Documentation:FIB book/Lateral Pedestrian-Vehicle Collision and Lower Limb Injuries

From UBC Wiki


Pedestrian safety is one of the biggest concerns related to public health and well being. Currently, vehicles are being modified and tested repeatedly for occupant safety but relatively little is being done to improve pedestrian safety. Research shows that lateral impacts account for 74% of all pedestrian-vehicle collisions.[1] When pedestrians are laterally impacted by a vehicle, the most common injuries they sustain are lower limb and head injuries.[1] This page explores the factors contributing to the severity of lower limb and head injuries due to lateral collisions, the development of pedestrian anthropomorphic test devices (ATDs), existing and developing pedestrian safety systems, and the future research and development priorities to improve pedestrian safety.


In 2017 alone, 5,977 pedestrians were killed in lateral vehicle collisions in the United States.[2] On average, a pedestrian was killed every 88 minutes in these crashes.[2] That is about 115 people every week, or more than 16 people per day.[2] Extensive research is being done to better protect pedestrians from vehicles, such as changing regulations, improving vehicle safety systems, or building better roads. This paper will be discussing and analyzing the effects of vehicle safety systems during lateral vehicle collisions.

Lower-Limb Injury

Image of a tibia fracture

In vehicle-pedestrian collisions, the lower limbs are typically the first body part to be directly impacted by the front-end of the vehicle, and therefore, are one of the most vulnerable parts of the human body during an accident.[1] Leg injuries from the bumper result in 38% of all non-fatal pedestrian accidents.[3] These accidents often cause ankle dislocations, fractures in the femur, tibia, and fibula, as well as ligament avulsion and condyle fractures in the knee joints.[1] Though these injuries might not be recognized as life-threatening (have lower AIS levels), they can cause other serious impairments that can significantly reduce quality of life.

Research shows that vehicle bumper height and vehicle speed are the two critical factors involved in determining the type and severity of lower extremity injury.[3] For example, an injury to the tibia and knee ligament could become an injury to the femur with an increase in bumper height. Further, at a speed of 20 - 30 km/hour, the main injury is to the knee ligament and at 40 km/hour is a fracture to the lower-extremities.[3]

Table 1. Impact of Vehicle Speed on Injury
Vehicle Speed 20-30km/hr 40 km/hr
Resultant injuries Knee Ligament Injuries Femur and Tibia fractures

Active muscles also play a significant role in pedestrian-vehicle lateral collisions. For example, the force experienced by the knee during the cadaveric condition is about 0.79 kN whereas during the reflex condition (using active muscles) is 3.1 kN. The FE simulations performed by the Indian Institute of Technology show that active muscles pull the tibia closer to the femur during impact, increasing the stiffness and stresses in the knee and bones, thereby resulting in fractures.[2] Additionally, the tight compression of the knee joint due to the active muscles results in strain in the knee ligaments. Though active muscles are hard to reproduce for cadavers and FE models, they are critical components in determining the type and seriousness of injury and must be taken into account.[2]

Head Injury

Head injury is another common injury type for pedestrian-vehicle collisions. The upper torso and the head usually follows the initial contact of the lower limbs with the vehicle bumper.[4] Although the head injury does not occur as frequently as the lower limb injuries, it is important to take head injuries into consideration as they are often associated with higher AIS scores. Head injuries account for about 31% of AIS 2+ injuries.[4]

It is imperative that attention is given to improving the safety of pedestrians on the road. An improvement in the vehicle safety system could effectively reduce the number of pedestrian deaths and injuries. Therefore, this paper will analyze the data that is currently used to build these safety systems and will provide appropriate recommendations to further improve them.

