Self-driving cars, or autonomous vehicles were once viewed as a futuristic idea that could possibly happen. Over the past 10 years this idea has turned into a real thing. With different companies now testing self-driving cars on the roads the thought of being able to own one of these cars is now a soon to be reality. I believe that with today’s technologies these cars could be a way to save lives lost from driver’s errors, and also a way to drive without having to pay attention to the road. The question I wanted to answer through my research was to find out actually how safe these vehicles are, and how they work.

Many people feel worried about these cars being on the road, but this is mainly due to the fact they are not educated on how these cars actually function and work. People feel as if a robot is driving the car which makes them feel uneasy, but actually the technology is very advanced. Uber’s version of the self-driving car has a LiDAR mounted on the top of the car which is used to create a 3D map of the area as the car drives by putting out 1.4 million laser points a second to have an accurate map for the car to follow(Han). LiDAR stands for Light Detection and Ranging, this is a remote sensing method that uses light in the form of a pulsed laser to measure distances to the Earth (Ocean Service). The car also features LiDAR modules on the front, side, and rear of the car to detect objects in hard to see spots. The car also has 20 different cameras on it to check for pedestrians, cars, or any other obstacle that could be in the way. The car uses a colored camera so that the map is in color the so car can detect traffic light changes as well (Han). Uber’s version of the self-driving car is just one of the many different companies trying and testing these vehicles in efforts to make these cars an everyday site. 

The technology used for self-driving vehicles is one that’s taken many years to get to the point where these vehicles could even get on the road. With advances in technologies such as digital mapping, improved LiDAR, and machine learning developers have been able to create software which is capable of driving a car and keeping its occupants safe (Urmson). Many different companies and groups have been trying to develop software and have the right technology to produce a self-driving car. One example is in 2009 Google started its program to design a self-driving car (Urmson). Since then Google has logged nearly 1,000,000 miles of autonomous driving on its self-driving vehicle (Lefrance). Having driven this many miles with no driver is a huge accomplishment, and a steppingstone in reaching the final goal of having these cars on the road like regular cars. Google’s version of the self-driving car is different from many of the major automotive companies in the industry. “Google is developing vehicles to be fully self-driving, requiring a rider only to tell the vehicle where to go , whereas the automotive companies are primarily focused on delivering advanced driver assistance systems that require the driver to remain in the steering loop” (Urmson). 

Developers of these vehicles hope that by creating a software to replicate a driver they can also create a safer driver. By logging hundreds of thousands of miles these vehicles developers can learn driving patterns for human drivers and can learn how to deal with these patterns. The software is so advanced these vehicles can react to drivers stopping suddenly, a driver switching over into its lane, and many other situations in which the driver is forced to react. Being that it is a computer there is a chance there could be accidents, but these software’s would know how to react to the accident to avoid injury to its occupants or others (Lafrance). This software could help reduce the number of vehicle accidents per year, and also reduce the number of deaths due to driver error. As humans we naturally make mistakes, developers are trying to make it where this software does not make mistakes. Over the course of the 1,000,000 miles googles autonomous car has driven it has only been in 16 accidents and none of which have been the fault of googles vehicle. This number is proof that these cars are much safer than having a human behind the wheel of a motor vehicle (Lafrance). 

When asked the question “How many lives can self-driving cars save?” researchers looked at the number of car crashes, and also the number of fatalities by car crashes, as well as how well the technology on these vehicles really is. In the United States in 2013 there were 32,719 fatalities by auto accidents (Lafrance). This is a large number, but a considerably lower number compared to the some 60,000 fatalities from auto accidents in the 1970’s (Lafrance). Researchers estimate that driverless cars could, by midcentury, reduce traffic fatalities by up to 90 percent. Which means that, using the number of fatalities in 2013 as a baseline, self-driving cars could save 29,447 lives a year (Lafrance). Looking at this in a long term view these vehicles could save nearly 300,000 lives in a decade. 

Many times as drivers we find ourselves paying attention to our phones, the radio, other people, and the things around us. These distractions are what cause car accidents, but also deaths. With these cars having a system scanning for cars and other objects these vehicles are always aware of their surroundings and the vehicles around them. By having a vehicle controlled by technology drivers could feel safer and not have to worry about making an error and getting into an accident.

One factor that has held back many companies in the recent past was the cost of the technology needed for the vehicles. Google’s self-driving car program is making strides at bringing these vehicles to streets around the country with some of their new innovations. Over the past decade the cost of a LiDAR scanning system which the vehicles use to process their surroundings has dropped more than 90% from the price of $70,000 per unit (LeBeau). Not only has the price dropped dramatically, but developers say that their software work better now than they did a year ago (LeBeau). This with the added price fall is great for people who would like to purchase or have access to these vehicles.  This drastic price drop is a giant stepping stone for putting these cars on the road and available to consumers (LeBeau). 

Making the technology available to consumers is one a huge role in progressing these vehicles, but another major factor is the cost that consumers must pay to have access to one of these vehicles. With another major leader in this industry being Tesla some may feel as if these vehicles are way out of their price range due to their reputation of making luxury vehicles with modern technology. Besides cutting the cost of the technology needed Google has also started testing their systems on less expensive vehicles that an average consumer could afford. By doing so they are appealing to a larger group of people, and also potentially saving more lives that could be lost due to accidents. 

