Periodic Reporting for period 2 - TransSec (Autonomous emergency manoeuvring and movement monitoring for road transport security)
Période du rapport: 2019-08-01 au 2021-01-31
The TransSec project addresses a new danger in European countries, which is the increasing number of attacks conducted with road vehicles. Today, the latest truck safety systems can prevent some of the most serious kinds of road accidents. They fail however in case of collisions intentionally caused by the driver. TransSec aims to initiate the development of supplementing security modules.
Project objective: Development and evaluation of systems to be used by trucks for secure road transport of (dangerous) goods. Preventing trucks and transport of goods to be misused for other purposes that may cause serious damages.
Specific objectives:
• Precise vehicle positioning and navigation on road (lane) and off road
• Vehicle movement monitoring for dangerous goods with critical area alarm/eCall
• Vehicle communication security for critical information exchange
• Onboard pre-crash environment detection of vulnerable objects on and off road
• Non-defeatable autonomous emergency manoeuvring for crash prevention on and off road
The TransSec project work has been finished at the end of January 2021.
The prototypes developed and tested within TransSec can prevent typical collisions – whether intended or by accident – that state-of-the-art safety systems cannot prevent. They provide a substantial improvement of road safety and road transport security.
Project objective: Development and evaluation of systems to be used by trucks for secure road transport of (dangerous) goods. Preventing trucks and transport of goods to be misused for other purposes that may cause serious damages.
Specific objectives:
• Precise vehicle positioning and navigation on road (lane) and off road
• Vehicle movement monitoring for dangerous goods with critical area alarm/eCall
• Vehicle communication security for critical information exchange
• Onboard pre-crash environment detection of vulnerable objects on and off road
• Non-defeatable autonomous emergency manoeuvring for crash prevention on and off road
The TransSec project work has been finished at the end of January 2021.
The prototypes developed and tested within TransSec can prevent typical collisions – whether intended or by accident – that state-of-the-art safety systems cannot prevent. They provide a substantial improvement of road safety and road transport security.
1. Precise vehicle positioning and navigation using GNSS systems incl. Galileo.
An analysis on positioning quality requirements with the TransSec use cases have been performed including several newer GNSS products with improved positioning algorithms, correction services, and 4 GNSS satellite systems: GPS, Galileo, GLONASS and BeiDou. A Galileo Positioning System has been developed. Authentication of GNSS Signals, additional measures against sensor manipulation and spoofing protection have been implemented. Several tests have been carried out with a truck in real road environments. Requirements on positioning accuracy were defined as 0.5 m (95 %) for truck movement monitoring. Reliable positioning accuracy on decimeter level currently is only achievable with (paid) positioning services. Remarkable was performance of Galileo standard alone positioning. Performance was comparable and in some cases even better than the GPS system, even though the Galileo system currently has fewer operational satellites than GPS system.
2. Road and Environment Map
Road and environment map uses a commercial navigation map enhanced with additional information. The availability of map geometry, topology and attributes has been investigated and the assessment of geometric map accuracies has been carried out using GNSS-based reference trajectories. Map-matching algorithms and electronic horizon providers have been analysed and extended to support TransSec use cases with focus on illegal path identification. Map-Aiding has been developed to replace classical map-matching algorithm, on-/off-road detection and wrong way detection have been implemented.
3. Environment Object Detection
Relevant sensors and hardware have been identified and specified. A clear emphasis was put on LiDAR, which is not a state-of-the art sensor for trucks. The developed system allows to monitor traffic scenes. The scene is virtually reconstructed by the system in 3D using v-SLAM algorithms. The system detects and classifies both static (e.g. lane lines) and dynamic objects (e.g. pedestrians, vehicles) surrounding the ego vehicle in real-time. A combination of camera-based vision technique and LiDAR-based object detection was finally included in the Local Dynamic Map. The proposed sensor combination provides a high spatial accuracy and higher robustness compared to current systems and is well beyond state-of-the-art object detection for trucks.
4. Vehicle Movement Monitoring
Vehicle movement monitoring provides two security modules, a map-based movement monitoring with critical area alarm and a driving situation analysis and classification based on dynamic environment sensing. The movement monitoring uses driving restrictions to identify illegal paths, critical areas and launches an alarm using eCall when a rule is violated. Risk situations are defined and can be detected. Trajectory-prediction is implemented and allows to predict collision probabilities. The developed algorithms showed promising results in simulations and on-road tests in WP8. They are suitable to reduce potential collisions during driving.
5. V-2-X Communication
Risk communication to entities (V2X) included vehicle to vehicle (V2V), vehicle to pedestrian (V2P) and also vehicle to infrastructure (V2I) communications. System architecture, relevant techniques, standards and security architecture are defined. Security and privacy issues have been addressed. Five demonstrators (V2V, V2P, V2I, V2X Security, and eCall) have been realized. A complete V2X communications chain with accompanied V2X cloud service was established, which can ensure Secure Risk Communication from vehicle to any entity (V2X) in the context of the TransSec approach and scenarios.
6. Autonomous Emergency Manoeuvring
The restrictions and capabilities of the current truck safety functions have been identified for further development. The results are prototype systems for lateral and longitudinal emergency manoeuvres, including a non-defeatable speed limiter. The system has the potential to improve road transport safety and security especially for vulnerable traffic participants.