Development of Pedestrian Anthropomorphic Test Devices


Thor 50th male ATD


The POLAR I is a pedestrian crash test dummy developed by Honda Motor Company, LTD.© (Honda) in 1998, based on an advanced frontal crash test dummy, Thor© (Test Device for Human Occupant Restraint).[5] The thorax, spine, and knee of the Thor dummy were modified based on post mortem human surrogates (PMHS) testing. A computer simulation model was developed using datasets from the Euro-SID ATD, and used to determine joint characteristics. The Thor neck, which is less stiff than the Hybrid III neck, was kept because it met kinematic corridors for frontal and lateral flexion.[5] The Thor ribcage was also left unmodified, because durability and stiffness were appropriate for representing human-like stiffness when impacted.[5] The spine was made more compliant in lateral flexion by replacing two cables in the lumbar spine with a single cable in the centre. The lumbar spine geometry was also changed from a rectangular to a circular section, and the original urethane material was replaced with neoprene. The three-segment abdomen that kept the Thor ATD in an erect posture was changed to a two-segment abdomen for POLAR I.[5] The Hybrid III femur was used, but the Hybrid III sliding knee was simplified to only have rotational freedom. The tibia was modified to fit the new knee, and the lower leg skin and flesh was simulated with Confor foam to represent the stiffness and damping. The lower arms were removed in the POLAR I ATD and replaced with a mass on the Hybrid III upper arm stub. As with the Thor ATD, instrumentation includes tri-axial accelerometers at the centre of gravity for the head, thorax, and pelvis. From testing, the body trajectory and head velocity were almost within the biofidelity corridors at 40 km/hr impacts, but were not biofidelic at lower impact speeds.[5]


In 2000, two years after the development of the first POLAR pedestrian ATD, Honda announced its second generation of the pedestrian dummy.[6] The development of the 2nd generation model was focused towards the knee joint, where the most common injuries occur during vehicle-pedestrian collisions. The cartilage and the ligaments in the knee joints were improved to better represent that of humans and instruments were added to measure the injury levels in eight body regions including the neck and the leg regions.[6]


The third generation of the POLAR ATD was announced in 2008.[7] The previous generations of the ATD were designed with the focus on pedestrian injuries from collisions with smaller vehicles such as sedans. With the increasing popularity of larger vehicles such as the SUVs, Honda developed a newer version of the POLAR pedestrian ATD. Biofidelity of the upper leg and the lumbar area of the ATD was improved to better mimic the pedestrian kinematics for collisions with taller vehicles.[7]

Hybrid III Pedestrian (HIII PED)

The Hybrid III male 50th percentile automotive crash test dummies, which are used to evaluate automotive safety restraint systems in frontal crash testing, were modified to develop the pedestrian version of Hybrid III 50th Male Pedestrian ATD (HIII-50M PED).[8] The lumbar spine was made straight instead of slouching, the legs were modified to allow for vertical posture (instead of sitting), and the knee sliders were replaced with blocks to prevent sliding and only allow rotation. These changes allow the HIII-50M PED ATD to represent a standing position. As with the original HIII ATD, forces and moments can be measured in the upper and lower neck, thorax, lumbar, femur, and tibia, and displacement can be measured in the thorax.[8]

The modifications listed above were also applied to other sizes of HIII ATDs to make the corresponding pedestrian versions. Hybrid III 5th female and Hybrid III 95th male automotive crash test dummies were adapted to develop the Hybrid III 5th Female Pedestrian (HIII-5F PED) and Hybrid III 95th Male Pedestrian (HIII-95M PED), respectively.[9][10]

Flexible Pedestrian Legform Impactor (Flex-PLI)

The Flex-PLI represents the 50th percentile male leg and is used to assess pedestrian lower leg and knee injuries.[11] During testing, the impactor is fired from a linear guide at 40 km/hr into the bumper of a static vehicle. A segmented assembly consisting of the tibia and femur of this legform are made of high-strength plastic, with a flexible knee joint. Springs and stainless steel wires simulate ligaments, and rubber and Neoprene foam sheets represent the flesh. Strain gauges bonded to the bones measure bending moments. Load cells are located in the femur and tibia, and displacement sensors are located in the ligaments of the knee.[11] Optional instrumentation includes load cells in the upper knee, and angular rate sensors in the upper and lower knee.

Advanced Pedestrian Legform Impactor (aPLI)

The aPLI is the successor to the Flex-PLI, also representing the 50th percentile male leg, but includes an upper body mass (UBM) that simulates the mass of the pelvis and torso.[12][13] Biofidelity was improved by redesigning the knee and rerouting ligament cables. Polyurethane flesh was used to make the mass distribution of the flesh and bone more accurate as well. The aPLI allows for more instrumentation than the Flex-PLI, as it includes load cells in more positions along the femur and tibia, and has improved angular rate, displacement, and acceleration measurement.[12]

Pedestrian Headforms

Pedestrian headforms are partial spheres used to measure head accelerations of a pedestrian during pedestrian-vehicle impacts.[14] A robotic arm with a magnetic or mechanical attachment and release launches the pedestrian headforms at car windshields and hoods. Head injury criteria (HIC) can be calculated from these tests. The partial sphere and flat cap of the headforms are made of aluminum and covered with a flexible elastomer, PVC, to represent skin. Triaxial accelerometers are mounted in the geometric centre of the sphere. Pedestrian headforms come in several sizes and masses to represent child and adult heads.[14]

The development of pedestrian ATDs has shown progress in biofidelity and inclusivity, but there is still room for improvement. These will further discussed in the Future Work section below.

