Validation Results
Författare
Summary, in English
This report describes the validation of the SASPENCE driver support system using an evaluation method which has been developed within the scope of the SAfe SPEed and safe distaNCE (SASPENCE) project. SASPENCE aims to develop a driver support system that focuses on keeping a safe speed and safe heading distance in relation to other road users and road infrastructure. The system warns the driver to adapt to the situation and consequently the driver actions are influencing the effectiveness of SASPENCE. The Human Machine Interface design has therefore been a key aspect in the development of the system. The developed evaluation method is capable of dealing with a large variation of circumstances (including driver behaviour) and in all steps of the evaluation (apart from straightforward hardware tests) the driver is considered.
The evaluation method consists of simulation-based functional tests, on-road functional tests, on-road subjective tests, and traffic impact assessment using simulation.
The simulation-based functional tests (using a driver model) have been undertaken in three stages using a probabilistic method to cope with the large variation of circumstances. In the first stage a large number of Monte Carlo simulations were performed to assess the effectiveness of the basic SASPENCE algorithm, and to identify the most critical situations. In the second stage, Software-in-the-Loop simulations have been used to reduce the number of critical situations in order to arrive at a practical number of Hardware-in-the-Loop (HIL) experiments. The HIL experiments confirmed the effectiveness of the results of Monte Carlo simulations, and they additionally provided estimates for dependability parameters (i.e. Safety, Reliability, Appropriateness and Timeliness). On-road functional tests addressed conditions that can not be realised in HIL experiments; however, Appropriateness and Timeliness could not be determined.
On-road subjective tests have been carried out in Spain and Italy by selecting 20 drivers at each location and letting each driver run a 50 km course with and without the SASPENCE system activated. Both experimental vehicles were equipped with systems for monitoring the driving session, recording camera images and vehicle motion data. Moreover, the Wiener Fahrprobe in-car observation method was used for the analysis of driver behaviour during the tests. From the users’ point of view, the system has been positively rated as being useful and well accepted by drivers, to an extent that the majority would be willing to pay up to 500 Euro to have it installed in their own vehicle. Also, from the experts’ point of view, the system was rated positively as for instance the number of conflicts and the number of events related to individual behavioural aspects decreased. The recorded vehicle data also confirmed the effectiveness of the system to reduce speed and increase heading distance.
The Traffic impact assessment simulations show that the impact of SASPENCE is most apparent for high congestion levels where the positive (safety) effect is an increasing Time-to-Collision. However, the large reduction in the mean speed adversely affects traffic throughput.
The evaluation method consists of simulation-based functional tests, on-road functional tests, on-road subjective tests, and traffic impact assessment using simulation.
The simulation-based functional tests (using a driver model) have been undertaken in three stages using a probabilistic method to cope with the large variation of circumstances. In the first stage a large number of Monte Carlo simulations were performed to assess the effectiveness of the basic SASPENCE algorithm, and to identify the most critical situations. In the second stage, Software-in-the-Loop simulations have been used to reduce the number of critical situations in order to arrive at a practical number of Hardware-in-the-Loop (HIL) experiments. The HIL experiments confirmed the effectiveness of the results of Monte Carlo simulations, and they additionally provided estimates for dependability parameters (i.e. Safety, Reliability, Appropriateness and Timeliness). On-road functional tests addressed conditions that can not be realised in HIL experiments; however, Appropriateness and Timeliness could not be determined.
On-road subjective tests have been carried out in Spain and Italy by selecting 20 drivers at each location and letting each driver run a 50 km course with and without the SASPENCE system activated. Both experimental vehicles were equipped with systems for monitoring the driving session, recording camera images and vehicle motion data. Moreover, the Wiener Fahrprobe in-car observation method was used for the analysis of driver behaviour during the tests. From the users’ point of view, the system has been positively rated as being useful and well accepted by drivers, to an extent that the majority would be willing to pay up to 500 Euro to have it installed in their own vehicle. Also, from the experts’ point of view, the system was rated positively as for instance the number of conflicts and the number of events related to individual behavioural aspects decreased. The recorded vehicle data also confirmed the effectiveness of the system to reduce speed and increase heading distance.
The Traffic impact assessment simulations show that the impact of SASPENCE is most apparent for high congestion levels where the positive (safety) effect is an increasing Time-to-Collision. However, the large reduction in the mean speed adversely affects traffic throughput.
Avdelning/ar
Publiceringsår
2008
Språk
Engelska
Dokumenttyp
Rapport
Förlag
Deliverable 20.70 SASPENCE, PReVENT Consortium
Ämne
- Infrastructure Engineering
Status
Published