Post 7, Human Driving is Soon to be Obsolete

Introduction

Should humans be banned from driving? I don’t mean immediately, and I don’t mean to imply they should be replaced with baboon drivers. I mean, when self-driving car technology enables cars to drive themselves with a degree of safety that surpasses humans, should they then  become mandatory? It might seem like a weird question, because depending on who you ask, you could get a variety of answers ranging from, “Obviously yes,” to, “Obviously no,” to, “It doesn’t matter, cars are never going to be that good, at least never in our lifetime.” Of course, if you’re following current news on self-driving cars, you know that they very well might be that good in our lifetime. Let’s step back for a moment though, and cover the necessary background information. The most commonly used scale for categorizing the autonomy of cars was developed by the organization SAE International, and rates at a level from zero to five, zero being fully controlled by a human, five being fully autonomous. Level 3 is where you could first call something a “self-driving” car, and it’s where most autonomous cars are right now. Level 3 is automated, but requires a human at the wheel to take over if a situation arises that the car’s systems can’t resolve. Waymo announced in November 2017 that they are testing cars with employees in the backseat, rather than at the wheel(). This means that as of now, level 4 cars are being tested. The Senate Committee on Commerce, Science, and Transportation had a hearing in 2016 with statements from prominent representatives from industry discussing all aspects of self-driving cars to determine how policy should be handled going forward (Hands Off). They raised issues such as federal regulations being necessary to allow autonomous cars to traverse state lines without legal issue, and the need for rigorous cybersecurity on the part of industry leaders. It’s no question that self-driving cars are coming. And when they do, humans should be banned from driving. We’re going to have to overcome the hurdles of trust and perceived loss of freedom, but human drivers should be replaced by self-driving cars for reasons both of efficiency and safety.

automationrating.PNG

Safety

If human drivers are able to be outclassed in terms of safety by self-driving cars, they should be replaced, as at that point every driver replaced will help to save lives. It’s not only possible, but likely that self-driving cars will become better than humans. A survey conducted by the NHTSA found that approximately 94% of vehicular crashes were caused by human error (Singh). That 94% is what self-driving cars are capable of reducing. A car doesn’t get drowsy, doesn’t drink, doesn’t undergo road rage, and doesn’t text. It can see in 360 degrees around it and never stops paying attention. A simulation run by Mark Morando et al was conducted to determine the potential safety impact at both a signalised intersection and a roundabout of introducing autonomous vehicles. They found that, “for the signalised intersection, AVs reduce the number of conflicts by 20% to 65% with the AV penetration rates of between 50% and 100% (statistically significant at p < 0.05). For the roundabout, the number of conflicts is reduced by 29% to 64% with the 100% AV penetration rate (statistically significant at p < 0.05)” (Morando et al). While these are simulations, and practical, real world testing is necessary as well, they still demonstrate a substantial improvement. And real world testing is happening, Waymo alone has over 5 million total miles driven by autonomous vehicles (Waymo Reaches 5 Million Self- Driven Miles), and every collision is simulated repeatedly to learn from and prevent a similar incident (Hands Off, 13). This is essential to the safety of autonomous cars versus humans. An accident can be learned from by a company, and a fix can be applied to every car in its fleet. Human error is never going to be done away with so long as humans exist. Safety isn’t the only potential benefit of self-driving cars, though, they have the ability to improve efficiency as well.

 

Efficiency

Self driving cars can improve throughput of roads, reduce travel times, and improve fuel efficiency, and the effects are more pronounced the more autonomous cars there are. In fact, it’s been estimated that, “cooperative adaptive cruise control (CACC) deployed at 10 percent, 50 percent, and 90 percent market-penetration levels will increase lanes’ effective capacities by around 1 percent, 21 percent and 80 percent, respectively” (Preparing a Nation, 5). This goes to show that it decidedly matters that self-driving cars dominate the road for their benefits to be reaped. Raw throughput isn’t all that will improve, as if self-driving cars really do end up safer, traffic caused by crashes will also go down. The notion of parking will see perhaps the most radical change of all. After all, if a car can drive itself, it can park itself. This means that parking need not be distributed evenly across cities. Roman Zakharenko predicts the rise of “parking belts” in cities, composed of parking structures outside of downtown areas, where cars will park themselves after dropping riders off at their destinations. This leaves more space to be made use of in city centers. There are only benefits to be had from having more self-driving cars on the road, so why isn’t there unchecked public support for them? Well, self-driving cars are a radical change, and people don’t trust them.

