Discussion

Without counting the comparison sites where no statistically significant effects were observed, 69.2% (i.e. 108 out of 156 possible results) of the results across all 3-years of road trials indicated a statistically significant effect or positive trend for rider behaviour due to the installation of PRIME road markings. These can be broken down into 47.4% that were statistically significant (i.e. 74 out of 156 possible results) and a further 21.8% indicated positive trends in the data (i.e. 34 out of 156 possible results).

Overall, the results for the PRIME road trials provide strong evidence for a range of beneficial effects of PRIMEs on rider behaviour on a range of bends. Across all three key measures (i.e. speed, position and braking) significant effects were observed at different sites during the trials.

There were no instances of statistically significant increases in speed, dangerous positioning, increases in braking or decreased use of the PRIME gateways. These observations provide further evidence that PRIMEs did not have a detrimental effect on rider behaviour. As such, even at locations where no statistically significant effects were observed, PRIMEs were no worse than not installing them at all.

This means that there could be benefits of PRIMEs at these locations which were not sensitive enough during statistical analyses to illustrate significant effects and where further research could provide further insights into rider behaviour. 

Speed, position, braking and use of the final PRIME road marking

A discussion of the results for speed, position, braking and use of the final PRIME gateway road marking is presented below.

Speed

There was a statistically significant reduction in speed at 10 of the 22 trial sites. The majority of bends throughout the trials were characterised as sharp left-hand bends. Some were approached from long straight sections of road where motorcyclists were more likely to be carrying speed into the bend (i.e. Appin House south, Taynuilt, Runacraig south, Dunira). In addition, some of the approaches (i.e. Rob Roy’s Dip west 1 and west 2) were on undulating or downhill sections that could have exacerbated the issue of carrying speed into the bend. Other trial sites were more technical and involved more complex bends and limited views around the bends (i.e. Loch Lubhair east, Runacraig north, Bonawe, Salmon Draft north).

Statistically significant speed reductions were in the order of 3.69% to 8.06% slower when PRIMEs had been installed. These reductions in speed were typically around 1.73mph and 3.28mph respectively. Previous research has indicated long-term speed reductions of 3kmh (1.86mph) to 4kmh (2.49mph) for all vehicle types when PCMs were installed (Martindale and Urlich, 2010), and these results would seem to support that finding.

Speed data across the PRIME road trial sites were within the prescribed speed limits and comparable with national data indicating an average free-flow motorcycle speed of 51mph (Department for Transport, 2015). In Scotland the average speed of motorcycles on national speed limit single carriageways is 60mph (Transport Scotland, 2021).

In this research, across the three years, the fastest average speeds were observed at Kingshouse south (54.5mph), Taynuilt (48.9mph) and Butterbridge (48.9mph). These observations were as expected as these were the most open stretches of road. The slowest average speeds across the three years were observed at Appin House north (36.0mph), Runacraig south (35.0mph) and Pulpit Rock (34.7mph) which some of the more technically demanding bends investigated.

Position at final PRIME road marking and apex of the bend

In previous research, speed reductions have tended to be relatively modest with lane keeping more reliably influenced (Stedmon, McKenzie, Langham, McKechnie, Perry and Wilson, 2021, 2022). 

For motorcycle position at the final PRIME road marking, statistically significant effects were observed at 15 of the 22 trial sites. In all cases motorcyclists were moving closer to the centre of the road, into the PRIME gateway position, and riding ‘through the gap’.

This provides evidence that PRIMEs influenced rider position across a range of bends. In addition, where significant effects were observed they were sustained throughout all the PRIME data collection weekends and showed that PRIMEs had a lasting effect at these locations.

For motorcycle position at the apex of the bend, statistically significant effects were observed across 13 of the 22 trial sites. This indicated positive changes in road position with riders taking a wider line around the left-hand bends.

This provides strong evidence that PRIMEs continued to influence rider position after the final PRIME road marking and through the bend itself. In all cases the effects were sustained over the PRIME data collection weekends ranging from 6 to 8 weeks.

Braking

While braking was one of the key variables measured in the PRIME road trials, this did not appear to be a high incidence activity. The low incidence for braking, in itself, was a positive observation as it can be dangerous for motorcycles to initiate braking on a bend. Such behaviour can destabilise the motorcycle which then tends to become more upright and travel in a straighter direction going forward. This often accounts for motorcyclists losing control on bends. Research has shown that motorcyclists tend to brake first on a bend before adjusting their steering but will tend to swerve first on a straight road before braking (Stedmon, Crundall, Crundall, Irune, Saikayasit, van Loon and Ward, 2010). 

