Are self-driven cars safer than human-driven cars?

Self-driven vehicles are no longer a fantasy. The first test samples of autonomous cars have already gone beyond the closed landfills. They may appear in the street and interstates of the world much earlier than is commonly believed. Several self-driven cars are already being used on the roads of Europe and America, such as Tesla with its Tesla version 7.0, Google Self-Driving Car Project, Ubuntu using self-driving cars for passenger travels. In 2018, the Baidu company is about to launch a pilotless vehicles. In general, there is no doubt that in the near future self-driving vehicles will occupy a firm place in human lives.

The information above is just in proportion as it approaches the future, fashionable technology creates more and more questions. For instance, previously, scientists have found that drivers want self-driving cars were able to “squeeze” pedestrians. They say that ethical questions are important, but in an emergency, a vehicle needs to save the lives of the driver and passengers not taking into account the price of any of the random victims. In addition, people’s confidence in the new technology also is not immense. For all generations, except for the very young, personal control over the car is of extreme importance due to several reasons. According to Nagy, people aged 16 to 54 are not ready to entrust the management of unmanned technology in full, but are willing to consider a hybrid variant.

Though such vehicles are considered to be safer than the usual transport, but there are some difficulties in adapting to the driving style of ordinary drivers, weather conditions and the quality of the road topping. The research in the behavior of vehicles on the road is very important for the further introduction of innovations. Therefore, the goal of the current study is to study three common problems of self-driving cars – weather conditions, traffic situation and, most importantly, safety by means of qualitative and quantitative research based on exiting literature, reports and studies on the matter. This study is of valuable significance since these three problems are the most discussed issues connected to self-driving cars when starting a research. Therefore, there is a need to accumulate and assess the existing information in order to create a solid viewpoint on the problem.

Literature Review

As mentioned, people now live in a technological age. Due to the emergence of various technological innovations and advancements it is almost no longer surprising that the industry was able to come up with a self-driving car concept. What’s more is that this self-driving or autonomous car is something that won’t be imaginary in the distant future; rather, there are many researchers and experts who are positive that the car and its benefits and advantages would be enjoyed by thousands or even millions of people. In this review of literature, the following key points are shown: (1) self-driving cars could be the solution to traffic congestion, parking problems, and other related issues; (2) self-driving cars could also be the answer to people who are not able or are not allowed to drive but would like to move to different places; (3) self-driving cars could be programmed to outperform human drivers, given that they may be programmed to become well-equipped in certain situations and contexts. In addition to that, there is also a possibility that self-driving cars could also become more sensitive to traffic signs and rules; and (4) self-driving cars could be the future of the transportation system of the world.

Sajan and Ray, for instance, believes that with the rising cases of truck-related accidents on the road, resorting to a systems perspective using the driving simulators, which is also is a term for self-driving cars is the best way to go. In their literature review, they tried to point out that by improving technology and incorporating it to the transportation systems, the current systems (which they described as outdated and no longer appropriate) will be improved and thus minimize multiple levels and cases of fatalities and losses. 

Traffic Regulations. Aside from traffic congestion and frequent road accidents, Ferreira focused on a car-related issue that was overlooked by most of the researchers and experts cited in this literature review: parking problems. Given that there are so many car-owners in the present times, it is isn’t unimaginable that aside from heavy traffic, there also are times when once these cars have reached the intended destination, there may be some problems when it comes to where they will have to leave their cars. According to the researchers, parking could also contribute to traffic congestion and will definitely have some implications on urban landscape; in this regard, they explained that the emergence of fully automated and mechanical parking systems were caused by the reduction of space needed by cars to park. This is also seconded by Fagnant and Kockelman who believed that self-driven cars could potentially become smaller than conventional ones, thus leaving more spaces for car parking.

Technology Advancements of Self-Driving cars. Butakov and Ioannou also mentioned that autonomous cars are one of the major technological breakthroughs due to technology: (1) Efficiency: through the autopilot mode of the cars, there is a high possibility that the cars could perform or drive better than an actual human driver, given that autopilots could be programmed to notice more things and become more sensitive in other factors than humans can process. (2) Limitations: through self-driving cars, there are safe and reliable operations within specific limits. (3) Responsibility: they believe that the transition could be safe and reliable since the driver knows what the autopilot could do and could not do. (4) Dynamics: the performance of the self-driving car should be predicted by human drivers.

