Tuesday 12 August 2008
More than a luxury?
Paul Cheeseman, Global Technical Director, Lloyd’s Register Group, discusses the idea that satellite navigation systems are no longer a luxury item but, increasingly, an essential safety tool
An investigation by the Dutch research institute TNO, which was carried out on behalf of Delta Lloyd Verzekeringen, TomTom International BV, Aon Nederland and Athlon Car Lease, has found that there is an overall positive influence on driving safety as a result of the fitting of satellite navigation systems (TNO report number 2007-D-R0048B; ‘Nomadic tendencies’ p.17, ITS International March/April 2008).
The work contradicts the opinion of those who suggest that such devices are a further distraction for drivers. It is not just the view of some private individuals that sat-nav systems can have a deleterious effect on safety.
For instance, paragraph 150 of the UK Highway Code states: “There is a danger of driver distraction being caused by in-vehicle systems such as satellite navigation systems, congestion warning systems, PCs, multi-media, etc.”; it ends by saying: “Do not be distracted by maps or screen-based information (such as navigation or vehicle management systems) while driving or riding. If necessary find a safe place to stop.”
However, if we assume for a moment that the positive case for safety is made, it raises a challenging point. As an inexpensive safety measure, it could be argued that fitment as standard of satellite navigation in road vehicles should be made compulsory.
“As Low As Reasonably Practicable”
In numerous contexts, legislation in many countries requires risks to be reduced to As Low As Reasonably Practicable (ALARP) and if technology offers an inexpensive, proven method of reducing risk further the law requires that the measure be adopted. Thus it becomes compulsory, rather than simply a good idea or marketing advantage.
For those not so familiar with this legal requirement, in the UK at least it dates back to the Health and Safety at Work etc Act 1974 which places duties on employers to ensure safety, ‘so far as is reasonably practicable’ (SFAIRP). Where no specific rules exist, the legal duty is to ensure safety SFAIRP. With respect to risk this is translated into a duty to reduce ALARP.
This is no absolute science but it often falls to the courts to decide, after an event, whether an employer was exposing its workers or those affected by its undertakings to an unacceptable level of risk.
The requirements are most usually illustrated by a tolerability of risk triangle (Figure 1), although this itself is not contained in the legislation.
Simply put, any activity with a risk level towards the top of the triangle – the unacceptable region – must not commence or must cease until risk reduction measures are put in place to make it at least tolerable, except in exceptional circumstances. Activities towards the bottom of the triangle, in the broadly acceptable region, need no further effort at risk reduction but need to be monitored to ensure that levels of risk do not change over time.
More typically, risks fall into the tolerable region in the centre of the triangle. Then, ALARP applies. This doesn’t mean safety at any cost, nor does it mean taking every measure available. Instead it requires that ‘…a computation be made… in which the quantum of risk is placed on one scale and the sacrifice involved in the measures necessary for averting the risk (whether in money, time or trouble) is placed in the other’ (Asquith LJ, Edwards v NCB 1949) where: ‘sacrifice’ is the net costs of a potential measure; and ‘quantum of risk’ is the safety benefit of the measure, a collective risk estimate quantified in Fatalities and Weighted Injuries (FWIs).
Deciding this is a competence-based judgement that may be supported by good practice and cost/benefit analysis techniques: if there is established good practice, this suggests that the practice is reasonably practicable; alternatively, an estimate based on the weighing up of costs and benefits will be made.
This approach of valuing safety benefits in financial terms versus risk reduction achieved is used for highway improvement schemes and throughout the public transport industry to prioritise projects and assess their relative effectiveness. Implications for sat-nav fitment
The ALARP approach raises some interesting issues for satellite navigation. Put another way: if it is not fitted and someone is hurt, a court may decide that the provider of a car was not demonstrably reducing risk to the best practicable extent, was therefore in breach of the legislation and is thus culpable.
In applying the two tests above, is fitment good practice? In many respects it is ubiquity which establishes good practice and any argument that it is okay for top-of-the-range cars but not less expensive ones to be equipped seems to introduce double standards for the safety of the occupants and others. Secondly, what is the balance between costs and benefits?
