The road ahead for driverless cars and autonomous transport looks to be bumpy
“Self-driving vehicles are significantly dependent on network connections – and are therefore extremely vulnerable as they use old software platforms and communication channels (Wi-Fi, GPS etc). Transport with artificial intelligence (AI) will create another problem as self-learning, based on numerous data points that don’t allow for context, make it hard to tell a hacked AI from one that’s not-so-correct.
“In general, the computerisation of critical systems (such as transport) requires a totally new security approach. By now, we live in the doomed world of insufficient IT security that stays apart from industrial security and safety.”
If an incident were to occur involving a driverless vehicle – such as a crash – who would be deemed responsible? This is a question many people are asking themselves, especially manufacturers and regulators. Jonathan Hewett, CMO of Octo Telematics, says data can help.
“The rise of telematics-enabled connectivity means our vehicles are more akin to data centres than ever before, and this is only the beginning of a brand new era of data-driven mobility,” he says.
“Take the example of crash detection from embedded telematics devices: up to 400 information points are sent from a single car per second before, during and after a crash, detailing the precise behaviour of that vehicle and enabling provision of a wealth of services, from automatic emergency calls to streamlined insurance.
“This same technology sits at the heart of autonomous vehicles. Telematics not only allows car manufacturers, insurers and legislators to understand what individual driving models look like and how they can be translated into an autonomous world, but also enables us to address critical issues such as determining liability in the event of a crash involving a driverless car.“
The tech just isn’t ready
There’s no questioning the advancements being made in the area of driverless transport. Some of the biggest names in technology and motoring have already demonstrated functioning models. However, that doesn’t necessarily mean they’re safe.
Luis Gargaté, business development manager at Critical Software, suggests that we’re still a while away from seeing autonomous vehicles that are safe and reliable enough to operate in the public domain. The issue is, there’s so much tech and and systems needed in a consumer-ready driverless vehicle. While companies are still testing systems, they can prove costly to research, implement and maintain to an acceptable standard.
“To make autonomous vehicles a reality, several things are required: a reliable control computer and algorithm; a sufficient number of (equally reliable) input sensors from which the car can ‘read’ its environment; and finally, the control actuators,” he says.
“Then there is also the potential for lack of maintenance, incorrect maintenance or even ‘hacking’ of electronic control units Safety-critical real-time hypervisors could help with the control, while assuring safety. This underlying technology could be installed alongside the autopilot control software itself.
“Such a system takes a significant amount of research and development, and there is also the small print to consider: the adoption of fully independent validation and safety assessments; full certification from independent authorities; and full adoption of appropriate development standards (such as the ISO 26262), for example. We’re sitting at the traffic lights: driverless cars are a certainty, but sufficiently safe, reliable driverless cars don’t have the green light yet.”
Driverless transport is imminent. Companies and governments from all over the world are showing an interest in this area – whether it be by setting money aside for research, testing fully functioning models or creating regulations.
But that’s not to say we’ll see autonomous cars, ships, trains and planes tomorrow. The fact is, the tech is still evolving, and there are a plethora of complex security and safety challenges that need to be considered before we can see this innovation entering the mainstream.