TPEX

Here we primarily consider autonomous vehicle as small car, taxi like replacements but they are deliberately referred to as autonomous vehicle to allow for more bigger vehicles to be imagined.

Science fiction has long imagined the autonomous vehicle and the possible benefits for the individual and society are easy to imagine. Perhaps for this reason there is historically limited opposition to their deployment.

Overall, results revealed that autonomous cars were perceived as a “somewhat low risk“ form of transport and, while concerns existed, there was little opposition to the prospect of their use on public roads.

Hulse, L.M., Xie, H. and Galea, E.R., 2018. Perceptions of autonomous vehicles: Relationships with road users, risk, gender and age.. online.

However, the potential threats of automated vehicles have long been discussed. Many of the fears related to automated vehicles perhaps can also be transferred to autonomous vehicles as well.

This paper has identified some of the cybersecurity threats to automated vehicles, with estimates of the severity of these threats and potential strategies for mitigating or overcoming these threats.

Petit, J. and Shladover, S.E., 2014. Potential cyberattacks on automated vehicles.. online.

Perhaps the autonomous vehicle offers yet more potential attack vectors. The anticipated actions may be more complex and feature differing potentially incompatible outcomes. Can we all get there are first! Within both the automated and autonomous vehicle spaces a variety of attack vectors can be considered depending on the opposing party.

Crash-Optimization Means Targeting

Lin, P., 2023-08-01. Why Ethics Matters for Autonomous Cars. online.

When an inevitable mechanical failure occurs the autonomous vehicle will almost certainly switch into a crash optimization approach. This inevitably will become targeting. Should the crashing vehicle prioritise, survivability of the passenger, minimising the damage to public infrastructure or not intruding or other road users. Would it be left to the car manufacturer, owner, government or to some other body to determine the best outcome.

When considering crash optimization it’s interesting to note that other desirable outcomes could also be considered. Consider crash optimization which seeks to preserve the driver of the vehicle and the infrastructure and is therefore is willing to sacrifice a person. Crash optimization of this sort would make it dangerous to be a pedestrian. While situations of this sort appear outlandish perhaps it is still possible especially if the pedestrians can be seen as lesser. Such tensions already exist in infrastructure design as calls to focus on active travel, pedestrian access, continue.

Necessary measures include... focusing on prioritising active travel and collective transport modes when building road infrastructure,

Cavoli, C., 2021. Accelerating sustainable mobility and land-use transitions in rapidly growing cities: Identifying common patterns and enabling factors.. online.

However the choice could be made to continue ignoring such calls and instead make the environment dangerous to pedestrians. Maybe the conflict surrounding low traffic neighbourhoods ‘LTN’ in the UK is an example of this? Making the environment dangerous to pedestrians could perhaps support revenue generation for car manufacturers, private hire or taxi firms and could also be used to suppress citizens.

Beyond making it dangerous it would also be possible to use crash optimization as a means to make it impossible to be a pedestrian. The accusation would perhaps be something like this; the vehicle and the infrastructure did everything correctly and the error was in the action of the human pedestrian. If the desire is to make it impossible to be a pedestrian, perhaps through an extension of jay walking rules, then autonomous vehicle that do not give way to pedestrians could quickly be used to create kinetic fences.

Perhaps though unexpected outcomes could be generated perhaps pedestrians could learn to game the systems as some have imagined with drivers.

If the crash-avoidance system of a robot car is generally known, then other drivers may be tempted to 'game' it, e.g., by cutting in front of it, knowing that the automated car will slow down or swerve to avoid an accident.

Lin, P., 2023-08-01. Why Ethics Matters for Autonomous Cars. online.

This could result in grid lock as pedestrians stroll down the main road with autonomous vehicles patiently queuing up behind them.

The use of autonomous vehicles to create kinetic fences would perhaps be a feature only intended to be accessible to governments. The legal framework placed around autonomous vehicles by the state is likely to be bounded by the practicalities of democracy and one would presume feature concessions to deal with concerns of civil liberties. However, if we imagine a democracy replaced with an authoritarian government then the potential risks of systems perhaps become more apparent.