Pedestrian Safety Systems

Bumper Design

In terms of pedestrian-vehicle lateral collision, it is the bumper of the vehicle that comes into contact with the pedestrian during the collision. Decreasing the stiffness of the vehicle bumpers will decrease the injury severity of the pedestrians in case of collisions.[15] However, the vehicle crashworthiness will be sacrificed if not enough energy is absorbed by the bumper or the survival space for the vehicle occupants is not ensured during frontal collisions.

Bumper of vehicle

Bumper Material Analysis:

A research team from the Dailan University of Technology have investigated optimizing the design of the bumper to capture both the pedestrian and occupant safety.[15]  The research team has investigated the performances of 3 types of bumpers: aluminum, aluminum-hybrid hybrid, and steel bumpers. The hybrid bumper consisted of layers of aluminum alloy and steel alloy joined by 38 steel rivets.  

The hybrid bumper has shown excellent performance in both the pedestrian and occupant safety. Table 2 shows the performance of each type of bumper. As can be seen from the table, the hybrid bumper has performed extremely well in terms of pedestrian protection, almost resembling the aluminum bumper’s performance. Also, in terms of occupant safety, the hybrid bumper has also performed well. The aluminum bumper was able to only absorb half of the energy that a steel bumper could absorb whereas the hybrid bumper absorbed about 80% of that of the steel bumper. This research has shown that it is possible to improve the pedestrian safety while not causing more harm to the occupants of the vehicle.[15]

Table 2. Performance of Each Type of Bumper[15]
Bumper Type Knee bending angle (deg) Knee shear displacement (mm) Tibia acceleration (g) Energy absorbed (kJ)
Aluminum 1.58 2.11 67.64 3.32
Hybrid 3.45 1.84 69.56 5.22
Steel 12.26 4.29 142.92 6.62

Safer Bumper Design:

To improve pedestrian safety, the redesign of the bumper and lower front of the vehicle can also be considered to further reduce lower leg injuries. The paper by Schuster et al., designs a safe bumper system to reduce pedestrian injuries.[16] The bumper system of a vehicle has a large influence on the vehicle and pedestrian impact performance since it comes into direct contact with the pedestrian’s lower limbs. Previous studies which recommend safer bumper designs have included lower bumper height to the ground, a lower stiffener as an alternative to lower bumper height, and a hollow bumper system to reduce tibia acceleration. Although all these alterations were shown to produce poor performance when used in the ECE-42 (low-speed damageability) test, the hollow bumper system was used for this study. To assess a pedestrian leg impact performance, a leg-form impactor is used. This leg-form impactor has two steel tubular structures joined by a knee joint allowing lateral knee bending and lateral knee shear, and is covered in flesh foam and Neoprene skin. While this study initially prototyped the lef-form with ligaments following the requirements set by Institut National de Recherche sur les Transports et Leur Sécurité (National Institute for Transport and Safety Research, France) (INRETS) with metal non-linear ligaments, the design was unstable when subjected to bend. As a result, this study used a simplified leg-form that has the same mass distribution and bending characteristics as a standard European Experimental Vehicles Committee (EEVC) test procedure without a shearing mechanism. Finite Element Analysis and physical tests using a Bendix impactor are used to test the leg impactor and the vehicle. From this study, Schuster et al. concluded that the bumper height, location, and stiffness of the lower stiffener are the most significant factors to reduce pedestrian lower limb injuries, while the bumper foam stiffness is not significant.[16]

Open hood of car

Hood Design

The first contact point during the vehicle-pedestrian collision is usually the lower extremities and the vehicle bumper, then follows the upper torso, head contact with the hood. Honda R&D department have designed a “pop-up” hood in an attempt to reduce the pedestrian injury for the head-hood collision during the pedestrian lateral collision.[17] The pop-up hood is a relatively simple mechanism which uses a pyro-actuator to lift up the rear side of the hood when a collision sensor detects a pedestrian collision. A POLAR pedestrian dummy was used to evaluate the safety system at 40 km/h vehicle impact speed. There were no noticeable differences for the head-hood impact speed, however the HIC value decreased by approximately 30% when the pop-up hood was activated. This is due to the increased space between the hood and the harder components of the engine bay which acted as a crumple zone.

