Filtern
Dokumenttyp
Sprache
- Englisch (25) (entfernen)
Schlagworte
- Efficiency (25) (entfernen)
Institut
A methodology to derive precision requirements for automatic emergency braking (AEB) test procedures
(2015)
AEB Systems are becoming important to increase traffic safety. Test procedures in testing for consumer information, manufacturer self-certification and technical regulations are used to ensure a certain minimum performance of these systems. Consequently, test robustness, test efficiency and finally test cost become increasingly important. The key driver for testing effort and test costs is the required repeatable accuracy in a test design - the higher the accuracy, the higher effort and test costs. On the other hand, the performance of active safety systems depends on time discretization in the environment perception and other sub-systems: for instance, typical sensors supply information with a cycle time of 50 - 150 ms. Time discretization results in an inherent spread of system performance, even if the test conditions are perfectly equal. The proposed paper shows a methodology to derive requirements for a test setup (e.g. test repeats, use of driving robots, ...) as function of AEB system generation and rating method (e.g. Euro NCAP points awarded, pass/fail, ...). While the methodology itself is applicable to AEB pedestrian and AEB Car-Car scenarios, due to the lack of sufficient test data for AEB Car-Car, the focus of this paper is on AEB pedestrian scenarios. A simulation model for the performance of AEB Pedestrian systems allows for the systematic variation of the discretization time as well as test condition accuracy. This model is calibrated with test results of 4 production vehicles for AEB Pedestrian, all fully tested by BASt according to current Euro NCAP test protocols. Selected parameters to observe the accuracy of the test setup in case of pedestrian AEB is the calculated impact position of pedestrian on the vehicle front (as if no braking would have occurred), and the test vehicle speed accuracy. These variable was shown in real tests to be repeatable in the range of ± 5 cm and ± 0,25 km/h, respectively, with a fully robotized state of the art test setup. The sensitivity of AEB performance (measured in achieved speed reduction as well as overall rating result according to current Euro NCAP rating methods) towards discretization and the sensitivity of performance towards test accuracy then is compared to identify economic yet robust test concepts. These comparisons show that the available repeatability accuracy of current test setups is more than sufficient for today's AEB system capabilities. Time discretization problems dominate the performance spread especially in test scenarios with a limited pedestrian dummy reveal time (e.g. child behind obstruction, running adult scenarios with low car speeds). This would allow to increase test tolerances to decrease test cost. A methodology which allows to derive the required tolerances in active safety tests might be valuable especially for NCAPs of emerging countries that do not have the necessary equipment (e.g. driving robots, positioning units) available for the full-scale and high tolerance EuroNCAP active safety procedures yet still want to rate active safety systems, thus improving the global safety.
Euro NCAP will start to test pedestrian Automatic Emergency Braking Systems (AEB) from 2016 on. Test procedures for these tests had been developed by and discussed between the AsPeCSS project and other initiatives (e.g. the AEB group with Thatcham Research from the UK). This paper gives an overview on the development process from the AsPeCSS side, summarizes the current test and assessment procedures as of March 2015 and shows test and assessment results of five cars that had been tested by BASt for AsPeCSS and the respective manufacturer. The test and assessment methodology seems appropriate to rate the performance of different vehicles. The best test result - still one year ahead of the test implementation - is around 80%, while the worst rating result is around 10%. Other vehicles are between these boundaries.
During the past five years, a Euro NCAP technical working group on pedestrian safety has been working on improving test and assessment procedures for enhanced passive pedestrian safety. After harmonizing the tools and procedures as much as possible with legislation, the work was mainly focused on the development of grid procedures for the pedestrian body regions head, upper leg with pelvis and lower leg with knee. Furthermore, the test parameters for the head and the upper leg were revised, a new lower legform impactor was introduced and the injury thresholds were adjusted or, where necessary, the injury criteria were changed. Finally, the assessment limits and colour scheme were refined, widening the range and adding two more colours in order to provide a more detailed description of the pedestrian safety performance. By abstaining from an assessment based on a worst point selection philosophy, the improved test point determination procedures that were introduced during the years 2013 and 2014 give a more homogeneous, high resolution picture of the pedestrian safety performance of the vehicle frontends. By using a uniform grid for each test zone approximately 200 test points, evenly distributed within each area, can now be assessed per vehicle. The introduction of the flexible pedestrian legform impactor in 2014 enables a more realistic injury prediction of the knee and the tibia using a biofidelic test tool. With the new upper legform test that has been launched in 2015 the assessment in that area is now focusing on the injured body region instead of the injury causing vehicle part and thus is aligned with the approach in the remaining body regions head and lower leg. At the same time, a monitoring test with the headform impactor against the bonnet leading edge is closing the possible gap between the test areas to identify injury causing vehicle parts that moved out of focus due to the introduction of the new upper legform test. The paper describes the new test and assessment procedures with their underlying philosophy and gives an outlook in terms of open issues, specifying the needs for further improvement in the future. In parallel to the work of the pedestrian subgroup, a Euro NCAP working group on heavy vehicles introduced a set of protocol changes in 2011 that were related to the assessment of M1 vehicles derived from commercial vehicles, with a gross vehicle weight between 2.5 and 3.5 tons and 8 or 9 seats. The paper also investigates the applicability of the new pedestrian test and assessment procedures to heavy vehicles.
