Recent reports that car production during 2021 and 2022 will have to decrease substantially due to the global shortage of silicon chips highlights the fact that our cars are now computational devices – realising Elon Musk’s 2015 description of Tesla being in the business of manufacturing “sophisticated computers on wheels”.
While the dream of the ‘fully autonomous’ vehicle remains a long way from being realised, many cars currently on our roads are stuffed full with computer technology – powering everything from digital dashboards through to cruise control and ‘cabin environment’ management’.
**
On the road to facial recognition
One computational feature that is beginning to be added to many high-end models is facial recognition. As is often the case with FRT, these applications are centred around various promises of safety, security, convenience and comfort.
For example, Mercedes-Benz was one of the first manufacturers to introduce in-car cameras and facial analysis technology to infer driver fatigue and trigger drowsiness alerts. Similarly, Hyundai’s ‘Driver State Warning’ system combines facial analysis data with car speed and angle to warn drivers if they are showing signs of unsafe driving.
Facial recognition technology is also being used for more ambient and aesthetic purposes. For example, Subaru’s ‘Driver Monitoring System’ will recognise the faces of up to five different drivers and adjust seating, mirror and in-car temperatures to fit the personal preferences of whoever is behind the wheel. Similar systems being developed by the likes of Jaguar, Land Rover, Mercedes-Benz and Chinese car manufacturer Bestune can also alter in-car entertainment and lighting choices to fit the mood of the driver and passengers.
**
So what’s not to like about this latest development in automotive sophistication?
Of course, there are all sorts of minor inconveniences that might arise from a car mis-recognising your state of alertness, or mis-judging you as being in the mood for a blast of loud techno music. However, more concerning are the ways in which in-car facial recognition might be contributing to the normalisation of technology that many people might feel has other deeply problematic uses.
In this sense, in-car facial recognition might be seen as a prime example of what Chris Gilliard and David Golumbia term ‘luxury surveillance’ – the willingness of middle-class consumers to pay a premium for tracking and monitoring technologies (such as GPS devices and home smart camera systems) that get imposed unwillingly in other guises on other marginalised groups.
This highlights the tricky nature of debates over the benefits and harms of the insertion of facial recognition technology into public spaces and everyday society. Indeed, cars are a good example of how seemingly innocuous facial recognition features are being quietly added to many of the most familiar and intimate settings of middle-class lives (such as the car), at the same time as major push-back against the broader use of this technology in public spaces, shops and schools, let alone by police and security forces.
At the moment, many middle-class people seem content to accept two different modes of the same technology. On one hand is the ‘smart’ convenience of being able to use one’s face to unlock a smartphone, pay for a coffee, open a bank account, or drive to work in comfort. On the other hand, might be a general unease at the ‘intrusive’ use of FRT in their child’s school or by their local police force.
People who drive a Mercedes Benz and own an iPhone might feel fine with the distinction that their personal uses of FRT are acceptable, and that they are in the privileged position of being able to benefit from these particular applications and the insights, convivences and controls that they add to everyday life.
Yet, this attitude could be seen as a slippery slope – weakening protections for how the same technology might be used on less privileged populations in more constrained circumstances. The more that FRT is integrated into everyday objects like cars, phones, watches and doorbells, then the more difficult it is to argue for the complete banning of the technology on grounds of human rights or racial discrimination.
Thus the downside of middle-class consumers continuing to engage with forms of facial recognition that they personally feel ‘work for them’, is the decreased opportunities to initiate meaningful conversations about whether this is technology that we collectively want to have in our societies. As Gilliard and Golumbia conclude:
“We need to develop a much deeper way of talking about surveillance technology and a much richer set of measures with which to regulate their use. Just as much, we need to recognize that voluntarily adopting surveillance isn’t an isolated choice we make only for ourselves but one that impacts others in a variety of ways we may not recognize. We need always to be asking what exactly it is that we are enthusiastically paying for, who ‘we’ are and who is the ‘them’ on the outside, and what all of us are being made subject to when we allow (and even demand) surveillance technology to proliferate as wildly as it does today”.
Automated Vision 