| Line 5: |
Line 5: |
| | Getting a car to drive forward between the [[white]] and [[yellow]] lines of a road is not as difficult as one might think. Sensors and [[computer]]s have advanced to the point that understanding the surroundings, and reacting to them is entirely feasible. Even fine-tuned maneuvering can be performed by modern [[technology]]. The truly complicated part of development is detecting and responding to the unexpected. Poor driving conditions can be an issue as well as road construction, but the erratic driving of others is probably the most serious concern. Those driving under the influence of [[alcohol]] in the vicinity of an autonomous car can particularly be an issue, since the computer will have a hard time trying to predict what will happen next.<br /> | | Getting a car to drive forward between the [[white]] and [[yellow]] lines of a road is not as difficult as one might think. Sensors and [[computer]]s have advanced to the point that understanding the surroundings, and reacting to them is entirely feasible. Even fine-tuned maneuvering can be performed by modern [[technology]]. The truly complicated part of development is detecting and responding to the unexpected. Poor driving conditions can be an issue as well as road construction, but the erratic driving of others is probably the most serious concern. Those driving under the influence of [[alcohol]] in the vicinity of an autonomous car can particularly be an issue, since the computer will have a hard time trying to predict what will happen next.<br /> |
| | An overlying problem is that experienced human drivers have a certain situational awareness. They know what to expect and prepare for, while the computer driver operates almost entirely from a "snapshot" view. While a human might see a red street light through the trees and begin to slow, the [[computer]] will not recognize it until the light is closer. Likewise for many other situations, like a pedestrian crossing the street or brake light coming on up ahead. A human knows what will happen next from experience, but the computer can only try to guess from records and guidelines.<ref name="bbc.com">http://www.bbc.com/future/story/20150824-what-may-be-self-driving-cars-biggest-problem</ref><br /> | | An overlying problem is that experienced human drivers have a certain situational awareness. They know what to expect and prepare for, while the computer driver operates almost entirely from a "snapshot" view. While a human might see a red street light through the trees and begin to slow, the [[computer]] will not recognize it until the light is closer. Likewise for many other situations, like a pedestrian crossing the street or brake light coming on up ahead. A human knows what will happen next from experience, but the computer can only try to guess from records and guidelines.<ref name="bbc.com">http://www.bbc.com/future/story/20150824-what-may-be-self-driving-cars-biggest-problem</ref><br /> |
| − | Another consideration is that if autonomous cars become common, what kind of driving skill with future drivers have? At a critical moment when the car tells them to take control, will they be capable of solving the problem? Also, what will the focus of the "psudo-driver" be? If an experienced driver was watching the road as if they were driving, they could probably react to a situation effectively. However, if "nothing ever goes wrong" so they get distracted with a [[Cellular phone|phone]], passenger, or even nap, they might be unable to intervene properly when needed.<ref name="bbc.com"/> | + | Another consideration is that if autonomous cars become common, what kind of driving skill with future drivers have? At a critical moment when the car tells them to take control, will they be capable of solving the problem? Also, what will the focus of the "psudo-driver" be? If an experienced driver was watching the road as if they were driving, they could probably react to a situation effectively. However, if "nothing ever goes wrong" so they get distracted with a [[Cellular phone|phone]], passenger, or even nap, they might be unable to intervene properly when needed.<ref name="bbc.com"/><br /> |
| | | | |
| | ===Progress=== | | ===Progress=== |
| | Despite all of the problems and complications of this development, progress has moved forward with this project. Scientists and engineers have actually been designing autonomous vehicles since the mid-1960s, but crucial innovation happened in 2004 when [[the Pentagon]]’s research arm began its Grand Challenge.<ref>http://www.nytimes.com/2010/10/10/science/10google.html?_r=2&partner=rss&emc=rss&pagewanted=all</ref> <br /> | | Despite all of the problems and complications of this development, progress has moved forward with this project. Scientists and engineers have actually been designing autonomous vehicles since the mid-1960s, but crucial innovation happened in 2004 when [[the Pentagon]]’s research arm began its Grand Challenge.<ref>http://www.nytimes.com/2010/10/10/science/10google.html?_r=2&partner=rss&emc=rss&pagewanted=all</ref> <br /> |
| | In recent years, The [[United States Government]]'s [[DARPA Challenge]] produced some credible self-driving cars, with teams from [[Stanford University|Stanford]] and [[Carnegie Mellon University|Carnegie-Mellon]] winning prizes in 2005 and 2007. [[Google]] was also joined the race, having announced it has tested autonomous cars (with a safety driver) on public roads. [http://googleblog.blogspot.com/2010/10/what-were-driving-at.html] | | In recent years, The [[United States Government]]'s [[DARPA Challenge]] produced some credible self-driving cars, with teams from [[Stanford University|Stanford]] and [[Carnegie Mellon University|Carnegie-Mellon]] winning prizes in 2005 and 2007. [[Google]] was also joined the race, having announced it has tested autonomous cars (with a safety driver) on public roads. [http://googleblog.blogspot.com/2010/10/what-were-driving-at.html] |
| | + | |
| | + | ==Security== |
| | + | Autonomous cars have shown a concerning trend towards being vulnerable to cyber attacks. Tencent Keen researchers have successfully hacked Tesla’s Model X in both 2016 and 2017. They were able to control doors, lights, and breaks in 2017.<ref>http://fortune.com/2017/07/28/tesla-model-x-tencent/</ref><ref>http://keenlab.tencent.com/en/2017/07/27/New-Car-Hacking-Research-2017-Remote-Attack-Tesla-Motors-Again/</ref><ref name="Cybrary 17-08-11">https://www.cybrary.it/0p3n/2017-08-11-unm4sk3d-cyber-security-news</ref> The vulnerability used for their 2016 hack of the car was presumably patched, but using a similar strategy, Tencent Keen was able to break in again in 2017. That vulnerability has reportedly been patched as well.<br /> |
| | + | There have been other occasions where researchers have taken control of such cars remotely, sometimes even gaining access to throttle control and power steering.<br /> |
| | + | Another potential issue is that autonomous cars can be deceived, whether by accident or design, into misinterpreting street signs and markings, and acting improperly and as a result.<ref name="Cybrary 17-08-11">https://www.cybrary.it/0p3n/2017-08-11-unm4sk3d-cyber-security-news</ref> |
| | | | |
| | ==Science Fiction Become Reality== | | ==Science Fiction Become Reality== |