Statistically, significant reductions in braking behaviour (i.e. braking late, braking on the bend and total braking) were observed across nine of the 22 trial sites. 

Apart from Pulpit Rock and Middle Kames, these PRIME road trial sites were generally approached from straights into and sharp left-hand bends. It might have been expected that motorcycles could be carrying excess speed into bends and so a significant reduction in braking at these sites is a positive outcome of installing PRIMEs.

Use of the final PRIME road marking

Across 18 of the 22 trial sites statistically significant results were observed for increased use of the final PRIME road marking.

This provides strong evidence that motorcyclists were riding ‘through the gap’ at the final gateway marking and therefore in the desired position prior to the bend. These effects were observed throughout the PRIME data collection weekends illustrating sustained effects.

Overall, PRIMEs had a range of effects on speed reduction, lane position, braking and use of the final PRIME road marking. Taken together, these trials provide strong evidence that PRIMEs have a positive effect on rider behaviour.

Sustained and long-term effects of PRIMEs

Transport Scotland supported further research to investigate the nature of potential sustained effects for PRIMEs at the trial sites.

In the 2020 PRIMEs trials there was a two-week interval between the PRIME 1 and PRIME 2 data collection weekends. In some cases, effects were sustained over this two-week timeframe across the trial sites:

• Appin House (Northbound) – sustained effect for position at the final PRIME road marking
• Appin House (Northbound) – sustained reduction in speed; sustained effect for position at the final PRIME road marking; and longer-term effect at PRIME 2 for position at the apex of the bend
• Kingshouse (Northbound) – sustained effect for position at the final PRIME road marking; and sustained increase in use of the final PRIME gateway
• Kingshouse (Southbound) – sustained effect for position at the final PRIME road marking; and sustained effect for position at the apex of the bend
• Loch Lubhair (Eastbound) – sustained effect for position at the final PRIME road marking; sustained effect for position at the apex of the bend; sustained reduction in total and late braking; and sustained increase in use of the final PRIME gateway
• Loch Lubhair (Westbound) – sustained effect for position at the final PRIME road marking; sustained effect for position at the apex of the bend; sustained reduction in total and late braking; and sustained increase in use of the final PRIME gateway
• Rob Roy’s Dip (Eastbound 1) – sustained effect for position at the final PRIME road marking; and sustained increase in use of the final PRIME gateway
• Rob Roy’s Dip (Eastbound 2) – sustained effect for position at the final PRIME road marking; sustained effect for position at the apex of the bend; and sustained increase in use of the final PRIME gateway
• Rob Roy’s Dip (Westbound 1) – sustained reduction in speed; sustained effect for position at the final PRIME road marking; sustained effect for position at the apex of the bend; and sustained increase in use of the final PRIME gateway
• Rob Roy’s Dip (Westbound 2) – sustained effect for position at the final PRIME road marking; sustained effect for position at the apex of the bend; and sustained increase in use of the final PRIME gateway

From these results it was clear that over a range of measures, the effects were still present two weeks later.

In the 2021 PRIMEs trials there was the opportunity to extend the data collection interval up to six-weeks or eight-weeks depending on the trial site. This allowed the research to investigate if effects were still present after a longer timeframe than the 2020 trials. 

Across the trial sites the following sustained effects were observed:

• Taynuilt – sustained trend in speed reduction; sustained effect for position at the apex of the bend; and sustained increase in use of the final PRIME gateway
• Inveruglas – sustained reduction in total and late braking
• Runacraig (Northbound) – sustained increase in use of the final PRIME gateway
• Runacraig (Southbound) – longer-term reduction at PRIME 2 for reduced braking on the bend and total braking; and sustained increase in use of the final PRIME gateway
• Dunira – sustained effect for position at the final PRIME road marking; sustained effect for position at the apex of the bend; sustained reduction in total braking; and sustained increase in use of the final PRIME gateway
• Bonawe – trend in the data for increased use of the gateway across all PRIME weekends
• Landrick Bends (comparison site) – no significant effects as expected.

These results show that the sustained effects were uniform across the weekends with no apparent drop in their effect. In some cases, an initial effect was observed (i.e. a reduction in speed at some of the trial sites) and then a return towards the Baseline in subsequent weeks (but still below the Baseline levels). 