State backed bad actors are useful to consider as we can presume they have a feature set beyond the typical user and likely encompassing the capabilities of an illegal user. Rogue employees with privileged access might also appear like this. What features of an automated system would be useful to the state?

Many of the imaginings of advocates of autonomous vehicles suggest the transmission of signals between infrastructure such as traffic lights, give way signs, roundabouts and a vehicle. With government control of the infrastructure and the autonomous vehicle being forced to obey instructions desirable law enforcement features are likely present. No more high speed car chases, though one hopes the film industry will be given an exemption! With complete control of the system it would also be possible for crash optimisation to sacrifice the citizen to prevent a crime through an ‘unavoidable’ collision.

Autonomous vehicle could also be used by the state to arrest or disappear individuals as required. If not the vehicle directly then the infrastructure could be used to manipulate the vehicle into breaking down or pausing at a particular point. More subtly rather than disappearing such manipulations could be used to inconvenience or delay protesters. In more extreme situations it could be possible for a vehicle to undertake a never ending journey, with individuals imprisoned within the autonomous vehicle. While potentially outlandish such circumstances offer a degree of deniability. An imprisonment could be blamed on a broken or incorrectly used autonomous vehicle perhaps like putting the wrong postcode into a GPS?

In more hostile circumstances autonomous vehicles could be used to create kinetic fences for crowd control. While roads already divide communities imagine one with a never ending supply of high speed vehicles that do not give way and no infrastructure for pedestrian crossing. Such a fence could be moved easily, sustained readily and utilised to kettle protest and isolate dissent. If required the movements or autonomous vehicles required to generate such fences could be declared as an unfortunate side effect of repositioning optimisation.

In the most crude of circumstances infrastructure could error in such a manner as to direct an autonomous vehicle to act in a way which kills its passenger. Either intentionally or through the selection of an intentionally dangerous route.

In addition to infrastructure based threats there could also be autonomous vehicles to autonomous vehicles communication threats as well. These are not necessarily a requirement of autonomous vehicles as the vehicles could sacrifice some efficiency in exchange for only reading the infrastructure. However the seamless high speed interactions at junctions imagined by many implies infrastructure mediated, if not direct autonomous vehicles to autonomous vehicles communication. Possible vectors could be ghost autonomous vehicles which appear present through fake signals but are not actually present. Such ghost autonomous vehicles could be used to manipulate traffic loads, pre positioning of vehicles or even position vehicles while in transit.

Another vector could be autonomous vehicles incorrectly stating their intended actions to cause collisions. This would be especially dangerous if autonomous vehicles signals or intentions become implicitly trusted. Of course it would also be possible for autonomous vehicles or other road users to act recklessly as well to trigger autonomous vehicle responses. g

In the worst imagined cases, drivers could play chicken on narrow roads or cliffsides, where an AV swerving could mean serious injury or death for its passengers.

Jenkins, R., 2016. Autonomous Vehicles Ethics & Law: Toward an Overlapping Consensus.. online.

While automated systems may also have crash-optimisation features a peculiar feature of autonomous vehicles is there potential deployment at large scale and to provide transit for individual users. Crashing planes through automated systems maybe be possible but it is attention grabbing. Crashing autonomous vehicle maybe far less noticeable. The crash rate of autonomous vehicles could be sustained artificially high to allow for such incidents while still being measurably safer than the current road network.

A good public proxy to these concerns would be the actions of government officials. Could we see the American president in an autonomous vehicle in America? Likely. Overseas? Less likely but maybe still possible. A British PM and American President in an autonomous vehicle together in either of their respective nations maybe.

The identification of these threats it’s not particularly novel, Rather the purpose is to reiterate the call for preventative action and to inform the public debate as others have done before.

However, it is crucial that the identified gaps are addressed before more vehicles with increasing levels of connectivity and automation are on the market.

Parkinson, S., Ward, P., Wilson, K. and Miller, J., 2017. Cyber threats facing autonomous and connected vehicles: Future challenges.. online.

Threats to Citizens from Autonomous Vehicle

As self driving cars appear on the horizon will they be safe for citizens?

Tenth Person

Future Imagining

I'm ready to future-ready my business!

References

About TPEX