Realising Advanced Incident Detection on European Roads
01/11/2011 - 31/07/2013
RAIDER is a research project of the cross-border funded joint research programme “ENR2011 MOBILITY – Getting the most out of Intelligent Infrastructure”, which is a trans-national joint research programme that was initiated by “ERA-NET ROAD II – Coordination and Implementation of Road Research in Europe” (ENR2), a Coordination Action in the 7th Framework Programme of the EC. The funding partners of this cross-border funded Joint Research Programme are the National Road Administrations (NRA) of Belgium, Switzerland, Germany, Netherlands, Norway and United Kingdom.
Incident detection is an essential capability for Road Authorities to manage their road networks and adequately respond to incidents. Issues with the quality of detection, such as a high false alarm rate, delays in detection, or inaccurate location of incidents, directly impact the operations. High quality of data enables faster resolution of incidents and pro-active measures to avoid traffic disturbance. Significant investments may be required to improve the detection quality with additional roadside detection systems. At the same time, new developments in nomadic devices and in-vehicle systems, and in third party services, may provide solutions that improve both incident detection quality and reduce the costs for National Road Authorities.
RAIDER is a research project aiming to improve incident detection systems in terms of the detection quality and costs of the system and individual components such as sensors, data fusion and incident detection algorithms.
The quality of existing systems and their components are reviewed against the user needs and requirements from National Road Authorities. Generic specifications are derived for sensors, data fusion and incident detection algorithms, based on which recommendations are made for necessary improvements of existing systems.
A second path is also followed to assess the feasibility, costs and benefits of innovations in the near future in roadside sensors, nomadic devices, and in-vehicle devices. Alternative system concepts for incident detection are explored to incorporate and fuse data from mobile devices. The quality criteria from the first approach also apply here as constraints on the quality of the devices.
Experts from the National Road Authorities are consulted early in the project to select their most pressing issues as reference cases for research. These could be for example; hard shoulder monitoring, and the detection of accidents and traffic congestion. Stakeholder consultations and two workshops with experts from the Road Authorities are planned to ensure the necessary input for the project and the dissemination and exploitation of the results.
The project does not attempt to focus on any particular incident detection systems or Road Authority., because custom solutions are not easily transferable to other systems and Road Authorities. Instead the project follows a methodology to deliver results that apply to all Road Authorities:
Generic specification of the quality requirements for sensors, data fusion and incident detection algorithms, in function of the user needs and requirements on incident detection functionality and performance.
Feasibility assessment of innovative roadside sensors, nomadic devices and in-vehicle systems to improve incident detection.
These results can be applied by Road Authorities to assess the performance of their existing incident detection systems, identify necessary improvements, and define specifications for tendering. The specifications and feasibility assessment also provide the necessary input for cost benefit analyses of such improvements.
Detailed Description of the Project
Road Authorities need high quality traffic data for effective traffic control and management of their road networks. Traffic data is necessary to maintain an accurate and up to date picture of the traffic situation, to detect the state of congestion or adverse weather conditions for example, and to set traffic measures to inform drivers, adjust speed limits, open extra lanes, or redirect traffic.
Traffic flow and safety can be further improved when traffic measures can be taken proactively to avoid the escalation of incidents into accidents or traffic jams. An incident is a disruption of traffic flow or safety, such as a slow vehicle, a broken down vehicle, an accident, or the build-up of traffic congestion near a ramp or weaving area. To respond adequately to incidents, high quality traffic data and incident detection are needed.
The quality of traffic data can be expressed in criteria for the functionality and performance of incident detection systems. Functionality criteria are for example the accuracy of the incident location and the time delay between occurrence and detection. If the location of a stranded vehicle could be detected within seconds, a lane can be closed immediately to avoid accidents, minimize traffic disturbance, and instruct emergency services accurately. On the other hand, if only lower quality traffic data would be available, such as inductive loop data, only the resulting traffic jam can be detected in high density traffic and minutes after the incident. In low density traffic, the incident may never be detected at all, while the safety risk may be even larger due to the higher vehicle speeds. Performance criteria are for example the false alarm rate and reliability of the incident detection systems. This will impact the operational usability of the system, for example when too many false alarms are generated, or when too many incidents are missed.
In the current situation, differences can be observed across Europe in traffic and incident management strategies, road design, detection technologies, variable message signs, driver behaviour, and incident and accident rates. Road Authorities have implemented their specific incident detection solutions, providing traffic data of different quality levels, enabling different approaches to incident detection and control.
