In recent years, while technology has improved, making society more efficient, efficiency comes at a cost. The invention of the cell phone has led to a new problem, texting while driving. Around twenty-six percent of all car accidents are caused by texting and driving, which amounts to hundreds of thousands of crashes per year. ArriveAlive, is a form of AI that detects when a person is driving and terminates all cellular data. The AA Algorithm prevents users from accessing games and social media while still allowing for emergency calls. Apple currently has an algorithm known as “driving mode.” The algorithm detects speed using the phone’s built in speedometer and disables notifications. A message displaying “driving mode activated” appears on the screen and requires a tap of the “I’m not driving” button on the bottom of the screen to regain normal function.
The issue of texting while driving has existed for as long as phones have been around. Below are some of the strategies some companies attempted to solve this issue. Apple currently has an algorithm built into the iPhone Operating System known as “driving mode.” The algorithm detects speed using the phone’s built in speedometer and disables notifications. A message displaying “driving mode activated” appears on the screen and requires a tap of the “I’m not driving” button on the bottom of the screen to regain normal function. The flaws with this algorithm are it lacks incentive and enforcement. Humans need a reason to not use their phones while driving because they tend to ignore the risks of doing so because the chances of getting into a car crash are statistically low. Nearly three-quarters of all drivers admitted to texting while driving according to CBS News. Most drivers with iPhones simply press the “I’m not driving button.” Another company that provides an incentive for drivers to put down their cell phones while driving is the Forest App. Forest is an application on a phone increases productivity by slowly growing a tree when opening the app. The user loses the virtual tree when closing the app, creating a free incentive to use the phone less. While proven to increase productivity, the app lacks enforcement. This app targets teenagers who have trouble managing their time and tend to browse social media for long, extended periods of time and lacks functions to prevent drivers from using their phones. A company that took a similar approach to our organization is Cell Control. The Cell Control IOS app disables notifications and text messages when driving. One of the application’s most celebrated features contradicts the purpose of the app. Cell Control allows users to customize when the app disables the cell phone’s functions, relying on the user’s self control to keep the roads safe. As proven in the statistics shown above, a majority of drivers lack self control. Furthermore, the app costs $8 a month, reaching few users. This, once again, contradicts their mission statement as they save few lives. Based on a survey from AT&T, over 90 percent of the people who text while driving already know the risk of texting while driving, but choose to do it anyway since they believe they can focus on multiple things at once. They also said that if they didn’t check their notifications immediately, they fear they would miss out on something important. This significantly reduces their reaction time and puts the people around them at risk. Statistics from the United States Department of Transportation show that around 1.6 million car crashes a year are due to the use of cell phones. Self regulation isn’t enough to stop this bad habit and many solutions currently available are optional even though in many places it is illegal to drive while using the phone.
During years 2012-2017, there were about 3030 distracted driving crashes each year. This leads to nine percent of all fatal crashes. NHTSA released a study on why in 2010 there were 33 thousand highway fatalities and 3.9 million injuries in the U.S. They found out that these crashes led to $277 billion dollars lost economically and another $594 dollars in societal harm. This accounted for 17 percent of the economic losses. From 2011 to 2015, it was reported that the number of cell-phone related crashes increased from 50 percent to 70 percent, a 40 percent increase. The Center for Disease Control and Prevention (CDC) reports that a third of all drivers between ages 18 and 64 read or sent text mails/emails while in traffic. In California, driving laws are much looser. A driver who kills another person because of texting or talking only gets a vehicle manslaughter charge and up to a year of jail time. Teens are at a greater risk as they are four times as likely to get into one of these accidents. Clearly, drivers are in need of a method of enforcement to promote safety.
With the use of artificial intelligence becoming a popular in many large tech companies, our idea isn’t that far away from becoming a reality. The ArriveAlive technology will take advantage of the newer and sharper cameras incorporated by smartphone manufacturing companies. While many of our societal problems have been solved by the government, safety remains a critical issue and the government will fund ArriveAlive because of communal concerns. Based on the previous statistics about the cost that distracted drivers created through damages, our technology would significantly help financially, taking away billions that need to be spent on repairs and emergency care.
The ArriveAlive technology will be a breakthrough in driving safety. All phones will eventually have the ArriveAlive software, so all drivers must obey driving laws which will be enforced. Other companies will begin to use artificial intelligence to run similar safety-related programs that enforce laws. No other company has touched upon artificial intelligence to solve this problem, but our organization has talented AI programmers. We estimate we will save 3500 lives a year by incorporating our software into the phone.
The first portion of our software would be the speedometer that checks whether the AI needs to be used. Since many phones already have built in speedometers, our software would simply use those and check if the user is moving over 15 miles per hour. If this condition is met, the AI would turn on. It would learn to detect if a person is in a car by using millions of images we give it of people driving. It would then associate drivers as people who are sitting behind the steering wheel, have a foot on the gas pedal, or any other signifiers that a person is driving. We would also flood it with images of people that it are not driving or that might be mistaken for drivers so it can learn what not to look for. As we give it more photos, the accuracy with which it determines drivers would be near perfect. Using a camera, it would compare its surroundings with the archives of photos it has. If it can determine that the user is not the driver, full access to their phone will be given. However, if they are driving, everything will be restricted except the use of emergency calls. A handsfree mode that gives directions will also be offered, but an address can only be typed in while the car isn’t in motion. It will also only give an audio for directions so that the driver never has to look at their phone while driving. After this has been determined, the images that the camera takes of the inside of the car will be deleted. As we would need to partner with large phone companies or cell service providers to implement this, we would first need to develop a nonprofit that could reach out and try to partner with these companies. With their help, the technology could be placed straight onto the phone. This would ensure that the consumers can’t just refuse to use it and resume their bad habits that put others at risk.