peopledon'tlikethefuture.png

Trust

Critics point out the many risks and challenges self-driving cars will run into as they develop, from technical malfunctions to the risk of a network of cars being hacked as an act of terrorism, but people are dying right now as a result of driver error, so we have to come to terms with the imperfections inherent to self-driving cars if we’re to cast off the imperfections of manually operated cars. Perhaps the biggest hurdle to mandatory self-driving cars is trust. If people don’t trust them, they won’t allow policy to make them mandatory. An argument someone distrustful of self-driving cars might use is that self-driving cars won’t prevent people’s deaths, there is always a risk of failure in some form or another, so we shouldn’t trust self-driving cars. This line of thinking is the perfect solution fallacy, comparing imperfect self-driving cars to an imagined perfect solution, making self-driving cars look as though they’re failing to accomplish the goal of safety, even if they are able to accomplish better than our current safety. After all, they only have to be better than human drivers to be a worthwhile replacement, so it follows they only have to be better than human drivers for people to trust them more than human drivers. This doesn’t seem to be the case though. A study conducted in 1969 indicates that “the public is willing to accept ‘voluntary’ risks roughly 1000 times greater than ‘involuntary’ risks.” (Starr, 1237). In this case, voluntary and involuntary risks refer to risks that one has a degree of control over versus risks where one has no control respectively. This, then, applies to cars, with driving yourself in a car being a voluntary risk and allowing a self-driving car to drive you being an involuntary risk. This is a massive issue for public trust of self-driving cars. If it holds, then self-driving cars don’t have to accomplish fewer than 30,000 deaths annually, but closer to fewer than 30. All I can ask of people is to be aware of statistics, and be open to trusting technology if data proves we can. Of course, trust isn’t the only thing on people’s minds when asked to take their hands off the wheel.

Freedom

Self driving cars are not in any way a significant reduction of freedom from driving oneself, and driving should not be considered a freedom worth preserving anyway if it runs such a high risk of fatality.  It’s natural to envision the car being taken out of your direct control as a massive loss of freedom. In reality, though, you lose very little when you transfer from a manually operated car to an autonomous one. There already exist rules of the road for people’s safety. Stop lights, lane markers, speed limits, and safety requirements for cars already prohibit absolute freedom while driving. Public roads and the laws that govern them don’t exist primarily for an enjoyable driving experience, they exist to safely and efficiently deliver people to their destinations. Even if you find it unbearable to imagine never driving again, private roads and race courses will still exist for recreational driving, no matter how much market penetration self-driving cars achieve. Finally, consider the legal perspective. The supreme court has held that free speech, even considering its status as a constitutional right, does not extend to creating a clear and present danger, famously providing the example of  “shouting fire in a crowded theater” (Schenck v. United States). In driving, while there is no intent to harm, the act of getting behind the wheel naturally puts those around you at risk. The principle of freedom of oneself vs. life and limb of others still holds. I understand the potential joys of driving. America has a strong car culture, and people can feel liberated behind the wheel. But those feelings can’t take precedence over lives.

Conclusion

In volume 100, issue 1 of Scientific American an article was published that said, “That the automobile has practically reached the limit of its development is suggested by the fact that during the past year no improvements of a radical nature have been introduced.” That was in 1909. It seems like science fiction, saying that humans might not be a necessary part of the equation in a car, but it’s possible. We’ve got to be open to a future where it’s true. People’s lives depend on it.

 

Works Cited

“Automobile and Motor Boat.” Scientific American, vol. 100, no. 1, 2 Jan. 1909, archive.org/details/scientific-american-1909-01-02 p.6.

Starr, Chauncy. “Social Benefit versus Technological Risk.” Science, vol. 165, no. 3899, 1969, pp. 1232–1238. JSTOR, JSTOR, www.jstor.org/stable/1727970.

Morando, Mark, et al. “Studying the Safety Impact of Autonomous Vehicles Using Simulation-Based Surrogate Safety Measures,” Journal of Advanced Transportation, vol. 2018, Article ID 6135183, 11 pages, 2018.

