Implications for Delivery
Speed Assistance Systems have the potential to support speed management approaches and practitioners should consider how to:
Improve Local Infrastructure to Support Speed Assistance Systems (SAS/ISA):
The effectiveness of ISA systems depends heavily on high-quality physical speed limit signage and accurate digital speed maps. Local authorities can play a vital role by auditing, maintaining, and enhancing local signage and contributing to reliable digital mapping to improve the performance of in-vehicle speed technologies.Evaluate and Implement Fleet-Based ISA Systems Where Feasible:
Local examples like Transport for London (TfL) demonstrate that ISA can be effectively deployed in controlled fleet environments (e.g. buses, vans). Road safety teams should explore similar tailored applications, using high-quality data and practical infrastructure adaptations to maximise benefits while avoiding user frustration.Balance Technology and Human Factors in Speed Management:
While ISA has the potential to reduce casualties, poor user experience (e.g. false warnings, annoying alerts) can lead to drivers disabling systems. Road safety professionals should advocate for integrated vehicle design approaches that enhance system reliability and consider the role of ISA as one part of a broader speed management approach.
Background
Excessive and/or inappropriate speed is a factor in many collisions, both contributing to the cause of collisions and strongly influencing the severity of resulting injuries. Limiting speeding can also help to reduce emissions of toxic pollutants and greenhouse gases. It may also help to improve traffic flow and reduce the risk of congestion. Road speed limits are intended to keep road traffic speeds at safe levels but cannot be perfectly correct for every circumstance.
Drivers must comply with the limit but also ensure that their speed is appropriate for the specific conditions of the time they are driving, accounting for weather, road surface condition, and other traffic, for example. Preventing excess speed can also help other collision avoidance or crash protection systems to be more effective.
Vehicles now have some capability to control speeds and better support drivers to keep to speed limits. The systems use speed sign recognition and/or digital road speed limit map data. These Speed Assistance Systems (SAS) can:
Inform drivers of the limits
Warn drivers if the vehicle speed exceeds a set threshold
Actively intervene to prevent vehicles from speeding
Maximum speed limiters are required on many commercial vehicles. These set a fixed maximum speed (i.e. 90 km/h for HGVs, 100 km/h for coaches), regardless of the speed limit of the road it is travelling on.
Intelligent Speed Assistance (ISA) is defined by regulation as a system to aid the driver in maintaining the appropriate speed for the road environment by providing dedicated and appropriate feedback. ISA is more advanced because it adapts the maximum speed to the prevailing speed limit.
Although a ‘mandatory’ ISA is technically feasible now, it has not been required on a widespread basis anywhere in the world because the system is reliant on the quality of both the physical (traffic signs) and digital (maps and speed limit databases) infrastructure. In many places, that infrastructure is not of a good standard.
Regulations mandating the fitment of ISA require that the speed limit is correctly indicated for 90% of the driving distance in a road test, though that can drop to 80% in a specific road environment provided the overall limit is still met (e.g. if 95% on motorway and rural it could drop to 80% in urban areas). However, events where the road sign was poorly positioned, or not in the regulated catalogue of road signs in the regulation, or the sign was worn or damaged can be excluded from the calculation. As such, in the real-world, performance can be significantly worse than the overall 90%. A ‘mandatory’ ISA that incorrectly limited a vehicle to 20 mile/h on a motorway would actively create a situation of high differential speeds, also statistically proven to correlate with a higher risk of collision.
The ISA Regulation does require audible warning when the speed limit is exceeded. An ISA that is minimally compliant with the regulation, operating in a region where speed limit signs are not well implemented could be wrongly identifying that a vehicle is exceeding the limit for 20% or more of the driving distance. Audible warnings that are highly alerting also tend to be highly annoying. As such, a system like this is very likely to annoy drivers such that they switch the system off, losing any benefit where the limit is correctly identified.
Preventing drivers from switching audible warnings off has been seen as beneficial in the past (for example, Euro NCAP protocols encourage designs that make it more difficult to switch systems off). However, there is some suggestion that where audible warnings are excessively frequent and cannot be switched off, they simply become ignored, probably subconsciously.
Drivers are exposed to a wide range of information, warnings, and interventions from modern vehicles during driving. It is beneficial if each system is an integrated part of the strategy for the whole vehicle and ISO standards exist to guide this design. In general, tonal audible warnings are reserved for situations where drivers must take urgent action (within the next 2-3 seconds) to avoid a critical incident (i.e. collision). When measured across a population, even a small increase in the average speed of traffic on a road can correspond to a larger increase in the number of casualties. However, in many individual situations it is not clear that it is an urgent need to reduce speed to avoid a collision. As such, Euro NCAP test protocols that encouraged fitment of speed limit information and warning systems from long before the mandatory requirements, required the warning to be visual and/or haptic and still retains that wording, despite the regulatory change making it impossible for manufacturers to comply with.