The redesign of the hood of the car can also be considered to improve pedestrian safety. In the paper by Kim et al., finite element analysis is used as a tool to understand optimal composite hood designs to mitigate impact of pedestrian head injuries.[18] As test procedures have been used to evaluate pedestrian protection features for new cars, FEA was found to be an effective tool to evaluate the Head Injury Criterion for quantitative analysis and to investigate materials and optimal designs to improve protection for pedestrians. Kim et al. used three finite element models for the vehicle, the headform impactor and the single hood to form the entire composite hood based on impact.[18] To simulate the impact, the headform impactor collided with the upper part of the vehicle hood, and the HIC was measured during the time of impact. This study used the ISO child headform impactor model and is composed of a backplate, sphere, and vinyl skin.[18] The composite hoods use composite fiber-reinforced polymers (CFRP) and the hybrid CFRP/GFRP laminates. The hybrid CFRP/GFRP laminate used CFRP layers on the outer layers to maximize bending stiffness of the design since GFRP has a low elastic modulus.[18]

The design objective of the composite hood is to increase the amount of strain energy for the structure when it is undergoing deformation. This means that the higher the strain energy absorbed in the hood during impact, the more head injury can be prevented since less energy would be transferred to the head. The first design maximizes the strain energy by making the composite hood softer to increase pedestrian safety upon impact. The second design minimizes the strain energy of the hood to make the hood stiffer for structural safety. It was found that an aluminum hood showed the highest HIC while the CFRP hood had the lowest HIC.[18] As the results from the CFRP and the CFRP/GFRP hood types were found to be similar, these two designs were found to be better and safer alternatives to steel and aluminum hoods for pedestrian impacts. It was also found that the alternative designs showed good weight reduction which is ideal for reduced fuel consumption. In terms of manufacturing cost, the CFRP/GFRP type hoods are preferred over the CFRP hoods. Finally, in terms of reduced impact force and pedestrian safety, the CFRP hoods have superior performance.

Drawbacks and Limitations

There are several limitations from previously conducted studies which greatly influence the implementation of new vehicle designs to improve pedestrian safety. First, many of the current studies only used isolated body parts when studying pedestrian-vehicle impact. This is not a good representation of the human body, since the mass of the body will affect the type and severity of injury. For example, the paper by Schuster et al fails to consider whole body parts when performing impact tests using the lower stiffener.[16] As the stiffener may increase the likelihood of injury to the tibia, fibula, or ankle joint,  these body parts should be included in the study. A similar scenario can also be observed with the pelvis and head impacts since they were not considered in this study. Similarly, the paper, “Improving pedestrian safety via the optimization of the composite hood structures for automobiles based on the equivalent static load method” by Kim et al. presented an analysis of alternative hood designs using only a headform impactor.[18] Other parts of the body such as the tibia, fibula, and pelvis should be considered in addition to the impact of the headform to improve biofidelity of the model.

Some of the other limitations of current studies include the oversimplification of parameters used in the simulated models, the lack of complete testing to represent real-world impact speeds, and the lack of a clear manufacturing plan usable to scale up vehicle designs. Schuster et al only performed tests using low-speed impacts and as a result, the study fails to account for the impact of vehicles at high speeds. Additionally, an increase in the overall vehicle length due to the addition of a lower stiffener will force manufacturing plants to change how they currently operate. Similarly, the study by Kim et al. presented alternative hood designs which can be used to improve pedestrian safety.[18] As the study used FEA, the study utilized parameters that are not fully representative of actual hood designs. The hood design was simplified to six layers of materials with optimized stacking angles and may be difficult to replicate during the manufacturing process since traditional manufacturing facilities are not equipped with precise enough tools to manufacture layered composite hood structures. More time and resources are still required for manufacturing companies to see if there is a financial benefit in implementing new designs and to transform their current manufacturing facilities to be compatible with the new vehicle redesigns.

