Skip to main content

Advanced Radar Tracking and Classification for Enhanced Road Safety

Final Report Summary - ARTRAC (Advanced Radar Tracking and Classification for Enhanced Road SafetyAdvanced Radar Tracking and Classification for Enhanced Road Safety)

Executive Summary:
The “Advanced Radar TRAcking and Classification for enhanced road safety” (ARTRAC) research project contributed to the Road Safety Programme 2011-2020 concerning improved safety for vulnerable road users (VRU). The main objective was to develop a novel 24 GHz radar sensor to protect VRUs in general and reduce the number of fatalities in particular.
The European commission described the ambitious goal in 2001 to reduce the number of fatalities by 50% in a ten years time frame. This requirement has not completely met so far, however a significant progress could be observed when the number of fatalities has been reduced by 25% in the last 10 years due to several technical assistance procedures and techniques. For the period from 2011 to 2020 again the goal is to reduce the number of fatalities on European roads by 50% [1]. The technical challenge of ARTRAC was to develop a 24 GHz radar sensor product in high volume series cars, as they are already applied in high-end limousines. Current sensor systems (cameras and radar sensors) show already excellent performance in target detection for driver assistance and comfort system functionality. However, the new ARTRAC radar sensor has also a target recognition feature integrated. This functionality distinguishes between radar echo signals from pedestrians, bicyclists, vehicles and static objects and is able to classify these different objects. Thereby as a reaction upon a critical situation an extended functionality like automatic deceleration, steering recommendation and driver warning is performed. The system developed consists of a 24 GHz radar sensor, a risk assessment and a vehicle control procedure. This safety system has a high potential to be launched because the majority of all components belong already to standard equipment in series cars.
There are a number of systems to detect pedestrians in the market. However, these combine radar and optical sensors and thus come with a high cost and therefore are available in middle class to high-end cars only. At the end of the project, the ARTRAC system was the first to use radar only allowing implementation also in compact cars. The demonstration vehicles used were a Volkswagen GTI and a Fiat 500L.
The ultimate highlight of the project was the Final Event held at the proving ground of Volkswagen in Ehra near Wolfsburg. About 50 persons, many of whom were from the production sector in addition to research oriented participant were presented with an overview of the work in the project and subsequently had the chance to take a test ride on the two demonstration cars. A video of the event and the test rides is available on the project web site.