Safety of Autonomous Cars. However, despite the evidence-supported arguments made by most of the researchers, claims like the ones provided by Butakov and Ioannou could somehow become too far from being realistic. Given that technology could somehow achieve this, still, the benefits they explained are not guaranteed to work all the time. As discussed by Kamath et al, there are potential reaction delays that these autonomous cars might possess, given that there are some contexts that the autonomous car was not programmed to become aware of. Although optimistic about the potentials of self-driving cars, Butakov and Ioannou are also open to the possibilities that there are some contexts and situations wherein self-driving cars could prove to be almost useless.

In their same paper, they have discussed that vehicular factors such as systems failures, roadway factors such as when there are large potholes or roadway debris that the GPS couldn’t notify the car about, and other sudden crashes caused by sudden human or animal movements could prove to be one of the biggest challenges faced by self-driving cars. In addition, these challenges are not mere challenges to overcome; these could actually cost lives if overlooked. In addition to that, both researchers also revealed that with the growing population and the ever-constant changes within societies, it is understood that there will be times when places will change – thus a need for map improvement and updating. Given that not all roads remain constant, it could prove to be a difficult task to update the maps and road information frequently, thus leading to potential accidents if the passengers of the self-driving car overlook these.

Self-driving cars in weather. In addition to that, Fagnant and Kockelman, though also positive about the benefits and advantages that self-driving cars could give to the users, still, they are determined about the way it could revolutionize the way people travel. In an instance, they believe that creating rules and regulations to govern these cars could prove to be a hard task. In addition to that, the authors claim that environmental problems such as snowy days, rainy days, etc., could be a problem for self-driving cars given that these cars are not yet programmed to withstand such terrible climate and weather The real test for the self-driving car is wet and slippery road, and most importantly – poor visibility. The designers and engineers claim that in five to ten years, the unmanned vehicle will be truly weatherproof. However, the thing is that in bad weather conditions, a human being behind the wheel has learned to overcome hard weather conditions, but computer is not capable of it yet. 

Given these advantages, it is noticeable that while the researchers in this paper were quite positive about the advantage and the advancement of self-driving cars, there is no guarantee that these cars could eliminate the traffic and other car-related accidents on the road. In fact, most of the researchers cited in this paper believed that certain factors such as: (1) Vehicular problems which includes systems failures and other problems on the machine or control of the car; (2) Constant need for updating the maps and road information; and (3) updating of the rules and regulations to cover the self-driving car could prove to be one of the biggest challenges and issues when these cars are effectively approved. In addition to that, it is also apparent that despite their programs and systems perspectives there is still a need for drivers and users to become well-acquainted and informed about the self-driving car in order to avoid accidents. Nonetheless, there is a possibility that self-driving cars could reduce accidents, thus raising the percentage of its safeness against human-driven cars.

Method

The methods to be used during the current research is the review of reports and researchers relating to self-driving cars of two most widespread and discussed manufacturers – Google and Tesla in order to evaluate their safety, ability to withstand the weather condition and to adapt to constantly changing traffic in urban conditions. The research allow synthetizing the data of self-driving cars in relation to the reliability of self-driving cars in urban conditions. 

The research is divided into three general research questions – safety, weather adaptability and traffic adaptability. The period of analysis is to concentrate on 2015 and 2016 reporting and research results in order to accrue the most recent updates of the software used. Additionally, only google and Tesla self-driving cars are to researched since as the most popular cars and one of the most researched and developed manufacturers they present monthly and weekly reports on the condition of their product design and testing, which creates the reliable source of information for the current study. 

The three questions are to be researched separately using different research methods – qualitative for weather, quantitative for traffic adaptability and mixed for safety. In qualitative research the technical specifications and the incidence of road accidents, failures and bags and their relation to weather is to be used. Traffic adaptability is to be assessed from the data on maneuverability, GPS tracking and ability to change the route will be investigated using a ten-point scale. Finally, safety is to be assessed from the standpoint of accidents and the decision the software takes in relation to the driver’s, passengers and pedestrian safety. The validity of data will checked from standpoint of the publication source and credits of the authors.

Results

Safety and reliability of self-driven cars are of an extreme importance. These are the issues causing the majority of arguments considering self-driving cars among public and professionals. Therefore, this study was aimed at investigation of three core issues related to self-driving cars – weather, traffic and safety. The results show that automatically controlled car can be named reliable in terms of traffic and safety (maneuverability and ability to avoid crashes). However, the weather conditions are still an important problem to solve.

Self-driving Cars and Weather. As it is known, rain is a big problem for unmanned vehicles. Slippery road surface and reducing visibility makes the automatic on-board systems literally go “blind.” Numerous tests at the site showed that the self-driving vehicles in heavy rain behave very unpredictably and can create emergencies and accidents.