One problem with AI is that there is always a chance that it lacks the judgement and experience humans have and could block a person from using a phone when they aren’t driving. This could anger consumers who have done nothing wrong but are being denied from the phone access they pay for. However, this issue isn’t nearly as troublesome as the case in which it fails to recognize that a person is driving. If people learn to cheat our system, or if it fails to pick up on a driver and they choose to use their phone, the same issue of distracted driving could still exist. While it would be significantly smaller than the risk without our technology, it would still expose a flaw in our design that can cause harm to other people. Another issue that could arise is a person’s privacy. While our AI would stop using the camera once it determines if one is driving, many may not ever want their camera to look into their car in the first place. People may have private belongings that they are worried companies might secretly save, or that they don’t want people to see what they have and may see it as an invasion of privacy. Our design for navigation also fails to cater to people who are deaf or hard of hearing. If a person isn’t able to hear their directions, it may be a hassle for them to pull over after every few directions to remember what turns they must take next. If they are on a freeway and forget what they previously read, they could end up taking a wrong turn and wasting time trying to figure out how to get back and track towards their destination.
The method ArriveAlive will use to earn revenue and incorporate our software into phones is by making our organization a nonprofit. We seek to request nonprofit permission and certification from the IRS through a Harvard lawyer. By creating a nonprofit, our organization will receive funding from the government. We are confident in receiving funding from the government because the government spends over $200,300,000,000 annually on texting related crashes. We save the government money. After partnering with the government, we will petition the government to pass a regulation which states all phone manufacturers must use our software in order to sell their phone in America. Currently, we have relationships with engineers in Google and Apple, so we’ll have an easier time partnering with things. FCC has a process where radio and cell phone manufacturers send a sample of their products to approve. The consultation fee allows FCC to make money and products must have FCC certification to enter the market. We test by patenting our ArriveAlive software and giving rights to companies to use. After they incorporate our software into their sample, they will give us a phone to test. We test by using a cart with walls surrounded by screen which simulates a driver driving. If the phone passes the test over an extended period of time, the phone will receive the AA certification. Financially, our company will be stable, requiring only four operating services engineers to program the ArriveAlive software. Other costs include lab space, corporate counseling lawyers to manage the legality of our organization, testing costs, and office/maintenance costs amounts to $1516500. The software should take around two years to complete. To be more specific on the software creating process, we will create a bot that simulates weak AI and program it to learn patterns.
Works Cited
- Artificial Intelligence: Face Recognition. (2018, June 1). Retrieved from
- https://www.everteam.com/en/artificial-intelligence-face-recognition/.
- Bill would require motorists to unhand their phones. (2006, June 20). Retrieved from
- https://www.post-gazette.com/frontpage/2006/06/20/Bill-would-require-motorists-to-unh
- and-their-phones/stories/200606200114.
- Chaudhary, S. (2019, August 7). Artificial Intelligence 101: How to get started. Retrieved from
- https://www.hackerearth.com/blog/developers/artificial-intelligence-101-how-to-get-started/.
- Christopher, Caleb, C., Caleb, C., Caleb, C. C. C., Caleb, C., WeeWorld, & WeeWorld. (2011,
- January 12). Build a GPS Speedometer: User Interface and Polish. Retrieved from https://code.tutsplus.com/tutorials/build-a-gps-speedometer-user-interface-and-polish--active-7150.
- Essex, A. (n.d.). Retrieved from
- Holzwarth, A., & Holzwarth, A. (2018, April 4). The Real Reason You Shouldn't Text While
- Driving. Retrieved from https://behavioralscientist.org/the-real-reason-you-shouldnt-text-while-driving/.
- http://www.ncsl.org/research/transportation/cellular-phone-use-and-texting-while-driving-laws.aspx.
- How to Build an App from Scratch in 10 Easy Ways. (n.d.). Retrieved from
- https://www.techaheadcorp.com/blog/how-build-an-app/.
- Op-Ed: Texting while driving is as dangerous as driving drunk. We need to treat it accordingly.
- (2018, April 4). Retrieved from https://www.latimes.com/opinion/livable-city/la-oe-wilson-texting-while-driving-20180404-story.html.
- Possible Solutions to Texting and Driving. (2019, April 8). Retrieved from
- https://distracteddriveraccidents.com/possible-solutions-to-texting-and-driving/.
- Reddy, C. (2019, May 12). Face Recognition using Artificial Intelligence. Retrieved from
- https://towardsdatascience.com/face-recognition-using-artificial-intelligence-fffa3b20ad5f.
- Santana, P. (n.d.). How to Start a California Nonprofit. Retrieved from
- https://calnonprofits.org/resources/starting-a-california-nonprofit.
- State lawmakers try to curb driver distractions. (2007, October 16). Retrieved from
- https://www.post-gazette.com/frontpage/2007/10/16/State-lawmakers-try-to-curb-driver-distractions/stories/200710160270.
- Strayer, D. L., Drews, F. A., & Johnston, W. A. (2003). Cell Phone-Induced Failures of Visual
- Attention During Simulated Driving. Retrieved from https://msu.edu/course/psy/802/snapshot.afs/altmann/802/Ch3-3b-StrayerEtAl03.pdf.
- The Dangers of Distracted Driving. (2019, May 10). Retrieved from
- https://www.fcc.gov/consumers/guides/dangers-texting-while-driving.
- Ways to Avoid Texting While Driving. (n.d.). Retrieved from
- https://www.idrivesafely.com/defensive-driving/trending/ways-avoid-texting-while-driving.
- Young, L. (2019, March 6). How to Build an AI Startup, and Do You Really Need To? Retrieved
- from https://thoughtbot.com/blog/how-to-build-an-ai-startup-and-do-you-really-need-to.