One step developers have taken to try and introduce self-driving vehicles to the public is developing and producing semiautonomous vehicles. These are vehicles which are capable of being put in a mode that allows the driver to remove his or her hands from the wheel, but from time to time they must have their hands on the wheel and control the vehicle (Del-Colle). Like autonomous vehicles these vehicles use cameras and sensors to scan the area around it and then a computer inside the vehicle analyzes what has been scanned to figure out where the vehicle is, and to understand how the road is going to be (Del-Colle). “By using a combination of sensors, radar, GPS, and cameras, these systems enable a car to interpret its surroundings and issue warnings or even brake the car if a collision with another vehicle or a pedestrian is imminent. As these systems become connected, our cars become smarter (Del-Colle)”. 

While semiautonomous vehicles are not fully self-driving they are a big step in the right direction. Many people prefer these vehicles over autonomous vehicles as they can still override the vehicles system to change line, brakes, make a turn, or for any other reason. Developers have recently been able to make the cars communicate with each other by using GPS transmitters and Wi-Fi transmitter’s (Del-Colle). By doing this the car is able to alert the other car when it brakes abruptly, is close to the car, or even in its blind spot (Del-Colle). 

Another important factor in having these vehicles on the road is state and nationwide laws. With autonomous and semiautonomous vehicles being a rather new thing many ask whether the laws of the road still apply to the driver, if the driver isn’t actually the person driving. This has caused lawmakers to take a look at their states policies and change them as self-driving cars make their way to the top of the automotive industry (Del-Colle). One example of a law that must be looked at is the law in the state of New York where the driver must keep a hand on the wheel at all times, but what if the vehicle their driving doesn’t need a hand left on the steering wheel (Del-Colle). As lawmakers adjust to the new automotive landscape approaching these companies must also adjust to the rules and laws in which the government decides to implement for self-driving vehicles (Del-Colle). 

In September of 2016 the federal government released its first set of rules regarding manufacturing and selling semiautonomous vehicles and self-driving vehicles (Hawkins). Under these new rules companies are required to provide information about their vehicle and its software to the government so it can be looked at to ensure that these vehicles will be safe to be used in public. The government wants to know how these vehicles function, how they record data, as well as how they will react to situations on the road (Hawkins). Another regulation the government implanted in its new set of rules for all the states to use as a guideline to ensure there are not some states that allow these vehicles and some that would not (Hawkins). These new rules also address how to inspect a self-driving vehicle and also how to inspect them (Hawkins). In the end the government has the final say whether or not these vehicles can be produced or allowed on the road, they can also recall a vehicle if they feel it is not safe enough to be on the road with other drivers (Hawkins). With the implantation of these rules it is a clear sign that self-driving cars are happening right before our eyes. This is the first set of rules that address the driver as the computer system rather than a human (Hawkins).  

Though there are many positives to self-driving cars there still is a risk something could go wrong. Although there have not been any fatalities due to these vehicles it is only a matter of time before it happens. This accident could be the incident that sets back the production of these vehicles, but it could also help improve these vehicles so they can react to situations accordingly to keep everyone safe. Since these vehicles are ran by computers there is always a risk that the software could malfunction, or that something in the computer could mess up and cause an error (Agrwal). Over the million miles drove google only feels partially responsible for one accident, and a minor accident at that. The accident was could by the system misjudging the distance between sandbags in the road and a city bus (Argwal). This can be accredited to the fact that computers depth perception is not the same as a human, therefor the computer must process the information faster than it can at sometimes and this was one of those times. This accident was a minor one in which companies can learn from this error and find a solution to fix this problem (Argwal). 

It is seemingly impossible to accurately judge how much safer these vehicles are than a regular driver, but statistically speaking these vehicles are on path to save hundreds of thousands of lives each year. As of now it is difficult to accurately tell how much safer these vehicles are than human driven vehicles (Bhuiyan). Once these vehicles are on the road and acquire more data it will be easier to accurately tell how safe these vehicles are. Due to this many people are skeptic of having a computer be in control of their safety. Developers continue to test these vehicles in hopes of developing a software capable of driving without human control. Testing is currently being done in multiple major cities around the country in order to prove that these cars are actually as safe as they are claimed to be. The question at hand now is how safe is safe enough (Bhuiyan)? This is a question that lawmakers and developers are struggling to find a correct answer for. With many companies claiming to have an autonomous vehicle produced and available by 2021 lawmakers must make a decision on what they consider to be safe enough (Bhuiyan). 

I feel that the test and research is not wrong, these vehicles can and will be safer than a vehicle driven by human drivers. There are countless times that drivers error is the reason for an auto incident, these vehicles have proven they might make mistakes but less than humans would. Others may argue that these vehicles can’t prove themselves in just a few million miles but this is an rather large amount of miles to have driven without a driver. It’s almost certain that if you would have made humans drive that number of miles and compared the number of accidents human driven cars would be responsible for more than self-driving vehicles. It’s only a matter of time before these vehicles will be a common site on your local highways, interstates, and streets. By allowing self-driving vehicles to be on the roadways I believe that this would cause a decrease in traffic accidents, in damage to property by automobiles, amount of money spent on healthcare due to auto accidents, and most importantly the number of fatalities caused by auto accidents. Through many test I feel these developers have proven that these vehicles deserve their spot on the road, and sooner than later. 