Other limitations regarding the incorporation of autonomous systems into vehicles to prevent pedestrian safety includes faulty detection of pedestrians, inaccurate collision prediction, and activation of these autonomous systems.[19]A study conducted by Hamdane et al. explored the use of sensors to trigger an Autonomous Emergency Braking (AEB) used as an active safety system to reduce pedestrian injuries due to a collision.[19] It was found that the field of view angle of the sensor model is critical to improve pedestrian visibility. Even at 35 degrees, the best field of view recommended by this study, only 60% of pedestrians are visible at a time of 2.5 s before impact.[19] Additionally, it was found that 50% of accidents could have been avoided if the AEB was triggered 1s before the impact. By analyzing the visibility of pedestrians by the sensor and the moment when the brakes are applied to avoid the crash, the paper highlighted the challenges associated with pedestrian visibility and the delayed deployment of the AEB. Hamdane et al. addresses the many challenges involved with incorporating new autonomous systems into vehicles and show that many of these systems still need to consider how the pedestrian is moving, where the pedestrian is located, and the accuracy of the embedded sensors.[19]

Future Work

Current studies aimed to improve pedestrian safety can consider full-body injury analysis and a feasible manufacturing plan usable by car manufacturers. For example, the paper by Schuster et al. presented an alternative design to bumpers to reduce lower limb injuries from pedestrian-vehicle impact.[16] Future work around the lower stiffener could include planning around how current manufacturing plants are able to incorporate the addition of the lower stiffener to their cars, how they are able to ship them since the overall length of the car is increased with the addition of the stiffener, and how to design a stiffener for the entire width of the vehicle. Additionally, further studies could investigate how the addition of the lower stiffener affects the tibia, fibula, ankle joint, pelvis, and head impacts. Similarly, further work from the study by Kim et al. could be done to test if the bumper and the lower stiffener can be redesigned using composite materials.[18] Additionally, more testing should be done using full-body impact tests to consider injuries other than the head.

Highly Automated Vehicles (HAVs)

The development of Highly Automated Vehicles (HAVs) aims to eliminate the human error of drivers and improve pedestrian safety.[20] Typically, pedestrians make eye contact with the driver or look towards the approaching vehicle prior to crossing the street. This interaction does not exist with driverless vehicles, so HAVs would need an alternative to assessing the pedestrians’ intentions. In Finland, Utriainen and Pöllänen performed a qualitative study on the potential of HAVs to prevent fatal pedestrian crashes when programmed to prioritize pedestrian safety compared to prioritizing efficient traffic flow. If an HAV decelerated or stopped every time it detected a pedestrian, it would likely avoid pedestrian collisions but would disrupt traffic flow or travel times, and increase the risk of rear-end collisions. If efficient traffic flow was prioritized, the HAV would only decelerate if a pedestrian were in the immediate collision course, which may not leave enough time to brake.


The study looked at 40 well-investigated, fatal pedestrian crashes in Finland from 2014 to 2016 and evaluated the likelihood of crash avoidance for the two conditions if HAVs had been used. The study implemented time-to-collision (TTC) analysis to determine the risk of collision and determined that a TTC greater than 1.5 s is sufficient to avoid collisions with pedestrians, either by stopping or evasive maneuvering. Crashes at pedestrian crossings related to driving with excessive speed, a driver failing to recognize pedestrians, the driver assuming the pedestrian would yield, and similar situations are likely preventable by HAV prioritizing pedestrian safety. For HAV prioritizing traffic flow, it may not be able to stop in time in cases where the driver did not recognize the pedestrian. Outside of crossings, such as standing in the road, jaywalking, or approaching oncoming traffic in the same lane, HAVs would have more difficulty preparing for all possible conflicts, since the expectation is that pedestrians obey traffic yielding rules. If HAVs stopped for pedestrians in all situations, pedestrians may become over-reliant on HAVs stopping even in roadways, and take more risks. In other crashes involving rear-end collisions, vehicles drifting out of the lane, or in a parking area, pedestrian fatality is likely preventable unless the pedestrian was already under the car when it started to reverse, or if the driver or pedestrian exhibited suicidal behavior. More research is required to determine how to strike a balance between pedestrian safety, flow of traffic, and other considerations for transport systems.

ATD and FEA Inclusivity

ATDs and FEA models need to be improved to consider different pedestrian ages, races, sex, and mass. As current pedestrian ATDs are designed to represent a 50th percentile male and a 5th percentile female, these designs group pedestrian responses across different ages, races, and sexes. Grouping pedestrians into a single ATD can reduce the accuracy of responses across categories. An example is the child dummy which is generally a scaled down version of a 50th percentile male dummy and therefore does not result in accurate testing. To make the design of ATDs more inclusive, pedestrian ATDs with different ages, races, and sexes should be developed and tested.