7. Vehicle Integration and Testing
The testing activities covered three main demonstrations:
• Test and demo of precise vehicle positioning, including positioning accuracy evaluation of several GNSS receivers with different GNSS positioning/correction strategies using different GNSS satellite constellations, with a special focus on Galileo system.
• Test and demo of object detection and situation prediction, which showed the capabilities and advantages of combining camera and Lidar based environment perception modules
• “Collision Prevention Road Testing” combining several tests. It included an update to the GNSS Multi Sensor Integration and evaluation. The results of the collision warning system prototype and the emergency manoeuvre prototypes have been tested using the test truck. The tests showed lane-level accuracy and successful detection of spoofing attempts.
An analysis on positioning quality requirements with the TransSec use cases have been performed including several newer GNSS products with improved positioning algorithms, correction services, and 4 GNSS satellite systems: GPS, Galileo, GLONASS and BeiDou. A Galileo Positioning System has been developed. Authentication of GNSS Signals, additional measures against sensor manipulation and spoofing protection have been implemented. Several tests have been carried out with a truck in real road environments. Requirements on positioning accuracy were defined as 0.5 m (95 %) for truck movement monitoring. Reliable positioning accuracy on decimeter level currently is only achievable with (paid) positioning services. Remarkable was performance of Galileo standard alone positioning. Performance was comparable and in some cases even better than the GPS system, even though the Galileo system currently has fewer operational satellites than GPS system.
2. Road and Environment Map
Road and environment map uses a commercial navigation map enhanced with additional information. The availability of map geometry, topology and attributes has been investigated and the assessment of geometric map accuracies has been carried out using GNSS-based reference trajectories. Map-matching algorithms and electronic horizon providers have been analysed and extended to support TransSec use cases with focus on illegal path identification. Map-Aiding has been developed to replace classical map-matching algorithm, on-/off-road detection and wrong way detection have been implemented.
3. Environment Object Detection
Relevant sensors and hardware have been identified and specified. A clear emphasis was put on LiDAR, which is not a state-of-the art sensor for trucks. The developed system allows to monitor traffic scenes. The scene is virtually reconstructed by the system in 3D using v-SLAM algorithms. The system detects and classifies both static (e.g. lane lines) and dynamic objects (e.g. pedestrians, vehicles) surrounding the ego vehicle in real-time. A combination of camera-based vision technique and LiDAR-based object detection was finally included in the Local Dynamic Map. The proposed sensor combination provides a high spatial accuracy and higher robustness compared to current systems and is well beyond state-of-the-art object detection for trucks.
4. Vehicle Movement Monitoring
Vehicle movement monitoring provides two security modules, a map-based movement monitoring with critical area alarm and a driving situation analysis and classification based on dynamic environment sensing. The movement monitoring uses driving restrictions to identify illegal paths, critical areas and launches an alarm using eCall when a rule is violated. Risk situations are defined and can be detected. Trajectory-prediction is implemented and allows to predict collision probabilities. The developed algorithms showed promising results in simulations and on-road tests in WP8. They are suitable to reduce potential collisions during driving.
5. V-2-X Communication
Risk communication to entities (V2X) included vehicle to vehicle (V2V), vehicle to pedestrian (V2P) and also vehicle to infrastructure (V2I) communications. System architecture, relevant techniques, standards and security architecture are defined. Security and privacy issues have been addressed. Five demonstrators (V2V, V2P, V2I, V2X Security, and eCall) have been realized. A complete V2X communications chain with accompanied V2X cloud service was established, which can ensure Secure Risk Communication from vehicle to any entity (V2X) in the context of the TransSec approach and scenarios.
6. Autonomous Emergency Manoeuvring
The restrictions and capabilities of the current truck safety functions have been identified for further development. The results are prototype systems for lateral and longitudinal emergency manoeuvres, including a non-defeatable speed limiter. The system has the potential to improve road transport safety and security especially for vulnerable traffic participants.
7. Vehicle Integration and Testing
The testing activities covered three main demonstrations:
• Test and demo of precise vehicle positioning, including positioning accuracy evaluation of several GNSS receivers with different GNSS positioning/correction strategies using different GNSS satellite constellations, with a special focus on Galileo system.
• Test and demo of object detection and situation prediction, which showed the capabilities and advantages of combining camera and Lidar based environment perception modules
• “Collision Prevention Road Testing” combining several tests. It included an update to the GNSS Multi Sensor Integration and evaluation. The results of the collision warning system prototype and the emergency manoeuvre prototypes have been tested using the test truck. The tests showed lane-level accuracy and successful detection of spoofing attempts.
Precise vehicle positioning, implementation and fusing of LiDAR technology with other sensors into a local dynamic map, V2X communication incl. eCall when entering critical zones and automated collision prevention covering intentional collisions are all well beyond state-of-the-art in trucking.
The higher accuracy and knowledge of the surrounding allows for greatly increased ADAS functionalities to prevent collisions or at least reduce damages caused by road traffic. Road safety and security in public areas is improved.
The higher accuracy and knowledge of the surrounding allows for greatly increased ADAS functionalities to prevent collisions or at least reduce damages caused by road traffic. Road safety and security in public areas is improved.