Abstract
Before highly automated vehicles (HAV) are the mainstream, partially automated vehicles (PAV) become prevalent. PAV are able to lateral and longitudinal vehicle control tasks in some circumstances, but a driver is always in charge of the driving. Mitigating and reducing the number of accidents is considered one of the most important benefits of HAV, but PAV can also enhance road safety. This study aims to evaluate potential safety effects of PAV by analysing three vehicle technologies’ possibilities to prevent real-world fatal passenger car crashes. In this study, PAV are determined as vehicles equipped with lane keeping assistance (LKA), automatic emergency braking (AEB), and adaptive cruise control (ACC). A crash-by-crash method enables evaluating each crash individually considering the actual restrictions and operational conditions of the systems.
The analysed data includes in-depth case studies on 506 fatal passenger car crashes in 2014-2016 in Finland. PAV could potentially have prevented 29% of these fatal passenger car crashes, and could deliver an annual crash cost reduction of 151 million euros. The potential to prevent crashes differs between the systems. LKA could potentially have prevented 27% of single-vehicle and head-on crashes while AEB and ACC could potentially have prevented 41% of rear-end, intersection and pedestrian crashes. As driver’s role for safe driving is crucial in PAV, driver-related risks are major obstacles for LKA’s operation. The system cannot operate due to driver’s attack of illness and in intendedly caused crashes as the driver should keep hands on the steering wheel and the system can be bypassed. The deficient visibility of lane markings such as the lack of lane markings or snow-covered lane markings is another obstacle for LKA’s operation. Excessive vehicle speed is the most typical obstacle, which would prevent the operation of AEB and ACC. Making these systems mandatory and by default always turned on in vehicles, the identified safety potential could be reached. The study also discusses possible paths to even greater safety potential by deploying advanced systems and developing the road infrastructure.
The analysed data includes in-depth case studies on 506 fatal passenger car crashes in 2014-2016 in Finland. PAV could potentially have prevented 29% of these fatal passenger car crashes, and could deliver an annual crash cost reduction of 151 million euros. The potential to prevent crashes differs between the systems. LKA could potentially have prevented 27% of single-vehicle and head-on crashes while AEB and ACC could potentially have prevented 41% of rear-end, intersection and pedestrian crashes. As driver’s role for safe driving is crucial in PAV, driver-related risks are major obstacles for LKA’s operation. The system cannot operate due to driver’s attack of illness and in intendedly caused crashes as the driver should keep hands on the steering wheel and the system can be bypassed. The deficient visibility of lane markings such as the lack of lane markings or snow-covered lane markings is another obstacle for LKA’s operation. Excessive vehicle speed is the most typical obstacle, which would prevent the operation of AEB and ACC. Making these systems mandatory and by default always turned on in vehicles, the identified safety potential could be reached. The study also discusses possible paths to even greater safety potential by deploying advanced systems and developing the road infrastructure.
| Original language | English |
|---|---|
| Title of host publication | ETC Conference Papers 2018 |
| Publisher | AET Association for European Transport |
| Publication status | Published - 2018 |
| Publication type | D3 Professional conference proceedings |
| Event | European Transport Conference - Dublin, Ireland Duration: 10 Oct 2018 → 12 Oct 2018 |
Publication series
| Name | European Transport Conference |
|---|---|
| ISSN (Electronic) | 2313-1853 |
Conference
| Conference | European Transport Conference |
|---|---|
| Country/Territory | Ireland |
| City | Dublin |
| Period | 10/10/18 → 12/10/18 |