“Preparing a Nation for Autonomous Vehicles Opportunities, Barriers and Policy Recommendations.” Eno Center for Transportation, Oct. 2013, http://www.enotrans.org/etl-material/preparing-a-nation-for-autonomous-vehicles-opportunities-barriers-and-policy-recommendations/.

Schenck v. United States, 249 U.S. 47, 49-51, 1917.

“Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles J3016_201609.” SAE International, SAE International, 16 Jan. 2014, http://www.sae.org/standards/content/j3016_201609/.

United States, Congress, Cong., Committee on Commerce, Science, and Transportation. “Hands off: The Future of Self-Driving Cars.” Hands off: The Future of Self-Driving Cars, U.S. Government Publishing Office, 2016. 114th Congress, 2nd session, report.

United States, Congress, Singh, Santokh. “Critical Reasons for Crashes Investigated in the National Motor Vehicle Crash Causation Survey.” Critical Reasons for Crashes Investigated in the National Motor Vehicle Crash Causation Survey, National Center for Statistics and Analysis, 2015.

“Waymo Reaches 5 Million Self-Driven Miles.” Medium, 28 Feb. 2018, medium.com/waymo/waymo-reaches-5-million-self-driven-miles-61fba590fafe.

“Waymo’s Fully Self-Driving Vehicles Are Here.” Medium, 7 Nov. 2017, medium.com/waymo/with-waymo-in-the-drivers-seat-fully-self-driving-vehicles-can-transform-the-way-we-get-around-75e9622e829a.

Zakharenko, Roman. “Self-Driving Cars Will Change Cities.” ScienceDirect, 15 Sept. 2016, www-sciencedirect-com.jpllnet.sfsu.edu/science/article/pii/S016604621630182X.

Post 5: Primary Research

To get some primary research on my topic, I conducted a survey. The survey consisted of eight questions. I wanted to find out some basic demographic information about respondents, and compare it with their trust of self-driving cars. I wanted the survey to be open to any demographic, with any person’s opinion being useful data. I was interested in finding out something about public perception of self-driving cars, and I definitely got results I wasn’t expecting.

The background information I decided to collect was age, level of education, amount of driving done in the past week, general location (urban, suburban, or rural), and if they had been in an automotive accident before. These were to see if there was any sort of trend on opinion of self driving cars across these categories. However, as far as age and location go, only 3 out of 26 respondents were over the age of 24, and only 2 lived in a rural area, making any sort of extrapolation in these aspects statistically questionable. Slightly more interesting was that of the 6 people who said they felt self-driving cars would be less safe, only 2 were drivers. This means that drivers were seemingly warmer to self-driving cars than non-drivers. Similarly, 4 out of the same six had been in an automotive accident before, perhaps having an effect on their opinions. Looking at the demographics was where I expected to get the most useful, interesting data, but it turned out that another aspect of the results took over.

By far the most fascinating thing I noticed from the results was a seeming 180 of opinion by some respondents in the last three questions. About three-quarters of respondents both said they thought that self-driving cars would be better/safer than themselves, and that they would feel either a little safer or much more safe if a large portion of other cars on the road were self-driving. Despite this, a majority (54%) of people said that if they were given a fully autonomous car without a steering wheel, they would not trust it enough to use it. This is seriously interesting to me, that people would say in no uncertain terms that they think a self-driving car would be better than them, but would still not trust one enough to ride in it. There are multiple explanations I can come up with to reconcile this. It’s possible that the first questions were very abstract in people’s minds, and so people’s risk assessment was more detached and objective, but the final question making things concrete made people consider their own personal safety as a factor in a way the previous questions didn’t. It’s also possible that the wording of the question to say “without a steering wheel” is what concerned people, as even though they had just said that self-driving cars would likely be better than themselves, people are willing to take much greater risks if they feel they have some sort of control, and removing the steering wheel feels like removing a safety net. Either way, it was insightful to see a thought process I hadn’t even really considered before.

Genre Transformation

I made a video

It’s pretty self explanatory. My target audience is broad, but posting on youtube, I think I’ll end up with a younger demographic watching the video (<25). I wanted to do something a little more lighthearted in tone to engage with comedy and leave people at least thinking about self-driving cars, and not as something to fear, or some sort of clinical technology.