Project Context and Objectives:
The European Union launched a programme in 2001, “Halving the number of road accident victims in the European Union by 2010: A shared responsibility”. Despite numerous safety initiatives and activities, it is clear that this ambitious target is still far from being met. There are a number of reasons for that. One of them is that the deployment of Intelligent Vehicle Safety Systems (IVSS) has not been as fast was anticipated in the 1990s. This, in turn, is due to a number of factors. Besides user awareness and outreach issues, a major problem has been the performance of the Advanced Driver Assistance Systems (ADAS) in all driving situations and the high cost of the applications. When drivers had a possibility to try out different ADAS, they have almost exclusively approved of them. However, they were willing to invest only modest amounts of money in advanced safety systems.
The work towards realising the ARTRAC vision is a logical continuation of activities from the late 1980s such as Prometheus and the DRIVE I project (Framework Programme 2), where incident detection based on environment perception and information systems for transport & traffic, guidance and cooperation between the vehicles were ranked among the core R&D areas in the years to come.
Today, safety technologies such as ESC (Electronic Stability Control) have shown outstanding capabilities for supporting the driver in hazardous situations. Despite their effectiveness, currently available lateral and longitudinal control systems are typically implemented as independent functions. This results in multiple expensive sensors and unnecessary redundancy, limiting their scope to premium-class vehicles. Vehicle safety technology is not democratic today.
The large Integrated Project PReVENT by major European safety technology developers recently showed the feasibility of a concept called the “electronic safety zone” surrounding a vehicle, which has considerable potential for enhanced road safety. Here again, the need for further developments was in improving the performance in all scenarios, integrating the different functions, and providing commercially feasible solutions. In this context, in addition to further enhancing ADAS performance and reducing the number of sensors used, ARTRAC needs to address the issue of making safety systems affordable and accessible to all drivers.
Traffic fatality risk has been significantly reduced in recent years for those inside vehicles through the use of initially passive measures such as seat belts, stiff passenger cell, and more recently active safety systems that incorporate sensors such as airbags for protection and anti-lock braking systems (ABS) and electronic stability control (ESC). Important to this development has been the development of mass production low-cost sensors making such safety systems affordable in the mass market vehicle range.
Much of the above has also improved safety of vulnerable road users (VRUs), but it is recognised that more can be done. The attention of vehicle manufacturers and regulatory authorities is now much more focused on the protection of road users outside the vehicle, particularly in urban scenarios where all traveller groups interact. This is turn builds on initiatives such as the so-called two phases of pedestrian safety in the European Union with the goal to halve the number of pedestrian fatalities from what they were in 2001 by 2010.
Research in the field of pedestrian protection has been undertaken for more than two decades. It is now being driven by parallel regulatory initiatives by the European Union and also globally: in 2002 pedestrian protection was introduced in the compendium of the UN regulatory forum of UN-WP29 (Working Party 29 of the United Nations). So far approaches have been to engineer those parts of the car that come into contact with a pedestrian in such a way as to minimise impact e.g. through specially adapted bonnets and bumpers. Other passive means include use of daytime running lamps to improve visibility. But in both cases it is the driver or the pedestrian respectively who has to take action. Sensing technologies, however, provide a means for automated collision avoidance. The challenge is to provide a reliable means of detection coupled with a fast reaction time for the vehicle.
Although there continues to be much research in this area, and especially on materials and the engineering of the vehicle body, this passive safety system approach is really not the ultimate solution: the vehicle should be prevented from hitting the pedestrian or cyclist in the first place. The goal of ARTRAC is to achieve a breakthrough in deployment of Intelligent Vehicles Safety Systems (IVSS) that will provide the function of collision avoidance as well as collision mitigation.
To achieve IVSS that will get market acceptance means in practice providing an extremely reliable detection system, a link to actuation of the vehicle to provide the appropriate response in a form that is affordable. Regulation can of course make safety features mandatory, but this is a slow process. If the aspirations of the European policies towards improving the road safety of all users are to be met quickly, then IVSS which can be easily integrated into all types of vehicles, perhaps even those already on the road, offers the best way forward.
The key element to achieve an Intelligent Vehicle Safety System which acts as a driver assistant to protect vulnerable road users is the obstacle detection system. The choice of technology and approach needs to take into account the requirements for achieving the low cost needed for mass market penetration of all vehicle classes, small size and robustness.
The consortium will develop a totally new safety system for the protection of vulnerable road users by means of advanced environment perception technology and active intervention by vehicle braking and steering. Decision strategies for active safety and driver-vehicle-interaction will be developed. The objective is significantly to improve current ADAS, and it requires major breakthroughs in system intelligence and decision making. In particular, three areas will be addressed by the project:
• New techniques for the dynamic prediction of a safe trajectory ahead.
• Decision strategies able to balance human and system interventions, keeping the driver in the control loop as far as possible.
• Advanced actuation system combining both emergency steering and braking.

More details may be found in the PDF attached.
Project Results:
The project could be divided into three consecutive milestones. The first milestone was the analysis of accident statistics, preceding projects and their activities and current sensor technology. From this, the requirements for the development of the radar sensor and the VRU safety system where derived and specified.
In the next phase, the sensor hardware and algorithms were developed, so were the on-board system interfaces and software. The second milestone was reached with the successful integration of the sensor into the demonstrator vehicles and first proof of the functionality.
In the last project phase, an in-depth testing and evaluation programme for all system parts has been specified and performed. The last project milestone was reached by finalizing the test runs and documenting the test results.

More details may be found in the PDF attached
Potential Impact:
The key result of the ARTRAC project is a safety system to protect vulnerable road users (VRUs) designed to be economically viable in the volume vehicle market. The safety system consists of both, actuators for controlling vehicle driving dynamics and the perception component for the vehicle’s surroundings. It was tested on two vehicles that pose the biggest hazard to VRUs in urban settings, namely compact cars.
However, the issue of market penetration is of major importance and this is very much an issue of cost. This has so far been the main barrier to the results of many previous funded projects bearing fruition in terms of widespread adoption of the technology, and it is only through widespread adoption that inroads into reducing road casualties will be achieved.
In turn, ARTRAC has supported the objectives of the Road Safety Programme 2011-2020 adopted by the Commission in July 2010. Here there is a commitment to ‘promoting the use of modern technology to make traffic safer’ (Objective 5) and to do more to ‘protect vulnerable road users’ (Objective 7). It is the goal of ARTRAC to achieve breakthrough systems in support of the latter by means of the former.

More detail in the PDF attached

List of Websites:
Project website: http://www.artrac.org
Contacts:
Hermann Rohling, Technische Universität Hamburg-Harburg
Email: rohling@tu-harburg.de

Axel Wegner, TuTech Innovation GmbH
Email: wegner@tutech.de

Project logo: