not a consensus on how the concept “seriously injured” should be measured at that time. The aim of this paper is to describe how to develop a measure to estimate the number of seriously injured people in Sweden. The results show that it is possible to estimate the health impact of road traffic accidents based on the definition of medical impairment. According to the results, 8389 people were seriously injured (permanently medical impaired more then 1%) in the road transport area in Sweden in 2013. If the number of people who had slipped or fallen down without any vehicle being involved is excluded, the number of seriously injured was around 4700. The result of the study has been included in the road safety goal for Sweden with the aim to have 25% less seriously injured people in 2020 than the average for the years 2006–2008. The results also showed that 91% of the seriously injured were transported by car, bike or by foot. The relatively high proportion of pedestrian fall accidents happening within the transport system irrespective of whether a vehicle was moving or not raises the question whether these types of accidents should be included in the reporting of formal road traffic accidents.

Development of Economical Vehicle Model for Pedestrian-Friendly Front-End Profile Study

An economical and deformable, hybrid model is developed for studying the effect of vehicle geometry on pedestrian fall kinematics and associated head injury. A simplified structure consisting of finite element surfaces and a Multi-body windshield is built using a series of iterative and no-iterative steps. The primary focus is not the stiffness characteristics of the structure but rather the fall pattern and kinematics data of the pedestrian due solely to the vehicle front-end shape.Comprehensive validation is carried out wherbey the fidelity of the model is reviewed for pedestrian crash kinematics and injury criteria as well as piecewise vehicle parts impact tests. The model is shown to hold up acceptably well against benchmarked values especially for the former, whereby very close head injury criteria values are obtained at identical impact locations. The model's notable features are its economical computational processing time and eases of modification.