The big question:
Self-driving cars operate effectively without human intervention—so, are the systems they're built on vulnerable to malicious attacks..?
The ability to find specific places around the country, including in our towns and cities, and rural areas is one of the key parts of any driverless vehicle.
With the public already a little skeptical about how well services using autonomous vehicles will work, the operators and system specialists cannot afford to get it wrong.
Cloud storage for maintaining up-to-date mapping systems
An autonomous vehicle is useless without the knowledge of its surrounding and a highly detailed map of where it is supposed to be going and how to get there quickly and easily.
Autonomous driving technology is a complex integration of technologies, consisting of three major technology areas.
These are:
(1) the algorithms, including sensing, perception, and decision-making code,
(2) the client system, including the robotics operating system and the hardware platform and,
(3) the cloud platform that is used for data storage, simulation, and HD map generation.
The first two systems are on-board, but it is essential that the mapping system is remote yet able to be accessed instantly, to deal with the constantly changing conditions on the road. To be able to achieve this, mapping systems have to be separate from the vehicle but instantly accessible to remain fresh and viable.
Fluid Roadways
While our roads may seem to be fairly stable, to a vehicle that is unable to ‘see’ in real time, and has to rely on information about its environment that it is being fed. Plainly, if that information is incorrect, then there can be serious problems and potential accidents waiting to happen.
The main problem here is that our roads are not static, and there are always changes and alterations going on. Roadways are constantly being subject to roadworks, closures, and disruptions such as traffic accidents.
To be able to overcome these obstacles, autonomous vehicles fitted with cloud-accessible HD maps and fitted used with deep learning algorithm software will be able to obtain information not only from satellites, but also from other autonomous cars, and even city infrastructure, to build up a full picture.
Cloud storage might be safe, but is it secure?
So cloud technology is obviously the answer to giving autonomous vehicles instant access to the most up to date information on traffic routes and issues, but how safe is that?
If the vehicle is in contact with a cloud storage area and being updated on maps and traffic situations, does that not leave them vulnerable to anyone with the technology to interfere with those signals?
The issue of malicious intrusion – either against the storage facility itself or any recipient to that information – have been around ever since the inception of the technology, but as the use of cloud storage has grown, so too has the desire from hackers to break into them.
Whether it is as a threat to make money from the storage facilities, as a simple prank, or just to prove the vulnerability of the systems, secure storage can always be attacked. And just one incidence of corrupted data could have severe ramifications for autonomous vehicles!
Rising to the Challenge
Autonomous vehicles represent a variety of challenges, not least the ability to navigate ever changing roads with complete safety, and that means maps of detail far greater than anything produced.
Cloud-to-car mapping systems allow everyone from suppliers and equipment manufacturers, to the users of driverless cars to get the most up-to-date information about the road network that they are using. Live, constantly updated HD maps on the cloud allow highly or fully automated vehicles to choose the optimal driving strategies based on roadway profile, lane curvature and terrain, congestion, detours, and other factors influencing driving safety.
The safety of these systems must be paramount and cloud computing systems are stepping up to the challenge of invasive hacking by considering new end-to-end encryption, similar to cryptocurrencies, making them very difficult – if not impossible by today’s standards to infiltrate. However, as hackers get smarter, systems are likely to have to continue to evolve.