ABSTRACT. Intelligent Transport Systems (ITS) are currently being developed in a lot of different industry sectors. The biggest challenge is to make those systems safe and reliable. Main focus of ITS in automotive industry is Highly Automated Driving (HAD). Highly Automated Driving is in an advanced development state and will get into production status in a few years. This new technology is required to be safe and reliable by design, not by bad experience and enhancements. In complex ITS, where system boundaries are not limited to single products from one OEM, safety and reliability is a question of trust. Can we trust those vehicles with variable HAD system boundaries, that they do not endanger participants and their surrounding field? From a technical point of view, a trust chain of all required technologies is essential. But which technology is trustful enough? Could it be surrounding field sensing (camera, laser or radar), global dynamic maps, local dynamic maps, GNSS, V2X, sensor data fusion or driving algorithms?

Who should be allowed to update maps? Can you trust the map authors? What has to be done to trust algorithms? What about data age? How is trust diminishing with growing data age? And what level of trust is really needed for highly automated driving? This keynote illustrates the problem of trust for highly automated driving and discusses high level strategies for getting trust into automated vehicles.

ABSTRACT. A new method for generating a separate two-dimensional free space grid map for ADAS based on data from radar sensors is presented in this paper. We introduce a free space model based on an inverse sensor model to compute the Gaussian-based free space probability for each cell of the free space grid map. A Bayesian approach is used for free space probability estimation independently from the occupancy probability, which enables increased amount of information for environment description. For this purpose, two free space grid maps are generated: The instantaneous free space map is generated in each measuring cycle and the accumulated free space map is generated once and updated in each measuring cycle. We describe how the free space grid maps are generated and updated by new observations. In contrast to other approaches, the detection accuracy is taken into account in the free space model. Finally we present the experimental results obtained from real world environments.

ABSTRACT. Localization is the process of nding ones relation to the surrounding stationary objects, and mapping is the process of determining the relation between the stationary objects. Mapping requires a sensing technique, and in addition to that a known (or estimated) location. If we neglect the possibility of external support such as GPS or street maps and instead consider localization using ranging sensors, we are facing two strongly interconnected problems that needs to be solved simultaneously. We have investigated the possibility to perform SLAM (simultaneous localization and mapping) in automotive using a forward-looking radar, primarily designed for ACC (adaptive cruise control) and PCS (pre-crash safety). Our positioning is not only relying on the inertial measurements speed and yaw-rate from the vehicle, but also
incorporates radar measurements and performs extended Kalman filter SLAM (EKF-SLAM). Using collected data from a single forward-looking radar, we have shown that it is possible to enhance the positioning performance without support from GPS. The heading of the vehicle was drifting as yaw-rate error accumulated, but when adding EKF-SLAM we mitigated this problem. However, this system was very sensitive to parameter settings, as well as radar misalignment, and needs a thorough on-line calibration. Robustness can be achieved by increasing the field of view or having side-looking radars.

ABSTRACT. The purpose of this research is to develop a method to detect driver drowsiness with pressure distribution sensors on the driver seat. Use of such sensors has an advantage that the measurement can be done in a non-intrusive manner, and that sensing data can always be obtained while driving. We conducted an experiment with a fixed-base driving simulator. The results suggested that driver body movement increases when driver drowsiness begins to increase.

ABSTRACT. Uncooled sensors for use in night vision are a promising technology for use in active driver-assistance systems (ADAS) in combination with active-type sensors. However, the sensitivity of conventional uncooled sensors is low because attempts to achieve high sensitivity with conventional sensors are hampered by their high resistivity, which increases thermal noise. To overcome this problem, we have developed new uncooled sensors that have a low resistivity and high sensitivity. This was achieved by improving the crystal quality of the detector materials. Consequently, the sensitivity of the developed sensors was about five times higher than that of conventional low-resistivity sensors. This result could expand the range of use of such sensors and might improve object-recognition accuracy in forward-facing ADAS.

ABSTRACT. We tried to develop the new localization method by simple 2D-plane map without deterioration of estimation accuracy. The boundary line has a lot of features e.g. changes of height, color and brightness, but they are sensitive for noises. From the robustness point of view, it is difficult to match the road boundary line with the boundary line on 2D map. The localization method using 3D point cloud matching or texture matching are so accurate, but these have disadvantage in adaptation to the change of environment. So, we decide to make the classifier to classify as road area or the other area, and propose the new localization method that has advantage in robustness by matching the identified shape of road area with the shape of the road on 2D plane map. First, we calculate the HOG features from the range data acquired by 3D LiDAR. Then, we make the road plane classifier applying SVM.

