Nils Lubbe

Serious injury prediction algorithm based on large-scale data and under-triage control

The present study was undertaken to construct an algorithm for an advanced automatic collision notification system based on national traffic accident data compiled by Japanese police. While US research into the development of a serious-injury prediction algorithm is based on a logistic regression algorithm using the National Automotive Sampling System/Crashworthiness Data System, the present injury prediction algorithm was based on comprehensive police data covering all accidents that occurred across Japan. The particular focus of this research is to improve the rescue of injured vehicle occupants in traffic accidents, and the present algorithm assumes the use of an onboard event data recorder data from which risk factors such as pseudo delta-V, vehicle impact location, seatbelt wearing or non-wearing, involvement in a single impact or multiple impact crash and the occupants age can be derived. As a result, a simple and handy algorithm suited for onboard vehicle installation was constructed from a sample of half of the available police data. The other half of the police data was applied to the validation testing of this new algorithm using receiver operating characteristic analysis. An additional validation was conducted using in-depth investigation of accident injuries in collaboration with prospective host emergency care institutes. The validated algorithm, named the TOYOTA-Nihon University algorithm, proved to be as useful as the US URGENCY and other existing algorithms. Furthermore, an under-triage control analysis found that the present algorithm could achieve an under-triage rate of less than 10% by setting a threshold of 8.3%.

Pedestrian crossing situations: Quantification of comfort boundaries to guide intervention timing

INTRODUCTION Technical systems that warn or brake for vehicle-pedestrian encounters reduce injuries more effectively the earlier an intervention is initiated. However, premature intervention can irritate drivers, leading to system deactivation and, consequently, no injury reduction whatsoever. It has been proposed that no intervention should be initiated as long as attentive drivers are within their comfort zones. This study aims at quantifying driver comfort boundaries for pedestrian crossing situations to offer guidance for the appropriate timing of interventions. METHODS Sixty two volunteers drove through an intersection on a test track at 30 and 50km/h. A pedestrian dummy was launched from behind an obstruction towards the driving path of the approaching car. Brake onset indicated discomfort. Time to collision (TTC), longitudinal and lateral distance were measured at brake onset. RESULTS TTC was independent of driving speed ranging from 2.1 to 4.3s with a median of 3.2s. Longitudinal distance ranged from 19 to 48 meters with an apparent difference between driving speeds. Lateral distances differed slightly, but significantly between driving speeds. The median was 3.1m (3.2m for 30km/h and 2.9m for 50km/h) and values ranged from 1.9 to 4.1m. Lateral distance in seconds ranged from 1.9 to 4.3s with a median value of 3.1s (3.2s for 30km/h and 3.0s for 50km/h). DISCUSSION TTC was independent of driving speed, trial order and volunteer age. It might be considered suitable to intervene in situations where, for example, 90% of drivers have exceeded their comfort boundary, i.e. when drivers have already initiated braking. This percentile value translates to intervention at a TTC of 2.5s (95% confidence 2.4-2.7s). The study was limited to Swedish nationals, fully aware drivers, and two driving speeds, but did not investigate behavioural changes due to system interaction. CONCLUSION This study showed that TTC at brake onset was a suitable measure for the quantification of driver comfort boundaries in pedestrian crossing situations. All drivers applied their brakes prior to 2.1s TTC.

Brake reactions of distracted drivers to pedestrian Forward Collision Warning systems

INTRODUCTION Forward Collision Warning (FCW) can be effective in directing driver attention towards a conflict and thereby aid in preventing or mitigating collisions. FCW systems aiming at pedestrian protection have been introduced onto the market, yet an assessment of their safety benefits depends on the accurate modeling of driver reactions when the system is activated. This study contributes by quantifying brake reaction time and brake behavior (deceleration levels and jerk) to compare the effectiveness of an audio-visual warning only, an added haptic brake pulse warning, and an added Head-Up Display in reducing the frequency of collisions with pedestrians. Further, this study provides a detailed data set suited for the design of assessment methods for car-to-pedestrian FCW systems. METHOD Brake response characteristics were measured for heavily distracted drivers who were subjected to a single FCW event in a high-fidelity driving simulator. The drivers maintained a self-regulated speed of 30km/h in an urban area, with gaze direction diverted from the forward roadway by a secondary task. RESULTS Collision rates and brake reaction times differed significantly across FCW settings. Brake pulse warnings resulted in the lowest number of collisions and the shortest brake reaction times (mean 0.8s, SD 0.29s). Brake jerk and deceleration were independent of warning type. Ninety percent of drivers exceeded a maximum deceleration of 3.6m/s2 and a jerk of 5.3m/s3. CONCLUSIONS Brake pulse warning was the most effective FCW interface for preventing collisions. In addition, this study presents the data required for driver modeling for car-to-pedestrian FCW similar to Euro NCAPs 2015 car-to-car FCW assessment. Practical applications: Vehicle manufacturers should consider the introduction of brake pulse warnings to their FCW systems. Euro NCAP could introduce an assessment that quantifies the safety benefits of pedestrian FCW systems and thereby aid the proliferation of effective systems.