In the 2022 PRIMEs trials there was a further opportunity to investigate any sustained effects between six-weeks and eight-weeks depending on the trial site. Across the trial sites the following sustained effects were observed:

• Dailnamac – sustained effect for position at the final PRIME road marking; sustained effect for position at the apex of the bend; sustained reduction in total braking and a trend in late braking; and sustained increase in use of the final PRIME gateway
• Pulpit Rock – sustained effect for position at the apex of the bend; sustained reduction in late braking and a trend in total braking; and sustained increase in use of the final PRIME gateway
• Butterbridge – sustained effect for position at the final PRIME road marking; a trend in late braking; and sustained increase in use of the final PRIME gateway
• Middle Kames – sustained effect for position at the final PRIME road marking; sustained reduction in braking; and sustained increase in use of the final PRIME gateway
• Salmon Draft (Northbound) – sustained effect for position at the final PRIME road marking; sustained effect for position at the apex of the bend; and sustained increase in use of the final PRIME gateway
• Salmon Draft (Southbound) – sustained reduction in speed; sustained reduction in braking; and sustained increase in use of the final PRIME gateway
• Carrick (comparison site) – no significant effects as expected.

During 2022, Transport Scotland supported further research by revisiting a site from 2020 and 2021 to investigate the nature of long-term effects over one-year and two-year periods

The results from this additional work indicated that PRIMEs had a long-lasting effect on behaviour across all three of the key criteria PRIMEs are designed for (i.e. speed, position and braking).

Motorcycle distributions on the Trunk Road Network

With 32,213 motorcycles observed, classified and analysed, it was possible to provide some wider observations of the nature of motorcycling in Scotland during the three year period of 2020 to 2022.

Throughout the analyses it became apparent that specific patterns existed in the nature of motorcycling in the West Highlands. Across the PRIME road trial sites there appeared to be directional effects in motorcycle traffic through the day with clear examples of positive-skewed distributions (i.e. more motorcycles in the morning) and negative-skewed (i.e. more motorcycles in the afternoon). This is represented below (Figure 10).

These profiles indicate directional effects of motorcyclists ‘riding out’ in one direction in the morning and then returning, in the opposite direction, later in the day. This could represent local motorcyclists who were only riding out for a day and heading home later rather than weekend tourers and those on longer touring holidays.

When the distributions are plotted on the map of the PRIME road trial sites, it is possible to identify typical routes that riders might take in this area earlier in the day (Figure 11).

(arrows indicate direction of travel in the morning)

The yellow circles indicate popular motorcycle meeting points (Glencoe, Tyndrum and Inveraray) and it is apparent from the motorcycle distributions that motorcyclists tend to ride out to these locations during the day and then return later in the day (Figure 12).

(arrows indicate direction of travel in the afternoon)

From the analyses specific motorcycle flow patterns were identified in the West Highlands. Clear directional effects exist in motorcycle traffic throughout the day. For any given location it is even possible to pin-point which hour of the day most motorcycles are expected to pass by, and typical numbers. This information could be invaluable for emergency response logistics, enforcement/education initiatives, pop-up road safety campaigns, etc.

Further research questions

Throughout the course of the 3-year research programme, a number of important questions have been identified that require further research that would underpin an implementation phase for installing PRIMEs more widely for casualty reduction.

What are the effects of PRIMEs on untreated roads?

One of the benefits of Transport Scotland’s investment in this research is that each trial site was optimised (i.e. resurfacing, lines/signs, barriers, and roadside vegetation/furniture management) prior to data collection. This was done to minimise any environmental variables that could affect rider behaviour. However, in the future it is likely that PRIMEs might be installed on roads where budget constraints do not allow for such engineering works. For PRIMEs to provide the cost-effective intervention they are hoped to achieve, it is vital that further research is conducted at sites where the road surface and environment has not been brought up to the best standard. This will help identify what sort of condition roads need to be in to be suitable for PRIMES and the development of an installation guide for councils and road authorities to use in the future.

Can greater speed reductions be achieved?

At locations where there are faster approaches to bends and where speed reductions were not observed at existing trial sites, it is important to investigate if a different PRIME road marking design might have a stronger ‘priming’ effect on rider behaviour. For example, installing a series of five gateway markings instead of three gateway markings used at all trial sites to date. This will allow the research to identify sites where speed might be influenced through the installation of more PRIME road markings or if three is an optimum number.

