It is well recognized that knowledge on the lane and road geometry is a prerequisite for reliable function of the majority of automotive safety applications, especially those that deal with vehicles’ lateral control. An effort to exploit all the available road and lane related information is presented in this work. It includes redundant information – from secondary sensor detections, e.g. stationary objects detected by the laser scanner – and data from dedicated sources of information, e.g. on-board digital map, and lane detection vision systems. The test vehicle system comprises a CMOS camera that identifies lane markings, while a laser scanner detects road borders position. Road geometry information from digital map databases is also fused with the above providing a complete representation of the road environment. In this paper the four individual estimators (data from digital map, camera, vehicle dynamics and laser scanner) and the combined fuser (with three different approaches) are presented and compared. A thorough examination of the system’s performance using real data recordings took place and the results are reported, as well.

The development of next generation driver-vehicle interaction systems should be focused towards obtaining a safe and sustainable mobility, with the aim to half the number of road accidents as proposed by the (2001). Mobility should be promoted in the future towards ‘intermodality’ in order to reduce traffic congestion and to optimise travel planning; however, towards this aim there is an increasing demand for on-board information systems. These needs together with the demand for new on-vehicle support and services and the need of the users to be connected to their own information cell (mobile phone, PDA, etc.) will unavoidably increase the number of interaction of the driver with the vehicle thus raising the potential risk of driver’s distraction and fatigue, which are among the main causes of road accidents.

Path prediction is the only way that an active safety system can predict a driver's intention. In this paper, a model-based description of the traffic environment is presented - both vehicles and infrastructure - in order to provide, in real time, sufficient information for an accurate prediction of the ego-vehicle's path. The proposed approach is a hierarchical-structured algorithm that fuses traffic environment data with car dynamics in order to accurately predict the trajectory of the ego-vehicle, allowing the active safety system to inform, warn the driver, or intervene when critical situations occur. The algorithms are tested with real data, under normal conditions, for collision warning (CW) and vision-enhancement applications. The results clearly show that this approach allows a dynamic situation and threat assessment and can enhance the capabilities of adaptive cruise control and CW functions by reducing the false alarm rate.

In safety automotive applications the system must me capable of early recognizing the maneuvers performed by the driver and the intention associated with them in order to take preventive measures or trigger warning alarms. This is done by the situation refinement level in the fusion system that processes the data provided by the on-board sensors. By recognizing relationships between entities of the road environment the system can react with more efficiency to the current situation. For example the intention of a lane change, the detection of an overtaking maneuver, the estimation of the lane in which the detected vehicle is located, help the system to decide which action must be taken in order to prevent an unwanted situation. This paper focuses on the investigation of methods regarding the identification of the maneuvering type and the intention associated.

Multi-sensor systems in automotive safety applications and sensor data fusion have become very popular in recent years. Sensors on board cars and active safety applications are increasing in number and the need to define a common method for object extraction and serving these applications has been recognized. Authors propose a high level fusion approach suitable for automotive sensor networks with complementary or/and redundant field of views. The advantage of this approach is that it ensures system modularity and allows benchmarking, as it does not permit feedbacks and loops inside the processing. In this paper this track level data fusion approach is introduced with the main focus to be on the data association of tracks coming from the on board sensors with distributed processing. The core of the proposed approach is the formulation of the data association problem in presence of multipoint objects and then the solution for (a) multidimensional assignment and (b) all around vehicle object maintenance. The motivation of this work is the research work that is carried out in the project PReVENT/ProFusion2 where the proposed algorithm is being tested in two experimental vehicles.

The development of a system that can be used for a safe, reliable, highly available onboard lane keeping support system is a critical research topic. One of the most important functions in driver assistant systems is the detection of unintentional lane departures. Current lane departure warning systems focus mainly in the detection of lane markings using vision sensors, such as CMOS cameras. In order to increase accuracy and robustness of such systems the utilization of digital maps is necessary. The goal of combining camera and map data is to extend the road geometry in further distances and eliminate false alarms based on unintentional maneuvers caused by the driver. The overall system efficiency is increased furthermore by using also vehicle dynamics and road geometry calculated using radar data.

Application of high level fusion approaches demonstrate a sequence of significant advantages in multi sensor data fusion and automotive safety fusion systems are no exception to this. High level fusion can be applied to automotive sensor networks with complementary or/and redundant field of views. The advantage of this approach is that it ensures system modularity and allows benchmarking, as it does not permit feedbacks and loops inside the processing. In this paper two specific high level data fusion approaches are described including a brief architectural and algorithmic presentation. These approaches differ mainly in their data association part: (a) track level fusion approach solves it with the point to point association with emphasis on object continuity and multidimensional assignment, and (b) grid based fusion approach that proposes a generic way to model the environment and to perform sensor data fusion. The test case for these approaches is a multi sensor equipped PReVENT/ProFusion2 truck demonstrator vehicle.

Automotive safety functions are divided, on one hand, into passive and active safety according to a time factor before an accident or a critical situation could occur; on the other hand, they focus on single slices around the intelligent vehicle, even if more and more sensors are added to improve system accuracy. The paper proposes new integrated functions under INSAFES sub-project of PReVENT merging and harmonizing lateral and longitudinal support functions, collision mitigation protective functions and road users’ protection systems. The advantage of INSAFES that is shown in the paper is that all functions share the same system resources — sensors and perception units — and trigger a given set of actuators and HMI controllers in order to communicate with the driver or intervene. The proposed system will be integrated and demonstrated in a passenger car and a truck.

