This paper makes the assertion that since passenger safety is of utmost importance when dealing with autonomous cars, any safety measures should only be cleared if it is guaranteed to work. On the road, it is impossible for an autonomous car to be able to predict (with certainty) how the given scenario will unfold. This leads to the vehicle having to analyze an infinite set of possible outcomes. In order to compute this, the researchers in this study propose that a hybrid verification method should be used, which would allow for the infinite set to be computed. This is done by only computing the …show more content…
The researchers state that human drivers tend to attempt to foresee what other drivers on the road will do in order to avoid danger, and that this foresight is something that should be emulated in autonomous vehicles. The safety assessment system would serve as this foresight. It is proposed that the system would detect dangerous situations and warn the vehicle about it so that action can be taken to change the trajectory of the car and avoid the danger. For example, if the autonomous vehicle is planning to make a left turn at an intersection but there is another driver approaching in the opposite lane, the safety assessment system would detect the threat and warn the vehicle to change its planned trajectory, in order to avoid …show more content…
That being said, many of the proposed intelligent transport systems require the transfer of data between vehicles, between vehicles and infrastructure, and/or between vehicles and the internet. With these connections, it is important to consider the importance of cybersecurity and how cyberattacks could prove to be a very large issue that must be addressed. This article describes the possible ways autonomous vehicles can be attacked through the internet, while also offering possible ways to protect against these attacks. There is a long list of possible attack methods, ranging from jamming the various sensors to inputting malware into the devices’