Assessing the potential of in-vehicle enforcement systems

Jason Barnes considers the social and ethical ramifications of using in-vehicle safety technologies to fulfil enforcement functions.

Although policy documents often imply close correlation between enforcement, compliance and safety – in part, as a counter to accusations that enforcement is rather more concerned with revenue generation – there is a noticeable reluctance among policy makers and auto manufacturers to exploit in-vehicle safety systems for enforcement applications.

From a technical perspective there is good reason to do so: these systems have access to the real in-vehicle state and can highlight what information has already been relayed to the driver; and as most of the necessary functionalities are already present as a consequence of the primary safety applications the addition of enforcement capabilities is a relatively inexpensive extension.  There is a comprehensive list of enforcement and compliance applications in which in-vehicle systems could take a role. These include speed, red light, illegal lane-change/turn and tailgating enforcement, and others of a commercial nature such as uneven/over-loading.

Electronic stability control systems are already widely deployed. These work by assessing a vehicle’s dynamic status using various sensors and can, in times of need, intervene in the driving task to adjust braking, steering and powertrain output to prevent or minimise loss of control. Onboard satellite navigation and other systems, meanwhile, can already provide precise fixes on vehicles’ locations in relation to their surroundings, especially in less densely urbanised areas. Coming technologies and applications will intervene to an even greater extent in the driving task whilst the two-way exchange of information between vehicles and/or infrastructure will make the association of a whole host of data streams far more readily achievable. In the case of a red light infringement, for instance, information on a vehicle’s position, speed and vector could be combined with nearby traffic signals’ status and used to confirm whether or not an offence had occurred.

Understanding concerns

The timidity over deployment of such ‘Big Brother’ applications of in-vehicle systems is both regrettable and, at the same time, understandable. There are significant potential safety gains to be made from wider, more rapid adoption of the technologies involved and anything which hinders this should be viewed with caution. At the same time, any discussion of in-vehicle enforcement systems has to call into question the effectiveness and accuracy of the enforcement systems which sit at the roadside – something which could be difficult politically. No state or national DOT is likely to promote to the public the idea that current roadside enforcement systems can never be 100 per cent accurate; neither would the majority of major vehicle manufacturers, who would also face polarisation within their customer base if they fitted systems which were seen to ‘spy’ on vehicles’ users. This all creates a market opportunity for independent systems manufacturers and providers until such time – and there has to be a degree of inevitability to this, although it may be some distance into the future – that in-vehicle systems are politically and publicly accepted as having an enforcement role.

Current legislation places great trust in the correct calibration of roadside systems but something in-vehicle, were it to be composed of elements deemed trustworthy enough, could provide a contemporaneous record. The question arises as to what form that system should take; attempts to use information from satellite navigation systems in court to challenge enforcement systems’ veracity have proven unsuccessful, for example.

A windscreen-mounted forward-facing camera system, meanwhile, might give a driver’s-eye view of events but would not provide information on the driver’s actions. Nor would it provide information on what goes on in a vehicle’s side and rear quarters, from where information on other drivers’ actions might be useful. This all presages, of course, data which could be held in-vehicle and/or downloaded once connected vehicle-type technologies are widely deployed.

The issue hinges on acceptability and prevailing opinion at present is that there would be a negative, privacy-related response from the public. However, whilst an in-vehicle solution requires an individual to be compliant, it provides him or her with a ready record of his/her and others’ actions.

It also introduces a burden of openness; switching off any such onboard system would raise suspicion as to an owner’s/driver’s intent and even suggest an element of forethought in doing so, whilst refusing to share with the relevant authorities any evidence held therein could result in accusations of withholding evidence and charges of attempting to pervert the course of justice when such data become relevant.