Assessment of Integrated Pedestrian Protection Systems with Autonomous Emergency Braking (AEB) and Passive Safety Components

Autonomous emergency braking (AEB) systems fitted to cars for pedestrians have been predicted to offer substantial benefit. On this basis, consumer rating programs—for example, the European New Car Assessment Programme (Euro NCAP)—are developing rating schemes to encourage fitment of these systems. One of the questions that needs to be answered to do this fully is how the assessment of the speed reduction offered by the AEB is integrated with the current assessment of the passive safety for mitigation of pedestrian injury. Ideally, this should be done on a benefit-related basis. The objective of this research was to develop a benefit-based methodology for assessment of integrated pedestrian protection systems with AEB and passive safety components. The method should include weighting procedures to ensure that it represents injury patterns from accident data and replicates an independently estimated benefit of AEB. A methodology has been developed to calculate the expected societal cost of pedestrian injuries, assuming that all pedestrians in the target population (i.e., pedestrians impacted by the front of a passenger car) are impacted by the car being assessed, taking into account the impact speed reduction offered by the cars AEB (if fitted) and the passive safety protection offered by the cars frontal structure. For rating purposes, the cost for the assessed car is normalized by comparing it to the cost calculated for a reference car. The speed reductions measured in AEB tests are used to determine the speed at which each pedestrian in the target population will be impacted. Injury probabilities for each impact are then calculated using the results from Euro NCAP pedestrian impactor tests and injury risk curves. These injury probabilities are converted into cost using “harm”-type costs for the body regions tested. These costs are weighted and summed. Weighting factors were determined using accident data from Germany and Great Britain and an independently estimated AEB benefit. German and Great Britain versions of the methodology are available. The methodology was used to assess cars with good, average, and poor Euro NCAP pedestrian ratings, in combination with a current AEB system. The fitment of a hypothetical A-pillar airbag was also investigated. It was found that the decrease in casualty injury cost achieved by fitting an AEB system was approximately equivalent to that achieved by increasing the passive safety rating from poor to average. Because the assessment was influenced strongly by the level of head protection offered in the scuttle and windscreen area, a hypothetical A-pillar airbag showed high potential to reduce overall casualty cost. A benefit-based methodology for assessment of integrated pedestrian protection systems with AEB has been developed and tested. It uses input from AEB tests and Euro NCAP passive safety tests to give an integrated assessment of the system performance, which includes consideration of effects such as the change in head impact location caused by the impact speed reduction given by the AEB.

Market penetration of intersection AEB: Characterizing avoided and residual straight crossing path accidents

Car occupants account for one third of all junction fatalities in the European Union. Driver warning can reduce intersection accidents by up to 50 percent; adding Autonomous Emergency Braking (AEB) delivers a reduction of up to 70 percent. However, these findings are based on an assumed 100 percent equipment rate, which may take decades to achieve. Our study investigates the relationship between intersection AEB market penetration rates and avoidance of accidents and injuries in order to guide implementation strategies. Additionally, residual accident characteristics (impact configurations and severity) are analyzed to provide a basis for future in-crash protection requirements. We determined which accidents would have been avoided through the use of an Intersection AEB system with different sensor field-of-views (180° and 120°) by means of re-simulating the pre-crash phase of 792 straight crossing path (SCP) car-to-car accidents recorded in the German In-Depth Accident Study (GIDAS) and the associated Pre-Crash Matrix (PCM). Intersection AEB was activated when neither of the conflict opponents could avoid the crash through reasonable braking or steering reactions. For not-avoided accidents, we used the Kudlich-Slibar rigid body impulse model to calculate the change of velocity during the impact as a measure of impact severity and the principal direction of force. Accident avoidance over market penetration is not linear but exponential, with higher gains at low penetration rates and lower gains at higher rates. A wide field-of-view sensor (180°) substantially increased accident avoidance and injury mitigation rates compared to a 120° field-of-view sensor. For a 180° field-of-view sensor at 100 percent market penetration, about 80 percent of the accidents and 90 percent of the MAIS2 + F injuries could be avoided. For the remaining accidents, AEB intervention rarely affected side of impact. The median change of velocity (delta-V) of the remaining crashes reduces only marginally at low penetration rates but this reduction increases with higher penetration rates. With 100 percent market penetration, one quarter of the vehicles still involved in straight crossing path accidents will sustain a delta-V higher than 17 km/h. Intersection AEB is very effective. Enabling a fast initial implementation of systems with wide field-of-view sensor(s) and ensuring a high market penetration over the longer term is essential to achieve high crash avoidance and injury mitigation rates over time. The standards for in-crash protection must be high to mitigate injury in the unavoidable, residual accidents.