ABSTRACT. Suddenly occurring disruptions on the road and especially at intersections – such as accidents, disabilities by vehicles and objects – may cause significant adverse effects on the traffic flow and in turn lead themselves to potentially dangerous situations. By using its integrated GNSS receiver and its mobile communication device the smart traffic cone provides timely and spatially accurate information about local disruptions in the road environment. It is a tool for police, rescue, recovery and service staff for detecting hazard areas that could not be detected so far.

ABSTRACT. This paper presents an artificial vision based video-sensor designed to detect pedestrian-vehicle conflicts at crossing points. This video sensor detects moving objects by isolating them from the background. Then the speed and trajectories are estimated using a Kalman filter. Potential conflicts are then predicted. This system has been tested at two real crossing points at the city of Salamanca (Spain).

ABSTRACT. Especially at intersections, pedestrians and cyclists are endangered by crossing vehicles. Modern sensor and radio communication technologies oer the possibility to detect dangerous situations and to prevent accidents. In the joint research project UR:BAN, the following is being researched and implemented prototypically: At intersections, so called vulnerable road users are being detected by radar, as well as their mobile devices, like smartphones. After the upcoming introduction of the new communication technology V2X, vehicles can distribute information gathered by their internal sensors. All these data sources serve a so called cyclist and pedestrian protection system. It calculates trajectories and predicts
possible collisions. V2X and mobile devices with Wi-Fi oer a return channel to warn all road users in time and in an appropriate way.

ABSTRACT. Still most commercial navigation tools used by pedestrians fail to encompass a comprehensive organization and prioritization of safety-related route qualities and accordant information in the user interface. To support pedestrian route choices to minimize potential dangers, we study in this paper user requirements for an enhanced pedestrian navigation system that considers safety related route quality parameters. Besides effectiveness, related factors of distance and time, safety was highly prioritized to become an explicit requirement for the conceptual design. The acquired data from an online survey provides the basis for pedestrian's classications and requirements regarding user friendly interfaces for mobile routing and navigation that enhance road safety.

ABSTRACT. Normally, pedestrians do not walk on streets but on sidewalks. Thereby, crossing the road is often necessary. Particularly in peripheral areas, dedicated crosswalks are rare and pedestrians are forced to cross roads apart from them. In order to integrate road crossing into a routing and navigation system for pedestrians a decision of where to cross the road is needed. In order to express how safe it is to cross the road at a certain location we suggest an index. Among others, the value of the index depends on various criteria as the roads' geometries, the traffic volume and the speed of the vehicles. In order to safely cross a road, a free gap with a certain length between two consecutive vehicles on a lane is required. The calculation of the index is mainly based on the probability of the availability of such a gap. All criteria are directly extracted from OpenStreetMap or derived from them.

ABSTRACT. Japan is facing up to the challenges of a rapidly aging society. In addition to measures to help vitalize this situation, the automotive sector is also working to resolve issues such as congestion and traffic accidents. With this background, My talk introduces two collaborative projects about intelligent vehicle research and development in Japan. The first project aims to develop an intelligent driving system to achieve a safe and secure traffic society for elderly drivers. The main purpose of this project is to realize an intelligent driving system incorporating an experienced driver model to help recover the deterioration in the recognition, decision-making, and operation capabilities of elderly drivers, and to achieve a significant improvement in road safety. The second is the Smart Traffic Flow Control Project. This project focuses on the fact that an advanced driver assistance system (ADAS) with driving intelligence has the potential not only to enhance safe and secure driving but also to reduce congestion. This paper uses these topics to describe perspectives about intelligent vehicle technologies.

ABSTRACT. Predicting future risk during driving in urban road is one of key solutions to enhance safety performance of vehicles. This study proposes a motion planning and control system based on collision risk potential prediction characteristics of experienced drivers. By optimizing the potential field function in the framework of optimal control theory, the desired yaw rate and the desired longitudinal deceleration are theoretically calculated. The validity of the proposed motion planning and control system is verified by comparing the simulation results with the actual driving data by experienced drivers. The safety performance of the intelligent driving system to avoid potential collisions is shown in pedestrian crossing scenario.