Except for corrective steering functions automatic steering is up to now only allowed at speeds up to 10 km/h according to UN Regulation No. 79. Progress in automotive engineering with regard to driver assistance systems and automation of driving tasks is that far that it would be technically feasible to realise automatically commanded steering functions also at higher vehicle speeds. Besides improvements in terms of comfort these automated systems are expected to contribute to road traffic safety as well. However, this safety potential will only be exhausted if automated steering systems are properly designed. Especially possible new risks due to automated steering have to be addressed and reduced to a minimum. For these reasons work is currently ongoing on UNECE level with the aim to amend the regulation dealing with provisions concerning the approval of steering equipment. It is the aim to revise requirements for automatically commanded steering functions (ACSF) so that they can be approved also for higher speeds if certain performance requirements are fulfilled. The paper at hand describes the derivation of reasonable system specifications from an analysis of relevant driving situations with an automated steering system. Needs are explained with regard to covering normal driving, sudden unexpected critical events, transition to manual driving, driver availability and manoeuvres to reach a state of minimal risk. These issues form the basis for the development of test procedures for automated steering to be implemented in international regulations. This holds for system functionalities like automatic lane keeping or automatic lane change as well as for addressing transition situations in which the system has to hand over steering to the driver or addressing emergency situations in which the system has to react instead of the driver.
The UN Regulation No. 79 is going to be amended to allow automatically commanded steering functions (ACSF) at speeds above 10 km/h. Hence, requirements concerning the approval of automatically performed steering manoeuvres have to be set in order to allow safe use of automatic steering on public roads as well as improve overall road safety for the driver and the surroundings. By order of the German Federal Ministry of Transport and Digital Infrastructure (BMVI), BASt developed and verified physical test procedures for automatic steering to be implemented in UN Regulation No. 79. The usability of currently available test tools was examined. The paper at hand describes these test procedures and presents results from verification tests. The designated tests are divided in three sections: functionality tests, verifications for the transition of control and emergency tests. System functionality tests are auto matic lane keeping, automatic lane change and an automatic abort of an initiated lane change due to traffic. Those tests check if the vehicle remains in its lane (under normal operating conditions), is able to perform safe automatic lane change manoeuvres and if it considers other road users during its manoeuvres. Transition tests examine the vehicle's behaviour when the driver fails to monitor the system and in situations when the system has to hand over the steering control back to the driver. For instance these tests provoke driver-in-the-loop requests by approaching system boundary limitations, like missing lane markings, surpassing maximum lateral acceleration in a bend or even a major system failure. Even further the driver and his inputs are monitored and if the system detects that he is overriding system actions or contrary want to quit the driving task and unfastens the seat belt, it has to shut down and put the human back into manually control and the responsibility of driving. The last series of test consists of two emergency situations in which the system has to react to a time critical event: A hard decelerating vehicle and a stationary vehicle in front both with no lane change possibility for the ACSF vehicle. Some of the tests, especially the emergency manoeuvres, require special target vehicles and propulsion systems. Since no fully automatic steering vehicles are available, a current Mercedes E-Class with Mercedes' "drive pilot" system was used. It was shown that the vehicle is automatically able to brake to a full stop towards a static Euro NCAP target from partial-automatic driving at 90 km/h, that it could brake towards a rapidly decelerating lead vehicle when travelling at 70 km/h, that it was able during partially automatic driving to remain in its lane in normal operation conditions and to perform a automatic (driver initiated) lane change while surveilling the driver- activities.