Providing latitude

In order to persuade potential users to do something which does not feel forced, there must be benefits, and these can be summarised as being able to: prove that the driver was behaving correctly; provide some form of blame apportionment; and reinforce perceptions of overall good conduct. The last-mentioned might have a beneficial effect in terms of insurance premiums but throws into focus another issue relating to automated enforcement: that if we keep improving automated enforcement to the point where every infringement is recorded, no matter how miniscule, enforcement actually begins to lose its credibility and support amongst the populace.

Police officers and others involved in enforcement activities will confirm the need for latitude in enforcement decisions, and that distinctions are lost once automation occurs.
How, for example, to differentiate or enforce against someone driving at the most appropriate speed for a given section of road (as is encouraged by some advanced motorists’ organisations) and not necessarily at the speed limit?

That implies some level of quality which is absent from automated systems and, as back office systems become increasingly automated, it will become a bigger issue; whilst there are parts of the world where legally there must be human involvement in the enforcement-handling process, there are others where a digital enforcement system working with an automated back office and a national vehicle register can generate an offence notification which is first touched by a person’s hand when it is picked out of the mail.

The recipient is then faced with a choice of paying up or mounting a challenge and going through a heavy-duty and potentially expensive legal process with no guarantee of successfully overturning the charge. The solution for many is to simply pay the fine and/or accept the penalty points, albeit with ill feeling, bad grace and a slightly eroded belief in the justice of the law.

It may seem counter-intuitive but if an enforcement system is too accurate then public belief and comfort is eroded.

Far from being black and white, automated enforcement has to be somewhat ‘softer’ and – clichéd though the phrase may be – deal in shades of grey. Penalising every minor transgression of an otherwise good driver will have a caustic effect on his or her belief in the correct functioning of the law so there needs to be a solution which allows drivers to build up some form of ‘credit’ for good behaviour but which is still able to catch habitual offenders. Happily, informed opinion has it that algorithms which can do this are not beyond the wit of man to create.

In fact, technically it is not difficult to achieve all of the above – or to make the technology tamperproof. It is already possible but we will see incremental improvements. For instance, Galileo/EGNOS will improve positional accuracies and system battery lives will improve. Video logging apps which can retain and use data (including those for mobile devices, although battery life is still an issue) are already entering the market.

Unintended consequences

However, there are elements of information sharing which, whilst they may not be problem individually, can be unwanted or harmful when taken collectively or used in ways un-envisaged originally. One need only look at social media to see this happening already, where apps can hook onto an individual’s status and add all sorts of information which results in something unexpected. Status and location updates can cause embarrassment if, for instance, someone is apparently doing something unrelated to work or is apparently somewhere they shouldn’t be within their hours of employment. In some respects, that’s only the mildest of what can happen – for instance, some individuals engage in practices in their personal lives which, whilst within the law, may be on the fringes of wider social acceptability. They may not want knowledge of those activities made generally public, whether to employers or loved ones, and there is scope for rather greater embarrassment and compromise. The fact is that, as has been said elsewhere, the quantitative change in the availability of information leads to a qualitative change in privacy. The data stores of the internet are effectively ever-lasting media – put simply, once something is out there, it stays out there – and purging wrong information from the huge number of locations and outlets where it can end up is nigh-on impossible.

Slowly, and in most cases unwittingly, we are deconstructing some of the fundamentals of our very existences. Over hundreds and thousands of generations, we have built up social structures and formed mental attitudes which are predicated on retaining at least some measure of privacy. Technology’s data rapacity challenges this.

This is important and relevant because it calls into question just how much information an in-vehicle solution should hold, and which bits of it should or need not be shared.

Perhaps, in the final analysis, this is where we should start from. ‘Big Brother’ is less of a concern than he/it should be but if we are to start down this path and if technology is going to start to challenge some of the fundamentals of our society we may need to enshrine others. We should possibly start with the premise that no record whatsoever of any individual’s
legitimate activities need ever be held without consent. Whilst a simple statement in itself, it is a quite monumental task to make real even from where we already are.

Any longer wait will only complicate the task further.

• The author gratefully acknowledges the help and advice of Dave Marples of Technolution in the preparation of this article.