As much of the transportation industry rushes to embrace a driverless vehicle future, Justin Fox takes a look at the bigger picture and asks what some of the unintended consequences could be
Connected vehicles and vehicles with autonomous features are already becoming a feature of our highways. Such technology is touted as being extremely efficient and free from the errors that human drivers inevitably make. The US Department of Transport’s outgoing Secretary of Transportation, Anthony Foxx (below), is already on board with the concept he believes will enable the prevention of the “94% of crashes that are attributable to human factors”.
Connected and autonomous vehicles (CAVs) are also already being touted as an environmentally-friendly way to reduce the number of private cars on our roads, with driverless taxis offering cardless payments that could provide incredibly cheap public transport.
Whilst this new technology seems very appealing, there are concerns from some quarters. In this article, we examine some of the concerns being raised about CAVs, and look at potential ways that risks will be minimised, as the technology becomes more and more mainstream.
How soon will we go driverless?
First of all though, there is significant debate over whether smart vehicles are even going to become widely prevalent anytime soon. Tesla CEO Elon Musk has never been one to shy away from embracing the bold and the innovative, and he argued in 2015 that not only will autonomous driving ‘just going to become normal’, but that ‘people may outlaw driving cars, because it’s too dangerous’. His point of view has merit, that whatever initial risks there are, it would be worth the trade off against the number of people killed or injured by driver’s mistakes on the road.
However Musk’s optimism is not shared by everyone with an interest in the development of such revolutionary models. Speaking on behalf of the non-profit Consumer Watchdog, journalist turned advocate John Simpson warned against developers ‘putting a glitzy spin’ on a technology ‘not remotely ready for deployment’.
AI, in particular in the automotive sector, is such a new and revolutionary technology, that it is perhaps not surprising that so many fears are being raised about its use, and about the potential for serious problems. Many of the concerns being voiced are scarily plausible and revolve around scenarios such as what would happen if the vehicle’s sensors malfunctioned, causing it to drive into the car in front of it? Many smart car systems do not have any means for a driver to take back manual control, so a system failure in these vehicles could be catastrophic. Of course, humans are generally much more fallible than computers, and more likely to make mistakes, but we all know that no software is 100% free from bugs, so the potential for problems is very real.
One issue that is potentially far more significant for the widespread adoption of CAVs is the possibility of hackers being able to take control of a vehicle remotely. The ‘Internet of Things’ has enabled all sorts of appliances and devices to utilise a permanent internet connection, from cars to kettles and toasters. That constant network connection is a potential source of vulnerability, and could allow hackers to take control of a car, in order to crash it or to steal it or its contents.
Hacking of everyday gadgets is becoming increasingly prevalent, with around 10% of people in England and Wales affected by cybercrime during 2016. While it might be annoying to have your smartphone or laptop compromised by hackers, there is at least no danger to your physical wellbeing. The same cannot be said for CAVs, and at a recent conference, security experts agreed that it was ‘almost inevitable’ that hackers would be able to exploit the systems inside driverless cars.
Protection from hackers
It appears logical that CAV technology needs to incorporate dedicated and robust protection for these vehicles, in much the same way as our smartphones and computers are routinely protected by anti-virus software and firewalls. The comparison with computer security is a good one, and demonstrates that even with a comprehensive security strategy and strong software defences, it is all too easy for a system to become compromised, as shown by the numerous high-profile corporate hacks in recent years. Just as with computer-based systems, it would appear that there can be no absolute guarantee of security and safety for smart vehicles, when faced with such a persistent and constantly-evolving threat.
One example of secure driverless technology that has been in use for many years is London’s Docklands Light Railway (DLR). Since driverless trains like the DLR cannot move from a pre-set route, they offer a safer alternative for public transport than driverless vehicles.
When looking at the security risks for smart vehicles, it’s also useful to examine the potential motives for hacking vehicles. In recent times, we’ve seen several major terrorist attacks involving lorries, such as in Nice and Berlin. Given the perceived success of these attacks, in the eyes of the terrorists, there is a very real chance that smart trucks could become the ultimate cyber-terrorism tool.
Looking at individual cars, however, it’s important to keep the risks in perspective. Your bank and personal identity details may have a significant value to a hacker, but ultimately, there is little financial gain to be had from hacking into a random private car. That said, hackers could potentially tap into private cars in order to track the owner’s movements, which could leave victims exposed to blackmail or ransomware (malicious software that blocks access to a computer system until a sum of money is paid). The threat, however remote, that a driverless car could be hacked and forced to crash, will undoubtedly prove a barrier to uptake for some. Therefore, is this a risk that is worth taking?
Even a CAV technology as seemingly innocuous as truck platooning could end up having unintended consequences detrimental to other road users. Jason Hodge, managing director of the UK’s Truck Locator, draws attention to the fact that “the UK road network does not easily lend itself to platooning”. He argues that “the short distances between motorway exits mean that a platoon of four lorries would effectively block the exits for car drivers”, and that “the large wins are for the USA, Australia and other, less densely populated countries, with a less congested national motorway systems.” Perhaps therefore it is the infrastructure that we should focus on upgrading, rather than the vehicles that use them.
Ultimately, we will all no doubt be forced to weigh up the risks against the convenience of CAVs. With today’s cars, we know the risks when we choose to get behind the steering wheel. In the driverless cars of tomorrow, though, there will be a different set of risks that may not be as obvious, and we as individuals must decide whether or not we accept them.
Good article and makes good argument/cautions with CAV technology. It is important to read and understand all perspectives and take them seriously. I do have a couple of observations
1. As an industry, lets not pretend that what we had yesterday or today is ok. We have huge transportation problems in this world - # of deaths, crashes, pollution, etc. We should not expect technological applications to be perfect and 100 percent fool proof. We have no choice but to embrace these technologies given the problems. Let's not pretend like we have alternatives.
2. There maybe a false assumption that hacking a vehicle has little financial implication. The future is my car is tied to my cell phone, my wallet, credit card, home, behaviors, etc. There are huge financial incentives to hack vehicles in the future. I can see a scenario where someone would hack my car and disable it until i pay cyber ransom.
I enjoyed reading the article. Please keep it up. The truth is no one really knows how technology can evolve and each of us is doing the best to connect the DOTs.
I 100% agree with the earlier comment regarding the "...false assumption that hacking a vehicle has little financial implication". While safety is sincerely a real concern and CAVs have the potential to address this, I am also very aware of mobility as a service (MaaS) for-profit companies that see the safety aspect as a "smoke screen" for what they really want in a CAV -- financial gain via reoccurring revenue transactions.