ABSTRACT. Industry Talk - not conference paper - invited

ABSTRACT. In automobile industry the ongoing development of vehicle functions plays an important role for safety and comfort. The enlargement of the driver’s environment perception is an enormous potential because additional information provide the opportunity of developing novel functions. In this context the Car2Car Communication Consortium launched the standardization for V2X-communication. A transceiver (CCU) offers a WLAN-based exchange of information between road users (Vehicle-to-Vehicle) or road user and infrastructure (Vehicle-to-Infrastructure). The CCU could be seen as a further sensor like radar and camera, which for example is able to receive information about the position of other road users and / or the signal phase of the relevant traffic light (TL). It is possible to design cooperative functions, which can process more than one sensor within a sensor data fusion. These functions can be subdivided into the scopes comfort and safety. For example a vehicle controls the own velocity by using an actuator due to a V2X-message sent by the relevant TL. As a result the vehicle is able to brake automatically in front of the TL and run-up automatically if the red signal phase is over. For a proper behavior of the comfort function, the detection of amongst others road users and lanes (vehicle’s environment) is necessary. This can be done via radar and / or camera. To ensure the safety of these kind of comfort functions, an implementation of a monitoring is necessary (“Safety Layer”). This scope has safety functions like the Automatic Emergency Brake (AEB), which can be triggered by receiving V2X-messages. Target of this paper is the discussion of a V2X-based Emergency Brake as a part of the functional Safety Layer.

ABSTRACT. ADAS in HCV.Scania Heavy-DAS Project (1st autonomous Truck). Automated Mining. Active Brake Systems.

ABSTRACT. Smart portable device used in car provides a cost effective approach to obtain driving dynamic signals and location information by utilizing its inertial sensors. MobileUTDrive is an Android App that we developed for in-vehicle data collection. This paper first describes motivation of using portable device for driving record, and then discuss the system design in hardware and software aspects. Finally, and details of software implementation is presented.

ABSTRACT. This study proposes a model to estimate the number of accidents using two parameters. One parameter expresses quality of active safety, and the other one expresses quality of passive safety. This model uses data on accident occurrence from 1999 to 2009. The estimated parameters indicate that during this period the quality of active safety did not increase, whereas the quality of passive safety did increase.

ABSTRACT. Merging into the traffic flow on an expressway is a challenging driving task. We aim to develop a driver assistance system for such a demanding driving task. Our previous study revealed that a driver’s decision making of the merging space in the oncoming traffic skill can be inferred by the driver’s behaviors after the driver was able to see the environment of the main lane. In addition, the study suggested that the merging behaviors were different among drivers and it seemed to depend on driver’s skill. Thus, the present paper proposed a method to characterize driver’s skill in merging operation based on the driver model and driver’s active longitudinal control toward a driver assistance system for merging.

ABSTRACT. For the purpose of enhancing road markings to prevent speeding on curved road, we focused on the vehicle speed inducement effect of sequence patterns of sequential transverse and lateral markings on a straight section of roadway leading into a transition section of the curved road. We tested the effectiveness of several sequence patterns through a driving simulation experiment with a driving simulator (DS) and a driver perception experiment with recorded moving pictures (MPs). The sequence patterns were characterised by patterns of progressively and concurrently reduced spacing of transverse lines on the lane surface and lateral poles on both the shoulder edge and the median strip. While estimating trends in spot speed that was perceived by a driver (test subject) who went into the curve entrance under the influence of sequence patterns, we examined a discrepancy between the perceived speed transition and the vehicle speed transition, by using hidden Markov model (HMM) on the estimation. We prepared four types of the sequence patterns including the types which had greater decrease rates of the intervals of spacing in the beginning, middle, and end sections than in the remaining sections among all the sections consisting of the pattern. The experimental results concluded that the sequence pattern type which had greater decrease rates in the end section than in remaining sections might be encouraged to be laid on the curved road, in terms of a safe vehicle speed inducement effect.

ABSTRACT. This paper describes a method of lane-level localization for automated driving using around view monitoring (AVM) camera. Today’s on-board sensors such as radar or camera do not reach a satisfying level of development from the point of view of robustness and availability. Thus, map data is often used as an additional data input to support these systems. A digital map is used as a powerful additional sensor. So we propose a lane map-based localization using an AVM camera. The maps are created beforehand using an AVM camera and RTK GPS. A pose estimation of vehicle was derived from a low-cost GPS and an iterative closest point (ICP) match of real-time sensor data to lane map. And the estimated pose was used as an observation inside a Kalman filter framework. The performance of the proposed localization algorithm is verified via vehicle tests in ITS proving ground. It has been shown through vehicle tests that good localization performance can be obtained. The proposed algorithm will be useful in the implementation of automated driving.