Post 3: Evaluation of Credibility

The article “The Real Problem With Self-Driving Cars” written by Robert X. Cringely (pseudonym of Mark Stephens), accuses the self-driving car industry of overreaching technologically and proposes an intermediate step in the progression of self-driving technology. Stephens uses a recent example of a pedestrian killed by a self-driving car to illustrate the point that self-driving cars aren’t ready to be brought to market. He proposes that stretches of highway and cars driving on it could be made to communicate with each other, increasing throughput by allowing cars to drive closer together at faster speeds without necessitating fully autonomous cars that can navigate more complex city roads. He claims in his conclusion that such an idea hasn’t been implemented yet due to corporate greed on the part of “mass transit people” and “self-driving car moguls.”

The article is decidedly not credible. It is posted on a personal website for seemingly persuasive purposes, and while it has citations, they are irrelevant to the main point of the article.

Mark Stephens himself is of reasonable credibility. He’s a public figure about whom information is readily available online. He has a masters in communications from Stanford and has written for high profile publications such as Forbes, Newsweek, and The New York Times. The main issue is that he has few qualifications that would indicate he’s an expert on self-driving cars, and so when his article lacks good citation, his claims aren’t trustworthy.

The problem with the post is that it has absolutely no sources for the main argument. It has sources for tangential information, but the proposed improvement has no substance to it besides an anecdote. For example, the claim that his idea “could be implemented today almost for free using GPS, WiFi, and cellular data” is conspicuously lacking any sort of source or evidence, even though this is a crucial factor of the argument. Another critical premise of his argument is the safety of this proposed system, saying that if two cars are travelling at 80 mph at one meter apart and “the car ahead slammed on its brakes and you bumped into it, chances are there would be no damage.” Not only does this use the weasel words “chances are,” there is absolutely no citation to back up this claim. The source he does keep referring to is the (unnamed) cousin of Paul Sen, the series director of Triumph of the Nerds. A person he claims to have talked to about this topic 23 years ago. This raises multiple, severe problems. Both the possibility that perhaps Stephens misunderstood some facets of the idea he’s presenting, and the possibility that the idea has been pursued adequately, but dropped due to unforeseen problems or complications. To give examples of where he does use citation, he links to an article confirming that a woman was killed by a self-driving car, an article that shows Tesla is working on technology dissimilar to what he proposes, to a blog detailing what “drafting” is, and to articles showing the financial and energy costs of California’s high speed rail project (without giving any sort of estimate for his proposal).

Citations

Cringely, Robert X. “The Real Problem with Self-Driving Cars.” I, Cringely, 26 Mar. 2018.

Schillinger, Liesl. “The Double Life of Robert X. Cringely.” Wired, Conde Nast, 5 June 2017.

Post Two: Rhetorical Analysis of an Argument for Self Driving Cars

Self driving cars are on the rise, but the issue of public trust remains. In the New York Times article “Driverless Cars Made Me Nervous. Then I Tried One,” published in October 2017, David Leonhardt relays his personal experience of driving a semi-driverless car, and uses his own experience of going from skeptical to trusting to try and convince others to have the same change of heart. He uses statistics and studies to make a case that humans are irrational in the case of trust as it relates to self driving cars, and asserts that people should be more trusting and open to them as they get better and better.

Leonhardt effectively blends Logos, Ethos, and Pathos in his article. He starts and ends with a development and reassertion of Ethos. The main body of the article is heavily Logos-reliant, but has Pathos skillfully woven in to add emphasis and make readers care.

Ethos and Trust

Ethos is important in the article, it’s what bookends the whole thing, and serves as a point of entry to open people up to both listening to and accepting the statististics in the body. What’s key is that Leonhardt uses Ethos to frame himself as an everyman, not an authority. Given that his target audience is people who are skeptical of self driving cars, his anecdote is used to set himself up as a person who was also skeptical of self driving cars. He’s reaching across the aisle, effectively saying, “I’m just like you, so you can trust me.” With this, he avoids establishing a confrontational tone, but instead gives readers the notion that if they only look at the evidence he looked at, they’ll naturally come to the same conclusion. With this as a backdrop, he moves on to the body of the article.