  1. 1.0 1.1 1.2 1.3 Masson, Serre, Cesari, C,T, D. "Pedestrian-vehicle accident: Analysis of 4 full scale tests with PMHS".CS1 maint: multiple names: authors list (link) Cite error: Invalid <ref> tag; name ":3" defined multiple times with different content
  2. 2.0 2.1 2.2 2.3 2.4 "TRAFC Safety Acts - Transportation". NHTSA’s National Center for Statistics and Analysis. April, 2019. Check date values in: |date= (help) Cite error: Invalid <ref> tag; name ":4" defined multiple times with different content
  3. 3.0 3.1 3.2 Matsui, Y (2005, April 7). "Effects of vehicle bumper height and impact velocity on type of lower extremity injury in vehicle–pedestrian accidents". Accident Analysis & Prevention. Check date values in: |date= (help)
  4. 4.0 4.1 Peng, Deck, Yang, Otte, Willinger, Yong, Caroline, Jikuang, Dietmar, Remy. "A study of kinematics of adult pedestrian and head impact conditions in case of passenger car collisions based on real world accident data" (PDF). IRC. 12: 82. line feed character in |title= at position 94 (help)CS1 maint: multiple names: authors list (link)
  5. 5.0 5.1 5.2 5.3 5.4 Akiyama, Akihiko; Yoshida, Suguru; Matsuhashi, Tetsuo; Rangarajan, N (03/01/1999). "Development of Simulation Model and Pedestrian Dummy". SAE Technical Paper – via SAE International. Check date values in: |date= (help)
  6. 6.0 6.1 Honda Motor Co., Ltd. (May 10, 2000). "Honda Develops a 2nd Generation Crash Test Pedestrian Dummy With Injury-Level Measurement Capability".
  7. 7.0 7.1 "Honda to Begin Using Third Generation Pedestrian Dummy in Crash Tests to Enable Evaluation of Injuries to the Lower Back and Upper Leg". Honda News Room. September 18, 2008.
  8. 8.0 8.1 "HIII-50M PED". Humanetics.
  9. "HIII-5F PED". Humanetics.
  10. "HIII-95M PED". Humanetics.
  11. 11.0 11.1 "Flex-PLI Legform". Humanetics.
  12. 12.0 12.1 Stielau, Katharina; Blauth, David (June 10–11, 2020). "Advanced Pedestrian Legform Impactor (aPLI)" (PDF). 16th International LS-DYNA® Users Conference.CS1 maint: date format (link)
  13. "aPLI Legform". Humanetics.
  14. 14.0 14.1 "Pedestrian Headforms". Humanetics.
  15. 15.0 15.1 15.2 15.3 Qi, C., Yang, S. (April 2020). "Multi-Objective Optimization Design of Hybrid Material Bumper for Pedestrian Protection and Crashworthiness Design". WCX SAE World Congress Experience.CS1 maint: multiple names: authors list (link)
  16. 16.0 16.1 16.2 16.3 Schuster, P., Staines, B. (February 23, 1998). "Determination of Bumper Styling and Engineering Parameters to Reduce Pedestrian Leg Injuries". SAE International in United States.CS1 maint: multiple names: authors list (link)
  17. Nagatomi, K., Hanayama, K., Ishizaki, T., Sasaki, S., Matsuda, K. (2005). "DEVELOPMENT AND FULL-SCALE DUMMY TESTS OF A POP-UP HOOD SYSTEM FOR PEDESTRIAN PROTECTION". Honda R&D Co., Ltd. line feed character in |title= at position 56 (help)CS1 maint: multiple names: authors list (link)
  18. 18.0 18.1 18.2 18.3 18.4 18.5 18.6 18.7 Kim, Dong-Hyun (2017). "Improving pedestrian safety via the optimization of composite hood structures for automobiles based on the equivalent static load method". Composite Structures. 176: 780–789.
  19. 19.0 19.1 19.2 19.3 Hamdane, Hedi. "Issues and challenges for pedestrian active safety systems based on real world accidents". HAL.
  20. Utriainen, R., Pöllänen, M. (April 15, 2020). "Prioritizing Safety or Traffic Flow? Qualitative Study on Highly Automated Vehicles' Potential to Prevent Pedestrian Crashes with Two Different Ambitions" (PDF). MDPI. line feed character in |title= at position 55 (help)CS1 maint: multiple names: authors list (link)