ABSTRACT. The motivation of this work is to enhance personal mobility vehicles (PMV) with autonomous vehicle technologies, such as autonomous vehicle platooning to realize new potentials and possibilities for future urban transportation system. This paper reports our development on the automatic vehicle following system building for autonomous platooning of PMV. We have developed a 3D sensing system for both object (front vehicle) detection and longitudinal distance sensing. We have conducted straight and curve paths vehicle following experiments using the automatic vehicle following system to fine-tune the control parameters. We have furthered our investigation to evaluate the performance of the developed system in autonomous platooning of two automatic following PMVs with a leading human-driven PMV. The experimental results have proven the effectiveness of the developed vehicle following system to achieve autonomous platooning of PMV.

ABSTRACT. In this paper, a control design approach known as interpolating control is used for cooperative vehicle longitudinal control in order to form vehicle platoons. The objective is to regulate the vehicles’ speeds and the spacings between the vehicles, from their initial conditions into a shared consensus. A discrete state space formulation is used to model the system, in which constraints are enforced. The interpolating control approach is implemented and compared with other methods such as MPC. The paper presents an implementation of the interpolating controller that includes integral action, for the purpose of improving the steady state performance in the presence of a disturbance. We show that this controller can indeed eliminate the steady state error, defined as the output of the system, if the disturbance is a bounded step function and the initial conditions are feasible.

ABSTRACT. To realize driver assistance systems based on automated driving technologies, intelligent vehicles need to recognize surrounding driving environments. On this point, sensing technologies with high-cost sensors have several problems for future popularization. Therefore, this research aimed at developing automated driving technologies with lean sensors via the enhancement of existing ADAS Horizon.

ABSTRACT. Based on an analysis of collision risk propagation, a vehicle Traffic Predictive Cruise Control (TPCC) system, responding to the change of downstream traffic situation, is proposed in this study to improve traffic operation, safety, and fuel efficiency of vehicle. The proposed TPCC system consists of four parts: (1) Collision risk calculator of a subject vehicle, which represent the state of the subject vehicle. (2) Vehicle Control algorithm only based on the collision risk of the subject vehicle, (3) Cooperative measure for representing downstream traffic state, which is based on the results of an analysis of collision risk propagation, (4) TPCC algorithm, which controls the vehicle by using both collision risk of the subject vehicle and cooperative measure. By using both collision risk of subject vehicle and cooperative measure, which represent the integrated collision risk of leader vehicles, TPCC is designed to proactively determine actuation of vehicle by adjusting parameters of vehicle control algorithm before high collision risk arisen from leader vehicles reaches to the subject vehicle. A simulation using the real vehicle trajectories from the NGSIM data validates the performance of TPCC algorithm with various market penetration rates. It is found that the proposed TPCC system can contribute to CO2 emission reduction, traffic flow stability, and safety improvement. Such results are due to the effects of suppression of the high collision risk generated from downstream traffic and removal of unnecessary fluctuation of speed.

ABSTRACT. This paper analyzes the potential safety benefit from autonomous acceleration of an electrified lead vehicle to mitigate or prevent being struck from behind. Safety benefit was estimated based on the expected reduction in relative velocity at impact in combination with injury risk curves. Potential issues and safety concerns with the operation and implementation of such a system in the real world are discussed from an engineering and human factors stand point. In particular, the effect of the pre-collision acceleration in reducing whiplash injury risk due to change in head posture and reduction of crash severity is also discussed. In general, this study found that autonomously accelerating an electrified lead vehicle can mitigate and prevent rear-end collisions and significantly increase the safety benefits from existing systems such as autonomous emergency braking.

ABSTRACT. This paper describes the factor of crossing collision involving bicycle at non-signalized intersection by analyzing near-crash incident. In recent years, the traffic accident shows gradual decreasing in the number of fatalities, whereas slowly decreasing the total number of accidents and injuries in Japan. In order to decrease the total number of accident, it is necessary to investigate the factors of accident. So the feature of crossing collision involving bicycle at non-signalized intersection was analyzed by traffic accident data and near-crash incident database collected with the drive recorders.

ABSTRACT. Collision Warning (FCW) systems that have customizable warning delivery settings may improve driver acceptance, thus increasing the benefits of such systems. In order to design FCW warning thresholds that match a driver's expectations, system designers need to characterize when the brakes are normally applied. However, a driver's normal braking behavior may vary with the driving context, e.g., traffic congestion or daylight conditions. This study examined over 2.6 million brake applications from the 100-Car naturalistic driving study to determine the effect of driver demographics (age group and gender) and driving context (day of week, time of day, travel speed, and traffic congestion) on brake application time. The results showed that both demographics and driving context were statistically significant indicators of the time to collision (TTC) that drivers applied the brakes during car following.

