|
Abstract
|
Any current discussion pertaining to the automotive industry will undoubtedly include the state of autonomous vehicle (AV) development. The eventual deployment of fully autonomous vehicles could bring many societal benefits including less roadway congestion and increased traffic flow, reduced emission of greenhouse gasses and more personal time for commuters. Additionally, there could be more transportation options for those with mobility challenges.
The most important benefit could be a significant reduction of traffic crashes and injuries. In 2018, more than 36,000 people died in motor vehicle crashes. While the road to autonomy will have bumps along the way, AAA believes that fully autonomous vehicles have the potential to save thousands of lives every year. Government and safety experts estimate that fully autonomous vehicles could prevent up to 80 percent of today's crashes.
While the concept of autonomous or "self-driving" cars was described as far back as the 1939 World's Fair, actual research and development has only begun to make significant progress over the last decade.
In 2018, AAA evaluated four vehicles equipped with active driving assistance (ADA) systems. In common language, these systems assist the driver with vehicle acceleration, braking and steering. SAE International defines various levels of driving automation within SAE standard J3016 and active driving assistance (ADA) systems currently available within the United States are considered to be Level 2 systems per this standard. To build upon work AAA conducted in 2018, the performance of five additional vehicles equipped with ADA systems was characterized.
Research Questions:
1. How do vehicles equipped with ADA systems perform during scenarios reasonably encountered in highway driving situations?
- Evaluated via closed-course testing
2. How do vehicles equipped with ADA systems perform during naturalistic highway driving?
- Evaluated on public highways and interstates
Key Findings:
1. In general, evaluated ADA systems performed according to expectations derived from the owner's manuals during closed-course evaluations.
- On fresh pavement with well-defined lane markers, all systems consistently traversed the testing lane with some lateral offset relative to the lane center.
- In aggregate, evaluated ADA systems made contact with a simulated disabled vehicle 66% of the time.
- For a simulated stop-and-go scenario, none of the evaluated ADA systems made contact with a lead vehicle for all tested deceleration rates.
2. Test drivers reached a consensus that over approximately 4,000 miles of total driving distance, combining longitudinal and lateral control into a single system did not significantly enhance the driving experience for the majority of encountered scenarios. On roadways utilized for evaluation:
- Lane-keeping events accounted for 73% of noted events.
- A total of 521 events were noted among all evaluated ADA systems. On average, this resulted in a noteworthy event approximately every eight miles.
Research conducted by the AAA Foundation for Traffic Safety evaluated the potential impacts of Level 2 automation equipped vehicles on driving behavior. A general conclusion of the work originated from a 12-month field evaluation of 50 participant-owned vehicles. Since "test" vehicles were owned and operated prior to the study, it can be assumed participants already had some degree of familiarity with the system. As a result, this data may be indicative of long-term adaptation and behavior effects. In general, this evaluation showed a greater impact on various driver behavioral measures when ADA systems were active compared to when these systems were not active but available. Eye-glance metrics indicated drivers had their eyes off-road and on non-driving-related tasks more frequently and for longer durations when ADA systems were active compared to any other system status.
This observation is troubling when combined with findings of this work in both closed-course and naturalistic environments. In both environments, clear examples of critical performance demerits that may result in a serious collision were observed.
Nothing contained within this report is intended to directly imply evaluated ADA systems are inherently unsafe. In totality, this work is intended to illustrate potential system inadequacies in the context of scenarios realistically encountered in a naturalistic environment. Additionally, all test drivers expressed some degree of distrust in each evaluated system and questioned the overall usefulness of integrating adaptive cruise and lane-keeping functionalities.
This work emphasizes the need for additional refinement of ADAS systems as a whole and for the automotive industry to collaboratively consider how current iterations of ADAS technology align with driver perception and behavior. Multiple facets relating to system programming, edge-case scenario design considerations, human/machine interactions and driver training should be considered in order to most effectively design and promote ADAS systems that provide the most benefit to the driving public. |