Angelos Amditis, ICCS, firstname.lastname@example.org Evangelia Portouli, ICCS, email@example.com
Timing and duration of tests
2012 – 2013
Location(s) of tests
Technical assessment Königsdorf, Frankfurt, Orbassano, Lommel, Hällered, Vårgårda, Ehra test track, Helmond
User assessment tests Aachen, Turin, Aachen, Hällered, Porrino
Continuous Support [CS]: This function intervenes with an additional steering torque when there is an object in the blind spot or in case of unintended lane departure. It warns the driver in case of rear end collision risk, VRUs, excessive speed, traffic rule violation. Curve Speed Control [CSC]: It warns the driver if speed is too high for an upcoming Curve. If ACC is also active, this function sets the ACC speed to a level which is safe for the curve. Enhanced Dynamic Pass Predictor [eDPP]: Informs the driver that overtaking is not recommended due to an upcoming path or due to oncoming vehicles which are detected by V2V communication or radar. Safe Cruise [SC]: This function reacts on changes in the traffic configuration when following a vehicle, for example it brakes softly when another vehicle cuts-in and identifies this vehicle as the vehicle to follow, it decelerates slightly when the driver is not attending the road. Lane Change Collision Avoidance [LCCA]: This function may intervene via automatic steering. Oncoming Vehicle Collision Avoidance/Mitigation [OVCA]: It attempts to keep the vehicle in its own lance through steering when an oncoming vehicle is detected. Rear End Collision Avoidance [RECA]: This function may intervene via steering and braking the vehicle. Side Impact Avoidance [SIA]: This function may intervene via automatic steering. Run-off Road Prevention [RoRP]: This function uses counter steering torque to avoid road departure and get the vehicle back into the lane. Collision Mitigation System (CMS): This function may intervene via braking and steering. The aim is to reduce impact speed and to optimise the impact orientation. Emergency Steer Assist (ESA): This function supports the driver when he/she tries to avoid an imminent collision by steering. The function monitors the environment and adjusts the Electric Power Steering to stabilise the vehicle.
Level of automation tested
SAE0 and SAE1
Tested use cases
The technical tests involved situations where the function should be activated. The same is true for the user-related assessment tests in driving simulators. Tests with instrumented vehicles in test tracks involved semi-realistic scenarios. Tests with instrumented vehicles on public roads involved drivers using the functions in a naturalistic way.
Tested transport system
Passenger cars Truck
Purposes of testing
- Assessment of driver behaviour (human, vehicle) / road user behaviour
- Assessment of impacts on safety (short-term, long-term)
- Assessment of user (driver, traveller, etc.) acceptance, usability, take up, etc.
- Technical assessment, proof of concept (incl. vehicle, background support systems such as communication)
Definition of baseline
In user-related assessment tests where a baseline condition was used, this was when users were driving without the function.
In tests where a baseline condition was used, the same persons drove with the function and without the function
Method of testing
Controlled field tests, Driving simulator, Interviews, Naturalistic field tests, Questionnaires, Simulation, Test track
Test fleet, participants and environment
Number and make of vehicles
BMW 535i (enhanced Dynamic Pass Predictor) Lancia Delta (Continuous Support) Ford Focus (Continuous Support, Curve Speed Control, Lane Change Collision Avoidance, Rear End Collision Avoidance, Side Impact Avoidance, Run-off Road Prevention) Volvo S60 (Continuous Support, Safe Cruise, Lane Change Collision Avoidance, Run-off Road Prevention) Volvo Truck FH-480 6*2 (Oncoming Vehicle Collision Avoidance/Mitigation, Rear End Collision Avoidance, Side Impact Avoidance, Run-off Road Prevention) VW Golf GTI VI (Collision Mitigation System) VW Passat B7 (Emergency Steer Assist)
Description and number of participants/drivers
User-related assessment tests Aachen (Dynamic Pass Predictor): Video presentation of function to 17 drivers from the public in two groups Turin (Continuous Support, Curve Speed Warning): Field tests on public roads, driving without and with the function, 24 persons Aachen (Continuous Support): Test track, 19 participants Aachen, (Side Impact Avoidance, Rear End Collision Avoidance): Test track, semi-realistic scenarios, 25 participants Hällered (Safe Cruise including Speed Support): Test track, 10 persons Hällered (Side Impact Avoidance, Rear End Collision Avoidance): Test track, 10 participants VTEC (Rear End Collision Avoidance, Side Impact Avoidance, Run-Off Road Prevention and Oncoming Vehicle Collision Avoidance): Fixed-base truck simulator (without-with the function), 31 participants Porrino (Collision Mitigation System): Simulator study, 59 drivers in 3 groups: control group (driving without the function), experimental group 1 (driving with the function but without knowing how the function worked) and experimental group 2 (driving with function and knowing how it worked). Porino (Emergency Steer Assist): Simulator study, 68 persons, in two different situations: highway and rural roads
Tested environment and facilities
The technical assessment tests were conducted mainly on test tracks (Orbassano, Lommel, Ehra, Frankfurt, Hällered), in laboratory environment (VeHIL) and in public roads (Königsdorf, Vårgårda). Various tools, like stationary and moving targets, were used, and reference measurements systems were used (normal GPS, DGPS, CANgos, etc.).
