Society appears to be on the brink of accepting driverless cars. Governments around the world are modifying laws, or introducing new ones, to cover the concept of a car driving itself and passengers to a given location. Car manufacturers and software companies are developing the systems to automate cars, helped by advances in technology that make it easier for a computer to analyse and process the information in order to drive. However, there remain significant difficulties that must be overcome; we are only just beginning to understand the implications of vehicles driving themselves. It will have ramifications for the way that we own, use and tax. Industry experts believe we are around a decade away from the wide scale consumer introduction of driverless cars and in the next ten years, we expect to see considerable development.
One area is that of device security, and by device I mean “car” (or truck) here. Cars already have an increasing reliance on computer and similar technology, but when a computer starts to drive the vehicle this gives a different tilt to the potential impact of those systems being hacked and compromised by malicious code. We’ve already seen how hackers are able to control vehicle driving systems such as engine output, braking and steering systems. If compromised, a self-driven vehicle becomes a significant risk to anything around it. There is another part of security and that is of privacy: self driving cars need to process an enormous amount of sensor data in order to be able to control the vehicle and respond to changeable conditions. The vehicle needs to be aware of other road and sidewalk users, and whilst this information is designed to be collected anonymously there may be identity clues embedded into the data collected. It’s a privacy minefield, but if used responsibly it could be very useful. One example is to support police and other authorities’ investigations into incidents: the driverless car has no bias in the data it records and is accurate. The data collected becomes a vehicle black box, similar in some respects to the black boxes in commercial airliners. Of course, this information needs to be secure and tamper-proof, which leads us back to the complex issue of in-car system security.
Another benefit of driverless vehicles is the ability to combine the data from each vehicle in order to streamline our road traffic network. If these vehicles could be networked, this could create the ability to significant reduce congestion and so air pollution and traffic delays. A road could anticipate busy stretches and adjust the flow of traffic into and out of the region in order to smooth the flow of people, and reducing the time it takes to move from point to point. Imagine the potential information that could be gleaned here for making our transportation systems work smarter and harder for us, whilst reducing the environmental impact? Depending on where one looks, the potential for networked smart cars reduces the risk of collisions by ninety percent based on the premise that most collisions are caused by something unexpected happening. In the case of a stream of networked vehicles, if something unexpected happens, every connected car would be made aware of the issue almost instantaneously.