Solving the infrastructure problems on a nationally or internationally consistent basis is a major investment, though one that is likely to be needed if automated vehicles become a common feature on roads. However, on a local basis there is much more scope to influence this, and any local authority can improve the quality of its speed signage. This could be to make sure it meets national standards and is positioned and maintained to allow both human drivers and advanced driver assistance systems (ADAS) cameras on vehicles to be able to clearly see it. In addition, digital infrastructure can be improved. For example, in order to enable ISA on London’s buses, Transport for London (TfL) have developed their own digital speed map. It uses techniques such as extending a main road speed limit a short distance into a side road to avoid many of the problems of false positives and has been successfully used to implement a mandatory ISA on their buses with relatively few problems.
TfL also took a similar approach to their internal fleet of vans, introducing a mandatory ISA system using a private sector map supplier, but with modifications based on those successfully used for buses.
What the Research Shows
There is considerable evidence that drivers often exceed the speed limit. DfT speed compliance data suggests that many vehicles exceed the speed limit but that it varies by road and vehicle type. For example, 44% of passenger cars speed in a 30mph limit but only 36% of articulated HGVs do. However, in a national speed limit, 9% of cars are speeding compared with 36% of HGVs (which are subject to a lower speed limit than cars on these roads). In 2022-23, speeding accounted for 16% of all road traffic offences in Scotland.
There is a huge quantity of information establishing clear correlation between the average speed on a road and the frequency and severity of collisions. Much of this is discussed in more detail in the Safe Speeds section. This research shows a clear correlation between increases in speeds and increases in casualties but also shows a correlation with the variation in speed around the average figure.
Predictions of the effect of ISA that are based on these statistical correlations will typically produce high estimates of effect. For example, according to Carsten et al. (2008) and Lai et al. (2012), advisory, voluntary and mandatory ISA could avoid 2.7%, 12.0% and 28.9%, respectively, of all road collisions (not only in relation to the target population) in the UK. This became 50% when the effect of mandatory ISA on fatalities was considered. The highest effectiveness is in 30 mph (50 km/h) limited areas (mainly for pedestrian collisions). However, studies of causation in individual crashes tend to produce much lower estimates.
In reality, there are limitations to both analytical approaches, the causation approach may not adequately capture cases where cause was uncertain, or where cause was not affected but severity was. In research conducted by the Department for Transport reviewing 1,053 crashes across 38 police forces, it was found that over a third were deemed to be speed-related after investigation. Similarly, with the statistical approach, it is an association that is proven between average speed and risk of death, and that does not prove cause. Other factors associated with a higher average speed on a road may also influence the risk of collision and the distribution of speeds around the average is also associated with a change of risk. Roads where the increase is driven by a small number of very high-speed events may be different to those where a large number of vehicles are faster by a small amount. The truth is likely to lie somewhere between the two; and this remains with speed and the potential effect of ISA being very significant.
A study of the actual effect of the ISA retrofitted to TfL’s van fleet used telematics devices fitted to the fleet. It identified that, on average, there was a reduction of 62% in the number of speeding incidents per 1000 km travelled in vehicles equipped with the ISA system. However, somewhat counter to expectation, the system had no effect on the magnitude of speeding incidents when they did still occur, or on other driving characteristics sometimes considered to be associated with higher speed (e.g. harsh braking, accelerating or cornering). The sample size and time frame in this study was too small to directly measure the effect on casualties.
More evidence is needed on the effect of ISA implementation, in particular newer ISA systems compliant with the General Safety Regulations.
Challenges to Implementation
In the EU, ISA that provides information on speed, and warning if the limit is exceeded, will now become universal on new vehicles due to regulation. Technically, the UK has not fully adopted the regulations that enforce this, but manufacturers will not build a different system for the UK compared with the EU and if an SAS is not standard in the UK (or any other member country) then Euro NCAP will not include it in the rating for cars. As such, it is also very likely to be universal in the UK too.
Speed has been controversial throughout the history of road safety. There has always been a substantial sub-set of drivers who like to speed and do not wish to be constrained either by the limits themselves or attempts to ensure they comply with the limits. This can be seen in long running debates about the popularity and marketing of vehicles that can exceed the highest speed limit in the country by a factor of two or three.