The user-related assessment was based on video presentations (Aachen), on tests with instrumented vehicles in test tracks (Aachen, Hällered), on field tests with instrumented vehicles on public roads (Turin) and on tests in driving simulators for critical situations (VTEC fixed-base truck simulator, Porrino). Some critical functions for which a simulator did not exist were studied in test track settings in safe, but semi-realistic scenarios (the test persons were aware of the functions). The field-trials with an instrumented vehicle involved observing drivers while they were using the function in a naturalistic way in real traffic conditions and comparing their behaviour when driving without the function. Observations were done by two observers, riding along in the car with the driver. The tests were complemented by questionnaires or interviews. Questionnaires included: subjective workload, experienced effects of the function, usefulness and satisfaction of the function, perceived benefits, usability, willingness to have and pay.
Legal and ethical aspects
Duration of testing
Technical assessment tests Königsdorf, Public road, Head on collisions, 23 test runs Frankfurt, Test track, Rear-end collisions, Unparking vehicle, Collisions with VRUs, 119 test runs Orbassano, Test track, Rear-end collisions, Head on collisions, Lane change collisions, Collisions with VRUs, Unintended lane departure accidents, Excessive speed accidents, traffic rule violations, 261 test runs Lommel, Test track, Rear-end collisions , Lane change collisions, Cross traffic collisions, Unintended lane departure accidents, Excessive speed accidents, 308 test runs Hällered, Test track, Rear-end collisions, Head on collisions, Lane change collisions, Unintended lane departure accidents, Excessive speed accidents, Traffic rule violations, 133 test runs Vårgårda public road: E6 motorway Gothenburg-Uddevalla, Rear-end collisions, Lane change collisions, Unintended lane departure accidents, 27 test runs Ehra, Test track, Helmond laboratory environment (VeHIL), Rear-end collisions, Cross traffic collisions, 37 test runs
Input parameters and assumptions of simulation tests
VTEC (Rear End Collision Avoidance, Side Impact Avoidance, Run-Off Road Prevention and Oncoming Vehicle Collision Avoidance) study with a fixed-base truck simulator The participants first drove a main route, comprising of one highway section and one rural road section, during which the scenarios related to each function would occur. Afterwards, the participants were informed about the functions and drove the function-related scenarios once again. For the main drive, a between-groups design was employed, with 16 participants driving with the function active (treatment) and the other 15 drove without function (baseline), to allow for measuring the effect the functions would have on participants’ reactions. All participants drove the concluding test drive of the function scenarios.
Porrino (Collision Mitigation System): Simulator study 59 drivers in 3 groups: control group (driving without the function), experimental group 1 (driving with the function but without knowing how the function worked) and experimental group 2 (driving with function and knowing how it worked). They drove approaching an intersection where another vehicle crossed from the right side where there was a high probability of an imminent collision.
Porino (Emergency Steer Assist): Simulator study 68 persons, in two different situations: highway and rural roads In both situations, drivers approached a stand still target object. The sample was divided in 4 groups: Control Rural Group, Control Highway Group, Experimental Rural Group, and Experimental Highway Group. Each group was composed by two variables; type of route (rural or highway) and test condition (experimental with function and control no function).
- Map data
- Two observers, one studied standardised variables such as speed behaviour, yielding behaviour, lane changes and interaction with other road users, the second carried out ‘‘free observations’’ such as conflicts, communication and special events.
Specifications for the data sources
Sensor data, specify: position, time, longitudinal and lateral distance, speed, acceleration, function status, warning type and level, driver manoeuvre, relative distance between vehicles
Key Performance Indicators (KPIs)
The following were calculated for the safety impact assessment: The speed difference between vehicles. The deceleration of the other vehicle. The lateral distance between vehicles. The lateral velocity of the host vehicle. The heading angle of the host vehicle with respect to the road (i.e., the angle of its velocity vector).
Situational data available
Subjective data collected
The data collected after the user-related assessment tests included: subjective workload, users’ opinions on the experienced effects of the function, their opinions on the usefulness of and satisfaction by the function, perceived benefits, their opinions about the usability of the function, their willingness to have and pay for the function.
Issues that affected the impact assessment
The safety impact analysis was based on the results of the technical and user-related assessment. Real rear-end and run-off road accidents of the GIDAS database were re-simulated with the function under study. Then, the changes for the different injury levels were calculated. Finally, the results were up-scaled to EU-level.
The results of the re-simulation of rear end collisions showed that RECA and ESA could prevent potentially up to 77% of the analysed rear-end accidents. The findings refer to ideal effects and do not considering unintended (adverse) driver reactions. The CMS function achieved the highest fraction of mitigated accidents. The function has also the potential to avoid accidents in quite a high number of cases (42%). In the re-simulation of road departure accidents, a complete avoidance was achieved in 5-64% of all cases.
The results of re-simulation of real accident were upscaled on European level. Resulting from an accident analysis both in GIDAS and CARE, adequate samples were matched and processed in the raking procedure. In GIDAS, the share of relevant car occupants on all car occupant casualties was determined. In CARE, the car occupants in accidents with either one vehicle or two or more vehicles and the conditions under which they occur, were selected. The raking procedure was used to account for differences between the conditions under which GIDAS accidents occur and the European accident scenarios as reported in CARE. The result of this work is a table with an estimation for relevant accidents on EU level for each scenario that is addressed by the interactIVe functions.
Deliverable D7.5 Impact Assessment of Developed Applications (link: https://lup.lub.lu.se/search/publication/4467672)