This paper describes the algorithms that are being developed for the perception layer of the PReVENT subproject LATERAL SAFE. These algorithms aim at achieving a reliable representation of the objects and their kinematics, present at the lateral and rear field of the ego-vehicle. The work presented in this paper is within the fields of radar tracking, sensor network processing, image and stereo vision processing, and integration and fusion of sensor- level processed data. The perception layer of LATERAL SAFE is a distributed sensor-level fusion system that processes in a central level the tracks of four tracking systems: a rear looking long range radar, two lateral short range radar networks and a system of lateral and rear looking cameras.

The question raised in this paper, for the first time, is how the JDL model can be applied in multi-sensor automotive safety systems, since new sensors are integrated on-board, while new functions support the driver, intervene and control the vehicle. The paper proposes a hybrid hierarchical structure and develops a suitable functional model, namely the ProFusion2 (PF2) model; PF2 serves the broad automotive sensor data fusion community as a conceptual framework of common understanding and it provides recommendations and guidelines for implementation of fusion systems. Reference implementations are given as complete examples from the major automotive research initiative in Europe (PReVENTproject)

Nowadays, drivers have to cope with a growing amount of information coming from on-board information messages, telematics and advanced driver assistance systems. The interaction between the driver and these systems is critical, since they may distract the driver from the primary task of driving. The paper, addressing this problem, aims at presenting the methodological framework for the optimization of human machine interfaces (HMI) in the automotive research area; thus, the proper communication and interaction strategies are designed, in order to deliver to the driver a message or a warning in the optimal way in terms of driver safety. The proposed methodology is adopted in the COMUNICAR project and relevant results are presented. Last but not the least, the AIDE integrated project and its vision is also proposed as the roadmap for future activities in the HMI sector.

There is a strong belief that the improvement of preventive safety applications and the extension of their operative range are achieved by the deployment of multiple sensors with wide fields of view (FOV). The paper contributes to the solution of the problem and introduces distributed sensor data fusion architectures and algorithms for an efficient deployment of multiple sensors that give redundant or complementary information for the moving objects. The proposed fusion architecture is based on a modular approach allowing exchangeability and benchmarking using the output of individual trackers, whereas the fusion algorithm gives a solution to the track management problem and the coverage of wide perception areas. The test case is LATERAL SAFE sensor configuration, which monitors the rear and lateral areas of the vehicle. Results show that with the given approach the system is able to maintain the ID of all objects in transition (an object enters a sensor's FOV) and blind areas (no sensor coverage)

The Preventive and Active Safety Applications project (PReVENT), contributes to the safety goals of the European Commission (EC). PReVENT addresses the function fields of safe speed and safe following, lateral support, intersection safety and protection of vulnerable road users and collision mitigation in order to cover the field of active safety. The majority of these functions are characterized by using perception strategies based on multi sensor platforms and multi-sensor data fusion. ProFusion as cross-functional activity has the responsibility to streamline the multi sensor data fusion in the functional field activities. This paper presents several aspects of the research work conducted in ProFusion2 (PF2). For the covering abstract see ITRD E134653.

thereisa strongbelief thattheimprovement of preventive safetyapplications and theextension of their operative rangewillbeachieved bythedeployment ofmultiple sensors withwidefields ofview(FOV). Thepapercontributes to thesolution oftheproblemandintroduces distributed sensor datafusionarchitectures- and algorithms foran efficient deployment of multiple sensorsthatgiveredundantor complementary information forthemovingobjects. The proposed fusion architecture isbasedona modularapproach allowing exchangeability andbenchmarking using theoutputof individual trackers, whereas thefusion algorithm gives asolution tothetrackmanagementproblemandthecoverage ofwide perception areas.The testcaseisLATERAL SAFE sensor configuration, whichmonitors therearandlateral areasofthe vehicle. Results showthatwiththegiven approach thesystem is abletomaintain theID ofallobjects intransition (anobject enters asensor's FOV)andblind areas(nosensorcoverage). IndexTermsDataFusion, Dataassociation, ADAS,lane change, lateral collision warning

This paper focuses on the solution of the problem of (onboard moving vehicles) multiple sensor data fusion systems. The proposed application uses distributed architectures that operate with sensors or sensor systems and give redundant or complementary information for moving objects. This architecture ensures a modular approach allowing exchangeability and benchmarking using the output of individual trackers, whereas the fusion algorithm gives a solution to the track management problem and the coverage of wide perception areas. The test case is a multi-sensor configuration, which monitors the rear and lateral areas of traffic. Results from simulations and real data show that the given approach allows maintenance of the ID of objects and recognition of the vehicle environment with acceptable rates of false alarm and misses

In Europe, a considerable part of lives lost in traffic accidents is due to inappropriate vehicle speed or headway. Excessive speed is one of the major causes of accidents on European roads, responsible for one-third of all road accidents. SASPENCE, as part of the EU founded integrated project PReVENT, is developing and evaluating an innovative system able to perform the reliable and comfortable safe speed and safe distance concept, which helps the driver to avoid dangerous situations related to excessive speed or too small headway. In this paper, the high precise reconstruction of the road geometry is discussed. This information about the road geometry is needed in the project to be able to calculate the risk of a given scenario and given speed. A sensor fusion method is proposed to combine information from a vision sensor, a radar system, digital maps, a GPS sensor and odometric sensors to one common description of the road geometry in front of the vehicle. Thereby, the sensor information is used to localize the own vehicle relative to the map data. The system has been tested using real data and the results are shown in the paper

The introduction of new advanced driver assistance systems and active safety systems opens new possibilities when it comes to accident prevention. With the introduction of multiple systems there is a possibility to create synergies, but also a risk of introducing conflicts. By keeping future integration possibilities in mind already in the early system design phase a flexible and intuitive system can be achieved. For the covering abstract see ITRD E134653.