The extremely high number of registered road fatalities each year does not tend to diminish requiring urgently a change of direction.
Every year the estimated number of road traffic deaths around the world costantly increases. According to the World Health Organization (WHO), in 2016 around 1.35 million people lost their lives in traffic accidents. This is equivalent to nearly 3700 deaths per day and to an average mortality rate of 18.2 deceases / 100,000 population. Overall, road injuries are the 8th most common cause of death with more people killed than tuberculosis or HIV/AIDS. An even worse scenario emerges considering that they represent the leading death cause for children and people aged 5–29.
In a statistical summary published by NHTSA’s National Center for Statistics and Analysis, the U.S. Department of Transportation claimed that 94 per cent of fatal road crashes can be attributed to drivers. In particular, accidents are, starting from the most frequent cases, due to:
recognition errors such as driver's inattention, distraction or low surveillance;
decision errors, including speeding, wrong assumptions and illegal manoeuvres;
performance errors, e.g. poor directional control or overcompensation;
non-performance errors mainly associated with sleep.
Within this dramatic context, automated vehicles are standing out as a potential solution for preventing avoidable deaths and saving tens of thousands of lives by limiting human mistakes.
SAE levels
Some companies in the transportation sector, and in particular Tesla Inc., have already achieved important milestones in the automation of vehicles by applying Machine Learning (ML) techniques in the so-called Advanced Driver-Assistance Systems (ADAS). Nevertheless, the road towards the full driver replacement is still significantly long and full of challenges. To the current state, autonomous vehicles (AVs) are in the prototyping phase and can be consequently tested only in closed or controlled environments.
The Society of Automotive Engineers (SAE) outlined a new standard made up of six levels to universally classify the autonomous driving skills of a vehicle.
As visible from the image above, the rating depends on the amount of intervention and attention demanded to the driver.
Self driving advantages
Autonomous vehicles are expected in the future to considerably enhance the living standards and the safety of the transportation system. In 2015, Lutrell et al. argued that substituting 90% of the cars in the United States with AVs could result every year in 25000 fewer deaths and $200 billion savings.
Self-driving vehicles will enable millions of people with disabilities and visual impairments to autonomously travel from one place to another. The same argument can be easily extended to children or drunk people since AVs do not require driving.
In comparison with traditional vehicles, it is expectable that autonomous cars will drastically reduce gas and energy consumption, that in most of the cases is indeed caused by bad driving human behaviours such as sudden accelerations or brakings. Moreover, while travelling in the city centre and after dropping the passengers, AVs might eventually park themselves in a more peripheral and less onerous location.
In general, self-driving cars will also increase the efficiency and the productivity of the time spent on the journey given that passengers can carry out tasks other than steering. Thus, they embody as well an excellent alternative for people with demanding jobs that traditionally would encounter some difficulties throughout the trip.
It is also possible to release a worldwide patch update to correct a failure whenever is detected in the program that controls the car. Finally, driverless transportation implies a dramatic cost reduction in the delivery field not strictly requiring the presence of a human to perform the service.
Nevertheless, to pave the road for autonomous vehicles and guarantee the safety of the citizens, countries must properly and with perseverance maintain their infrastructure as well as define together new laws and rules.
References
World Health Organization. "Global status report on road safety 2018: Summary". No. WHO/NMH/NVI/18.20. World Health Organization, 2018.
Schwarting, Wilko, Javier Alonso-Mora, and Daniela Rus. "Planning and decision-making for autonomous vehicles." Annual Review of Control, Robotics, and Autonomous Systems (2018).
Rojas-Rueda, David, et al. "Autonomous vehicles and public health." Annual review of public health 41 (2020): 329-345.
Wiseman, Yair. "Autonomous vehicles." Encyclopedia of Information Science and Technology, Fifth Edition. IGI Global, 2021. 1-11.
Osman Ors, Ali. "The Role of Machine Learning in Autonomous Vehicles." www.electronicdesign.com, 3 Dec. 2020.
U.S. Department of Transportation. "Automated Driving System 2.0: A Vision For Safety." www.Nhtsa.Gov, 1 Sept. 2017.
Singh, Santokh. "Critical reasons for crashes investigated in the national motor vehicle crash causation survey". No. DOT HS 812 115. 2015.
US Department of Transportation. "Preparing for the Future of Transportation: Automated Vehicles 3.0." (2018).
Centers for Disease Control and Prevention, "Road Traffic Injuries and Deaths A Global Problem." 14 Dec. 2020, www.cdc.gov/injury/features/global-road-safety/index.html.
Luttrell, Kevin, Michael Weaver, and Mitchel Harris. "The effect of autonomous vehicles on trauma and health care." Journal of Trauma and Acute Care Surgery 79.4 (2015): 678-682.
Redmon, Joseph, and Ali Farhadi. "Yolov3: An incremental improvement." arXiv preprint arXiv:1804.02767 (2018).
Liu, Wei, et al. "Ssd: Single shot multibox detector." European conference on computer vision. Springer, Cham, 2016.
The images in the blog are either copyright free or designed from scratch. Some figures presented in this article are also created leveraging elements extracted from the following vector images:
https://www.freepik.com/vectors/car Car vector created by macrovector.
コメント