I have not attempted a detailed ALARP calculation but it is quite possible to assess the level of risk reduction achieved with the satellite navigation system and work backwards to identify what the mandatory spend threshold would be.
It is less obvious who the guilty party might be when or if something goes wrong. Private owners who because of the cost chose not to specify an option on a car could probably only seek redress from themselves – although we might wonder about the position an injured passenger or other road user may take. However a car manufacturer which insisted on optional, not standard fitment may be on more difficult ground.
Here, we reach well-trodden earth with the infamous Ford Pinto and the decisions which resulted in a less safe car because of the cost of the redesign compared with the possible legal costs for damages.
The current legislative regime is different but the logic is rather similar. The OEM is an obvious target but what about the car-hire or leasing company, or the company providing vehicles for use by its employees? They too have the ability to reduce risk by choosing to fit satellite navigation; to choose not to do so may also demonstrate culpability.
After all, we would never accept that headlights should be an optional extra and airbags have become good practice now in the majority of vehicles. In time, features become mandatory but until then it seems cost/benefit arguments have to apply.
The existence of the technology on its own is not sufficient to precipitate a decision. Now however, that they are demonstrably a ‘good thing’, have sat-nav devices matured from ‘nice to haves’ to become essential safety equipment?
Implications for data providers
If the device manufacturers are rubbing their hands in anticipation of a sales bonanza, they might do well to pause and think further. The research findings raise further interesting risk issues similar to those which are currently facing the rail industry. Let us assume we accept the idea that satellite navigation is a safety improvement tool.
The route-finding logic and supporting data then takes on a safety-related role through the routes it selects. At a level crossing of course, this affects not just the car occupants but rail passengers and staff as the two usually separate systems interact. On roads around the world, the largest single cause of risk is the road-rail interface.
There are numerous ‘funny’ press reports of drivers using satellite navigation that end up in dead ends, stuck under bridges or turning onto railway lines at level crossings. The route-finding logic has the power to significantly alter the level of risk, for good or bad, through the approach it adopts to selecting routes. That might, for example, involve routeing traffic away from level crossings to nearby roads with bridges or underpasses.
In Figure 2, the map shows a level crossing which was the scene of a fatal accident in 2005. Yet there is a perfectly reasonable route for through traffic via a bridge just a few hundred metres to the north.
A risk assessment study identified the particular local features of the level crossing to be causing a higher than typical risk and the railway has subsequently fitted half barriers to supplement the active road signals which were in use and working at the time of the accident.
Using sat-nav systems’ route-finding algorithms to direct traffic via the bridge rather than the crossing reduces the risk to both the road user and railway passengers and staff. In reality, achieving this apparently simple change has proved to be nigh-on impossible as finding the data owner (and likely there are several different data sets) has not been possible.
Yet ALARP would suggest this is not just a desirable change but, given the low cost of a simple programming change, one that seems reasonably practicable and therefore mandatory under health and safety law.
Once the benefits of a low-cost safety improvement have been demonstrated there becomes not a choice but a compulsion to act.
Businesses take decisions and make investments which impact upon safety because they are necessary to fulfil a legal requirement and they are sensible from a commercial/business perspective
These are two different sets of criteria. This article has argued that when safety improvements are being considered and their cost is less than the monetary value of the safety benefit it would be expected that the improvements should be implemented.
We have discussed a very simple change to install satellite navigation systems and ensure they are programmed in a way which would appear to have a direct effect on the level of risk experienced by both car occupants and others. Intuitively, some roads are safer than others due to their size, design or geographic features and risk-based route-finding optimised for safety could take this into account.
Yet, it is not yet clear with whom the responsibility lays or who may be interested in co-operating to produce a viable risk-based system.
Even if this incentive of improved safety is not convincing enough an argument, there is the worst-case scenario of avoiding the need to mount a defence against liability when a case is brought alleging that a road vehicle was, for example, routed to expose it to a known hazard when it could easily have been routed in a safer way.
While there may be many reasons why different interest groups might wish to manipulate traffic flow using route-finding algorithms and the potential for vested interests to dominate is a concern.
However where there is a clear safety benefit, this surely suggests that decisive action is needed to reduce these well-known risks through some simple route-finding adjustments.
Paul Cheeseman, Global Technical Director, Lloyd’s Register Group