The research of the weather influence onto self-driving car safety was concentrated around three weather issues – rain and snow, fog and icing, The data on weather effect was collected for two car involving the prevalence of accidents or failure due to the weather conditions. Basing on Google and Tesla report for 2015 and 2016 as well as statics for road accidents due to bad weather conditions in the US, the results stored in table below were obtained.

 

 It can be seen that three weather conditions explored cause more accidents with ordinary cars. However, it should be noted that there are thousands of ordinary cars in the US and weather is the cause for only 110 accidents in total and only 4% of all accidents statistics. Considering Tesla and Google there are only several Tesla self-driving cars on the US roads and 2 types of Google cars. Therefore, the prevalence of accident in self-driving cars is higher, that in common ones. 

Self-Driving Cars and Traffic. Researchers estimated that after the mass release of unmanned cars on the roads and ubiquitous car sharing, about 80% of the cars will be taken off the streets. Additionally, automatic system of cars. Lights and traffic jams detection are said to decrease the number of cars on the roads, reduce the “road snakes” and ease the parking. In order to investigate that, the traffic indicators for both cars were evaluated.

Using qualitative method of reports evaluation and since both cars communicate with each other almost immediately, their response to the presence of other cars on the road is more flexible than that of common cars. Unfortunately, it is impossible to investigate the issue as whole due to a small number of self-driving cars on the roads. However, both reports studied provide a strong qualitative data on the maneuverability of cars and their communicational skills. Therefore, both cars are reliable enough.

Safety. For this question, the mixed method was used in evaluating the two performance report. First, the prevalence of accidents for both cars was calculated.

Afterwards, the accidents were evaluated for the cause, and those relating only to automatic driving system were detected. Therefore, for Google car automation-related accidents comprised 5 for two years, and for Tesla – 2 for 2 years. The overwhelming majority of accidents happened as the result of mistakes made by other road users Again, the limitation and weakness of this particular part of study is the small number of self-driving cars on the roads. However, relating to percentage data, one can see that automation-related accidents for both self-driving cars is less than 1%. Therefore, it can be said that self-driving cars are rather safe in terms of accidents prevalence.

Discussion

Basing on the research results, one can say that self-driving car in some aspects can be safer than human-driven cars. Considering safety, which is the greatest concern for self-driving cars the result have shown that automatic car are more reliable than common ones, since the car system completely eliminates the human mistake. Each year about 600,000 people in the U.S. die as a result of road accidents. This is more than 2.1% of all deaths and is comparable with the number of deaths caused by major “killers” such as malaria and tuberculosis. Road traffic injuries is the second leading cause of death among people aged 5-25 years. However, relating on the study results and reports data, self-driving vehicles will exclude abuse speed since it is the main factor contributing to road traffic injuries in most countries. Young men in particular tend not to comply with the relevant speed limit. Reducing the average traffic speed by 1 km/h leads to a decrease in the number of fatal accidents by 4-5%. Additionally, the cars will exclude drunken driving since the car does not allow a person to get behind the wheel himself, if he is in a state of intoxication. Drinking and driving increases both the likelihood of accidents and severity of injuries.

Self-driving cars can also solve traffic problem – under condition that everybody will use automated cars, instead of common one. Self-driven cars can communicate with each other and are connected to the constantly updating GPS tracking systems, detecting traffic jams, repairing and other every minute. Therefore, every single self-driven car is capable of detecting the most suitable way, which is going to reduce the number of traffic jams. The results of the study confirm the theoretical data.

Weather remains the greatest problem for such kind of cars. Snow, rain and so on are difficult not only for the human drivers. For example, the snow prevents the automatic detection system to read traffic lanes and “instinctively” following them. Similarly, the falling rain and snow can seriously interfere with laser sensor readings of the surroundings. According to Google allegedly the weather is one of the major causes of accidents and failures of self-driving vehicles. Mercedes-Benz Company is developing the solution to this problem – they intend to increase in the number of sensors and their improvement. For example, the orientation not by the road lanes, but on roadsides or row of trees along the road, if the strip is covered with snow.

Conclusion

The conducted study has shown that self-driving cars can be safer that human driving ones. However, there are several conditions and limitations to such a conclusion. First, the number of self-driving cars on the roads does not allow making comprehensive research on their safety. The great emphasis in the automated car technology is made on the ability of car communicate with each other and on GPS tracking and constant updating. Since common car are not able to do the same, any study, including the current one, is limited. The literature in the field are also controversial since the technology is only being developed there is a lot of gaps and weaknesses to be eliminated in the future. The authors mainly rely on successful results of studies, rather than on concerns. However, the small number of cars, which use the self-driving technology, limits all studies reviewed. Therefore, future research is to fill the gap of individual evaluation of the technology, but researching it in relation to other cars, including common vehicles used today.

Related essays