Usually, Road Authorities operate roadside-based detection systems and message signs, which are relatively expensive in ownership and in maintenance. New technologies, such as innovative roadside sensors, nomadic devices and in-vehicle devices, may provide higher quality data to improve incident detection. These technologies enable the tracking of individual vehicles and dramatically increase the temporal and spatial resolution of incident detection. In-vehicle systems can also provide information about road surface conditions and short term incident warnings.
The functionality, performance, costs and benefits of these systems are significantly different, and also differ with the current situation. Innovative roadside sensors still require a significant investment by the Road Authorities. Mobile devices, on the other hand, may require smaller investments in communication systems or services from third parties. The accuracy, delay times and reliability may be less and require minimum penetration rate of those devices and complex data fusion.
Interestingly, mobile devices are based on international standards that may also facilitate the harmonisation of incident detection in the longer term, even though the migration from the current situation may be different for different Road Authorities. The quality criteria for incident detection equally hold for the new technologies. This concept provides the basis for this project.
RAIDER follows a methodology to deliver results that are applicable to all European Road Authorities:
• Stakeholders are consulted to assess their operational experiences and issues with existing systems, and to define user needs and requirements for incident detection systems (now and in the future) as a reference for evaluating new technologies. Experts from the National Road Authorities are consulted actively as stakeholders to ensure outcomes are fit-for-purpose by selecting the top priority types of incidents and most pressing issues as reference cases for research.
• Quality criteria for the functionality and performance of incident detection systems and their major components; sensors, data fusion and incident detection algorithms. These quality criteria are derived from the user needs and requirements on incident detection functionality and performance.
• System concepts for incorporating new technologies for road side systems and for utilizing in-vehicle systems and nomadic devices.
• Estimating costs and benefits as well as the feasibility of innovative of proposed incident detection systems.
• Generic specification for incident detection systems, and their major components (sensors, data fusion and incident detection algorithms) in terms of quality requirements.
• Obstacles to implementation of potential systems and recommendations for mitigation.
Key tasks and deliverables:
• Stakeholder workshop to collect and consolidate the user needs and requirements.
• Stakeholder workshop to disseminate the generic specifications for incident detection systems.
• Intermediate working documents on user needs and requirements, data quality criteria, and generic specifications.
• Summary report with final results and recommendations for facilitation of implementation of potential solutions.
Benefits to the Road Authorities
This project will identify anticipated future developments and international best practice in incident detection to support Road Authorities’ network management duties. This will include:
• User needs and requirements – a consolidation of the stakeholders’ operational and technological experiences with existing systems and demands in the near future.
• Identification of new and emerging incident detection technologies with indication of system performance in terms of strengths an limitations for application in the near future
• Feasibility assessment of incident detection technologies – mapping of the requirements against appropriate incident detection technologies’ performance, and guidance for demonstration of the business case (case benefit analysis).
Overview of Work Packages
To coordinate the project as a whole and manage the process to deliver results to time, cost and quality. With project team members in different countries, stakeholders of different nationalities, research material in different languages and highly interrelated work packages, efficient and effective project management is crucial to success.
To obtain the user needs and requirements on incident detection and management from stakeholders.
To identify the major experts as stakeholder for incident detection to represent the ERANET partners, NRAs and operators, and to organise stakeholder consultations.
To consolidate for the types of top priority incidents with stakeholders.
To organise workshops to consolidate user needs and requirements, and to disseminate the specifications to stakeholders.
To assess and analyse the stakeholders’ user needs and requirements, operational and performance issues, from WP2, Task 2.1
To define generalized quality criteria for incident detection systems and major components.
To review and select future technologies relevant for the selected types of incident detection
To review new and near-future technologies for incident detection capabilities.
To review existing technologies with alternative primary purposes for incident detection capabilities.
To assess obstacles and issues related to using these alternative sources of incident detection data.
To provide a feasibility assessment for potential incident detection solutions.
WP5: Final results and recommendations (TNO)
To summarize the results and recommendations for the stakeholders in a final report.
To assess feedback from the second stakeholder workshop and update D2.1 and D4.1
Toon Beeks, TNO
Oude Waalsdorperweg 63
2597 AK The Hague - The Netherlands
firstname.lastname@example.org,Tel. 31 888 668 007