ABSTRACT. Car-to-X technology enables vehicles to directly exchange information with other vehicles or with roadside infrastructure components using standardized communication and message protocols. So far, several cooperative Car-to-X use cases were defined to improve road safety and traffic efficiency, which will be introduced in different stages. In this paper, we will evaluate these use cases of the early deployment phase – such as electronic emergency brake lights or green light speed optimal advisory – with respect to their suitability to automated driving. We will see that their current specification only have minor contributions for improving automated driving scenarios. However, we will also see that by improving this use cases we can have a significant impact on autonmated driving technology. Moreover, our evaluation shows that Car-to-X is an essential base technology for cooperative automated driving scenarios.

ABSTRACT. In this research, we focus on improving acceptability of proactive collision avoidance systems to elderly drives by using passive information sharing with drivers. In this paper, the information sharing mainly consists of visual contents. As a result of the evaluation experiment by a driving simulator to investigate the effectiveness of the information sharing, we confirmed that passive information sharing were able to improve the acceptability.

ABSTRACT. Autonomous vehicles allow the driver to be out of the loop of the driving task, under particular conditions. Interaction between the driver and the technical agent is crucial especially in autonomous mode and in handover processing. It is essential to identify the information requirements to meet the driver’s needs. Technology advancement has introduced Augmented Reality, which is said to enhance vision. In this paper, we explore the use case of lane change. By applying a cognitive method, we extract information related to lane change maneuver. We present what the Augmented Reality representation will potentially look like at the end of our study. Finally, future research avenues are outlined.

ABSTRACT. ISO 26262, the functional safety standard for automotive electric and electronic (E/E) systems, requires a controllability assessment to be made as part of the hazard and risk classification process. As well as influencing the function’s Automotive Safety Integrity Level (ASIL), the verifiable controllability may also limit the functions intervention options and intensity during normal operation. For electric driven vehicles this limits their accident-avoidance/-mitigation potential. For an in-wheel motor driven electric vehicle it is questioned whether the failure of a motor could lead to a risk. It is obvious that the result of the risk assessment depends on the operating scenarios chosen. As numerous factors define a driving situation, the possible detailing of these factors is unlimited. In a previous paper, we have presented the results of a study regarding the controllability of a vehicle driven by in-wheel motors using a simplified linear bicycle model. In this paper we extend the previous work by qualitatively and quantitatively identifying the hazards associated with in-wheel motors and by quantify the vehicle level effects that could be expected using validated detailed multibody vehicle models in both straight line and cornering events.

ABSTRACT. Due to the globalized development of vehicles and advanced driver assistance systems (ADAS) in combination with the large variety of traffic situations all over the world, there is an increas-ing need of evaluating the effectiveness of ADAS on the basis of international real traffic data. This can be done with the help of so-called pre-crash matrices (PCM), which describe the vehicle dynamics in a defined time before the collision. After the publication of the first IGLAD data (1.550 accidents) a study was done dealing with the creation of pre-crash simulations out of international accidents. For the first time the benefit of an ADAS can be evaluated prospectively in the wide variety of global traffic accident scenario. This paper provides an overview of the challenges that come with merging data from different investigation areas. The main focus will be on the methodology to derive PCM from this international d a-tabase. This also includes the definition of minimum requirements to enable the simulation of the vehicle behavior in the pre-crash phase. Furthermore, methods were developed how to deal with unknown data with regard to the different data quality and quantity. Finally the paper shows the unique possibility to analyze active safety systems from a global point of view by implementing and assessing an exemplary ADAS for different global traffic accident scenarios.

With the work done within the study, especially with the definition of minimum requirements and the de-veloped methods, it is possible to create pre-crash simulations not only for upcoming iGLAD releases but also for other international accident databases.

ABSTRACT. Elucidation of the relationship between brain activity and driver behavior may assist in the development of new driver models for next-generation driving-assistant systems that adapt to drivers' individual characteristics. However, multiple regions of the brain are involved in driving, so it is first necessary to investigate the role of each region. In this paper, we examined the relationship between driving skill and parietal lobe activity. We performed experiments using a driving simulator featuring a curved course modeled on a real test course. When drivers steered around the curve, data on their driving operations and the state of their vehicle were recorded, and their cortical activity was measured using functional near-infrared spectroscopy. Subsequently, jerk, which is the derivative of acceleration with respect to time, was utilized to divide drivers into the high-skilled and low-skilled groups. We found that high-skilled drivers operated the accelerator pedal and steering wheel smoothly while steering into the curve. Simultaneously, the parietal lobe was more active in the high-skilled group than in the low-skilled group at the entrance to the curve. The parietal lobe is known to integrate sensory information from various modalities. Therefore, our findings suggest that the integration of sensory information strongly influences driving skill.

ABSTRACT. French government decided to support the SVRAI project (Saving Lives through Road Incident Analysis Feedback) to answer the question "Can incident data analysis help to avoid accidents?" Thus a 12 months data collection involving 50 public vehicles, fitted with a dedicated EDR named EMMA, was carried out. This paper focuses on results concerning vehicle dynamics: 339 incidents were collected, among which 70 % concern the lateral direction, with 70 % in right hand turns vs 30 % in left ones. Based on additional data synthesizing complete travels, 3D representations of dynamic parameters are also shown interesting to characterize driver performance.

ABSTRACT. This paper developed a driving support system to reduce the driver’s dilemma at signalized intersection. When facing an amber signal phase, the drivers occasionally come across the dilemma whether they should stop or pass the intersection stop line. The proposed system provides the remaining green time (RGT) to the drivers in real-time to support the stop/pass decision making. Driving simulator experiment equipped with the RGT indication system showed that the proposed system significantly reduced the probability of facing the dilemma and encouraged the drivers to decelerate earlier and stop before entering the intersection. Also, logistic regression analysis (LRA) revealed that approaching speed of a vehicle is the most significant factor to affect the stop/pass decision before the amber onset whereas remaining distance and acceleration rate affect it much more after fallen into the amber phase.

ABSTRACT. Post-crash safety technologies are an important part of efforts to eliminate traffic accident-related fatalities. They are developed to enhance crash safety, and are therefore based on data collected from airbag control module (ACM) operation. In particular, delta-V (technically defined as the velocity vector difference before and after impact) data are very important in determining that a collision has occurred and evaluating accident severity. The aim of this study was to assess delta-V characteristics and to clarify the performance of event data recorders in ACM operation with focus on delta-V values recorded during collision with a tubular metal guardrail.

ABSTRACT. Autonomous braking systems have potential benefits in active safety systems and Advanced Driver Assistance Systems. Ideally, emerging driver assistance systems which can automate certain driving aspects would apply braking in a human-like fashion. A better understanding of driver braking behavior can assist active safety and driver assistance system designers to better tailor the vehicle braking pattern to the driver and driving context. The objective of this study was to determine the potential effect of driver age and gender on braking profile. The approach of this study was to extract braking patterns in normal driving from the 100-Car Naturalistic Driving Study. Braking events with a closing lead vehicle were identified and extracted from the database. For each braking event, maximum brake force and braking profile was calculated from the instrumented vehicles. The result of the study shows that driver age and gender, as well as vehicle speed at start of braking, all have a statistically significant effect on driver braking profile. The results of this study have substantial implications for improving future autonomous braking system design to better tailor the system activation time to individual driver according to age, gender, and vehicle speed.

ABSTRACT. Over the last decade, collision avoidance technologies targeting rear-end collisions have been introduced by many vehicle manufacturers. However, evaluation of the real world performance of these systems are rare. The objective of this study was to evaluate the real world effectiveness of systems called Forward Collision Warning and Brake support combined with Adaptive Cruise Control (CWB+ACC). These systems were introduced as optional equipment in Volvo car models in 2006. The data analyzed comes from a detailed, representative dataset based on insurance claims. The rate of rear-end frontal collisions was compared for cars with and without CWB+ACC, controlling for different generations of CWB+ACC as well as presence of Low-speed Emergency Braking functionality. For cars with CWB+ACC, rear-end crashes with frontal impacts were reduced with 38%. Also, the data showed a clear progress in crash avoidance efficiency as a function of CWB+ACC development. For the third generation of CWB+ACC, the estimated collision avoidance effect was 45%. In future studies, the additional safety performance that collision avoidance technologies bring in the form of crash mitigation needs to be investigated.

ABSTRACT. This paper presents enhanced versions of two metrics: the Time-to-Brake (TTB) and the Brake-Treat-Number (BTN), which are used as measures to describe the degree of being critical of traffic situations. The main idea is to include road information as input to obtain a more advanced prediction of the leading vehicle.

The results, illustrated by an example using real data, show a better assessment of the collision potential hazard and no false alarms.