It is clear from the evidence that speeding is an important factor in collisions and ensuring 100% compliance with limits would result in substantial casualty reduction. However, it is also clear that speeding happens all the time (e.g. 44% of cars speeding in a 30mph limit) and that collisions happen very rarely - in 2023, one casualty every 2.5 million vehicle miles and one fatality every 200 million vehicle miles travelled. Although across a population the effect of speed is clear, the lived experience of most individual drivers is that they can speed for a large proportion of their driving time without ever getting involved in a collision. So, it may be perceived as much less dangerous than the statistics demonstrate it to be.
ISA is no exception to the long-standing controversy and has faced criticism in the professional and mainstream media, despite surveys of the public and governments indicating much more widespread acceptance of the need to limit speeding. The flaws (e.g. high rates of false detections) in the current implementation of regulatory ISA have the potential to legitimise some of that criticism.
In terms of implementing further vehicle-based speed interventions, it is therefore very important to be able to understand the actual effect of current implementations of ISA on speeding; objective information on driver acceptance (and switching off of systems); and whether there are strong differences in performance from different vehicles. With that understanding can come the evidence to support improvements to current speed assist systems, with the twin aim of increased driver acceptance and, through that, improved compliance with the limits. Once the information and warning systems are working more effectively, then intervention systems could also be considered on a wider scale.
In the meantime, road authorities and digital mapping services could work with the vehicle industry to improve the current systems without any change to the passenger car fleet. Also, well-engineered ‘local’ systems (with good supporting infrastructure) could be a solution for certain fleets, following a similar example to TfL’s bus and van fleets.
Sources
Carsten O., Fowkes M., Lai F., Chorlton K., Jamson S., Tate F., & Simpkin B. (2008). Intelligent Speed Adaptation ISA - UK : final report. Department for Transport (DfT), London.
Department for Transport. (2023). Contributory factors in fatal collisions: Comparing STATS19 with post-investigation recording, 2021 data. https://www.gov.uk/government/statistics/contributory-factors-in-fatal-collisions-comparing-stats19-with-post-investigation-recording-2021-data
Dodd, M. (2022). Analysis of the effectiveness of retrofit ISA on TfL’s fleet vehicles (Project AVS161b). Transport for London; Apollo Vehicle Safety. https://content.tfl.gov.uk/analysis-of-the-effectiveness-of-retrofit-isa-technology-on-tfl-fleet-vehicles.docx
European Transport Safety Council. (2013). Intelligent Speed Assistance – Frequently Asked Questions. European Transport Safety Council. https://archive.etsc.eu/documents/Intelligent_Speed_Assistance_FAQs_2013.pdf
Highway Loss Data Institute. (2018). Compendium of HLDI collision avoidance research (HLDI Bulletin Vol. 35, No. 34). Insurance Institute for Highway Safety. https://www.iihs.org/media/7560e1bf-fcc5-4540-aa16-07444f17d240/Io54Pw/HLDI%20Research/Collisions%20avoidance%20features/35.34-compendium.pdf
Lai, F., Carsten, O., & Tate, F. (2012). How much benefit does Intelligent Speed Adaptation deliver: An analysis of its potential contribution to safety and environment. Accident Analysis & Prevention, 48, 63–72. https://doi.org/10.1016/j.aap.2011.04.011
Scottish Government. (2023, June 20). Road traffic offences. In Recorded Crime in Scotland, 2022–23. Scottish Government. https://www.gov.scot/publications/recorded-crime-scotland-2022-23/pages/13/
Taylor, M. C., Lynam, D. A., & Baruya, A. (2000). The effects of drivers’ speed on the frequency of road accidents (TRL Report 421). Transport Research Laboratory. https://www.trl.co.uk/uploads/trl/documents/TRL421.pdf
Theofilatos, A., & Macaluso, G. (2017, June). Intelligent Speed Adaptation, Speed Limiter and Speed Regulator. European Road Safety Decision Support System (SafetyCube), European Commission. https://www.roadsafety-dss.eu/assets/data/pdf/synopses/Intelligent_Speed_Adaptation_Speed_Limiter_and_Speed_Regulator_03072017.pdf
Transport Scotland. (2023). Contributory factors to reported road collisions. In Reported road casualties Scotland 2023. https://www.transport.gov.scot/publication/reported-road-casualties-scotland-2023/contributory-factors-to-reported-road-collisions/
With thanks to Apollo Vehicle Safety for their contribution to this content.