Logos and Pathos, The One-two Punch

The body of Leonhardt’s article is fairly dense with statistics, having a link to one every two or three paragraphs throughout the body, and he capitalizes on them effectively. This technique is best exemplified in two sentences, “More than 37,000 Americans died in crashes last year, most from human error. In my community, the heartbreaking toll included a mother, father and their teenage son, killed when a speeding car slammed into their car on one of those busy suburban roads.” This is so effective because it has a linked-to statistic of deaths (Logos), followed immediately by a specific, concrete example, intended to tug at the heartstrings (Pathos). Even if readers’ eyes glaze over, and breeze past the statistic, the case of the family comes in and gives weight to the number. The next paragraph is another blend of Logos and Pathos, where Leonhardt says that cars kill more people than guns. Thus, he reasons, if you care deeply about guns, it makes sense that you should similarly care about deaths from car accidents. This is an appeal to Pathos as it tries to tap into the outrage the audience is likely to feel about guns, but it gets there through Logos, as it’s a simple logical argument that if one death count is higher, you should rationally care more.

Overall, the article has compelling rhetoric, and taps into all three basic forms, Logos, Pathos, and Ethos. Leonhardt orders and blends them well to make a solid case.

Works Cited

Leonhardt, David. “Driverless Cars Made Me Nervous. Then I Tried One.” The New York Times, 22 Oct. 2017, http://www.nytimes.com/2017/10/22/opinion/driverless-cars-test-drive.html.

Post One: An Introduction To Cars Without Drivers

Self driving cars. You’ve almost certainly heard of them, but if you haven’t, fret not, their name is their definition. Of course, things get more complicated, they are an experimental technology at the forefront of innovation and they’re quite literally a matter of life and death. They offer to both improve traffic flow, and prevent accidents. So the way I see it, self driving cars are inevitable. In some form or another, they’re happening. Therefore, the question of whether self driving cars should see widespread use or not is irrelevent, what I’m interested in is how they come into widespread use. Boiling it down to its simplest, there are three main players: drivers, companies, and government. I hope to look at how each deal with and respond to the notion of self driving cars.

How did I get so interested?

Youtube. Two videos in particular, and here they are, if you’re inclined to watch. I was always vaguely aware of self driving cars, but these videos made me sit up and pay attention because they made self driving cars and what they can do real, they made it personal. The traffic I sit in and the mistakes I make behind the wheel are at stake, and could be wiped away within my lifetime. Now I don’t see self driving cars as a novelty anymore, I see them as a small piece of science fiction becoming reality before my very eyes.

Article 1

So, I’ve really been hyping up self driving cars and what they can do, but things get both messy and interesting when you take a look at what they can’t do. And boy are there problems with the things. It’s understood that self driving cars aren’t perfect yet, and so it’s pretty much common practice right now that self driving cars have a person behind the wheel to take over in an emergency. But New York Times author John Markoff points out the issue with this thinking in the aptly-titled article, “Robot Cars Can’t Count on Us in an Emergency.” In it, he discusses the ways companies such as Google and Toyota are having to contend with the distractable nature of human drivers. He looks at expert opinions, and presents the case that while fully manual driving and full automation would both be stable, the middleground of a car needing a human backup in emergencies isn’t, and that the opposite is preferable, a computer giving a hand to an occasionally distracted driver. The article ends on a very cogent point backed by a statistic, the question of whether people would even trust self driving cars at all. The article illustrates a clear two-way street of distrust between companies and drivers, where neither is perfect, and a solution has yet to be reached.

Article 2

The government, of course, is necessarily involved in the issue of self driving cars, but before they can even get to specifics of how they should regulate them, they have to sort out who can regulate them in the first place. In the New York Times article, “U.S. Regulators Seek Public Views on Self-Driving Trucks and Trains” David Shepardson brings up the issue that drivers are under the purview of state government, while cars themselves are regulated federally. So when the two combine, who does the decision fall to? He says the decision is going to be decided on later this year, and points to a decision that already needs to be made, with General Motors trying to start a driverless ride-sharing service. And so another point of contention is developing. Companies are eager to press forward on self-driving cars, as the financial incentives for being the first into a new market are astounding, but the government has to wrap red tape around everything and ask important questions about safety.

In Conclusion…

I was absolutely on the side of the unmitigated benefits of self driving cars at one point, but even with an optimistic view of the long term, I’m now seeing the convoluted mess the short term holds. And the topic’s relevancy is undeniable. I never saw any self driving cars when I was younger, but new just walking out into the city, I see them consistently. Now is the time to be watching them.