Trauma management (TM) covers two types of medical treatment: the initial one provided by Emergency Medical Services (EMS) and a further one provided by permanent medical facilities. There is a consensus in the professional literature that to reduce the severity and the number of road crash victims, the TM system should provide rapid and adequate initial care of injury, combined with sufficient further treatment at a hospital or trauma centre. Recognizing the important role of TM for reducing road crash injury outcome, it was decided, within the EU funded SafetyNet project, to develop road safety performance indicators (SPIs) which would characterize the level of TM systems" performance in European countries and enable country comparisons. The concept of TM SPIs was developed based on a literature study of performance indicators in TM, a survey of available practices in Europe and data availability examinations. A set of TM SPIs was introduced including 14 indicators which characterize five issues such as: availability of EMS stations; availability and composition of EMS medical staff; availability and composition of EMS transportation units; characteristics of the EMS response time, and availability of trauma beds in permanent medical facilities. Basic information on the TM systems was collected in close cooperation with the national expert group. A dataset with TM SPIs for 21 countries was created. It was demonstrated that the countries can be compared using selected TM SPIs. Moreover, a more general comparison of the TM systems' performance in the countries is possible, using multiple ranking and statistical weighting techniques. By both methods, final estimates were received enabling the recognition of groups of countries with similar levels of the TM system's performance. The results of various trials were consistent as to the recognition of countries with high or low level of the TM systems" performance, where in grouping countries with intermediate levels of the TM system's performance some differences were observed. The SafetyNet project's practice demonstrated that data collection for estimating TM SPIs is not an easy task but is realizable for the majority of countries. The TM SPIs" message is currently limited to the availability of trauma care services. Further development of the TM SPIs should focus on characteristics of actual treatment supplied, based on combined police and medical road crash related databases.
Assessment of the effectiveness of Intersection Assistance Systems at urban and rural accident sites
(2015)
An Intersection Collision Avoidance System is a promising safety system for accident avoidance or injury mitigation at junctions. However, there is still a lack of evidence of the effectiveness, due to the missing real accident data concerning Advanced Driver Assistance Systems. The objective of this study is the assessment of the effectiveness of an Intersection Collision Avoidance System based on real accidents. The method used is called virtual pre-crash simulation. Accidents at junctions were reconstructed by using the numerical simulation software PC-Crashâ„¢. This first simulation is called the baseline simulation. In a second step the vehicles of these accidents were equipped with an Intersection Collision Avoidance System and simulated again. The second simulation is called the system simulation. In the system simulation two different sensors and four different intervention strategies were used, based on a time-to-collision approach. The effectiveness of Intersection Collision Avoidance System has been evaluated by using an assessment function. On average 9% of the reviewed junction accidents could have been avoided within the system simulations. The other simulation results clearly showed a change in the principal direction of force, delta-v and reduction of the injury severity.
The main focus of the benefit estimation of advanced safety systems with a warning interface by simulation is on the driver. The driver is the only link between the algorithm of the safety system and the vehicle, which makes the setup of a driver model for such simulations very important. This paper describes an approach for the use of a statistical driver model in simulation. It also gives an outlook on further work on this topic. The build-up process of the model suffices with a distribution of reaction times and a distribution of reaction intensities. Both were combined in different scenarios for every driver. Each scenario has then a specific probability to occur. To use the statistical driver model, every accident scene has to be simulated with each driver scenario (combinations of reaction times and intensities). The results of the simulations are then combined regarding the probabilities to occur, which leads to an overall estimated benefit of the specific system. The model works with one or more equipped participants and delivers a range for the benefit of advanced safety systems with warning interfaces.
The Traffic Accident Research Institute at University of Technology Dresden investigates about 1,000 accidents annually in the area around and in Dresden. These datasets have been summarized and evaluated in the GIDAS (German Accident In-Depth Study) project for 13 years. During the project it became apparent that the specific traffic situation of a covert exit of a passenger car and an intersecting two-wheeler involves a high risk potential. This critical situation develops in a large part due to the lack of visibility between the driver and the intersecting bike. In this paper the accident avoidance potential of front camera systems with lateral field of view, which allows the driver to have an indirect sight into the crossing street area will be presented.
The project UR:BAN "Cognitive assistance (KA)" aims at developing future assistance systems providing improved performance in complex city traffic. New state-of-the-art panoramic sensor technologies now allow comprehensive monitoring and evaluation of the vehicle environment. In order to improve protection of vulnerable road users such as pedestrians and cyclists, a particular objective of UR:BAN is the evaluation and prediction of their behaviour and actions. The objective of subproject "WER" is development support by providing quantitative estimates of traffic collisions at the very start and predict potential in terms of optimized accident avoidance and reduction of injury severity. For this purpose an integrated computer simulation toolkit is being devised based on real world accidents (GIDAS as well as video documented accidents), allowing the prediction of potential effectiveness and future benefit of assistance systems in this accident scenario. Subsequently, this toolkit may be used for optimizing the design of implemented assistance systems for improved effectiveness.