How do PRIMEs affect right-hand bend behaviour?

In 2020 it was possible to investigate PRIMEs at four right-hand bends. However, for the main RST funded trials left-hand bends were the primary focus. In order to extend the evidence base of PRIMEs it is necessary to include further right-hand bends in future trials so that the scientific knowledge takes both left and right-handed bends into account.

Can PRIMEs communicate different hazards to motorcyclists?

The research to date has focused predominantly on left-hand bends. Some of the research in 2020 began to investigate right-hand bends and PRIMEs were shown to provide benefits. As well as reducing motorcycle casualties on bends, more research is needed on the use of PRIMEs for other hazards that motorcyclists are vulnerable to. PRIMEs could be installed on approach to rural side-roads where drivers of emerging vehicles might fail to see motorcyclists and/or fail to judge their speed accurately. PRIMEs could be installed to help initiate lateral movement prior to the junction and present an additional cue to drivers.

How do other road users react to PRIMEs?

By virtue of collecting the motorcycle data over the last three years, video data now also exist for every other vehicle at each site during the same timeframe. It would be possible to analyse other road users (i.e. cyclists, cars, lorries, buses, etc) to understand if their behaviour has changed due to the installation of the PRIME road markings. This is a very important question as an underlying aspect of installing PRIMEs is that they should not conflict with the needs of other road users. Although PRIMEs are designed specifically for motorcycles, if other road users derive some benefit from them then this also needs to be understood. Equally, if the PRIMEs have any kind of impact on other road user behaviour then early insights into this could help with design revisions.

Which site characteristics influence the success of PRIMEs?

Underpinning the current findings, a key research question is which sites obtain the most benefit by installing PRIMEs and where might PRIMEs be less effective so that even the low cost of installing them is not wasted. This could be achieved by on some form of meta-analysis of existing and new data combined with expert motorcyclist reviews of specific trial sites to identify any road environment factors that could inform future implementation and roll out of PRIMEs more widely.

Scotland’ Road Safety Framework and motorcycle casualty reduction

Transport Scotland recently published its ‘Road Safety Framework to 2030′ outlining a long-term goal for road safety where no-one dies or is seriously injured by 2050 (Transport Scotland, 2021). It proposes a ‘Safe Systems’ approach to road safety delivery as set out in the National Transport Strategy Delivery Plan (Transport Scotland, 2020). In relation to the concept of PRIMEs, the current research addresses the following pillars:

Safe speeds

Speed limits in a Safe System are designed for crash-avoidance and reducing physical impact. Key factors that should be taken into account in any decisions on local speed limits are:

• history of collisions
• road geometry and engineering
• road function
• composition of road users (including existing and potential levels of vulnerable road users)
• existing traffic speeds
• road environment (Transport Scotland, 2021).

With these factors in mind, PRIMEs offer a potential tool for supporting speed limits where roads have already been brought up to the best possible standard. With the observed reductions in speed and no statistically significant increases in speed, PRIMEs may therefore provide a means for maintaining safe speed limits rather than drastically reducing them. However, coupled with improved position on the road and reduced braking on bends this would appear to be supporting the rider experience more holistically rather than focusing on one specific measure of performance for safety.

Safe road use

Road users should pay attention to the road ahead and the task in hand; adapting to the conditions (weather, the presence of other users, etc.); travel at lower speeds; and give sufficient room to all other road users, no matter what their mode of travel (Transport Scotland, 2021). PRIMEs may provide motorcyclists with a tool that allows them to adapt their behaviour to the road environment and which other road users may also use as a cue for demanding bends and the presence of motorcyclists. In this way PRIMEs may help ensure that road users are risk-aware, can develop coping strategies for demanding situations, and act appropriately to keep themselves and others safe on the road (Transport Scotland, 2021). This was demonstrated by the positive results for road position both at the final PRIME road marking and at the apex of the bend.

safe roads and roadsides

The environment is designed to reduce the risk of collision and to mitigate the severity of injury should a collision occur. This can be achieved through design, maintenance and the implementation of strategies to reduce casualties on the roads (Transport Scotland, 2021). This can also be promoted through positive behaviours and safer sharing of spaces, the appropriate use of speed limits and signage that provides a much more affordable and sustainable way to protect the most vulnerable road users. PRIMEs provide a low-cost and easily maintained casualty reduction initiative working in harmony with other interventions such as bike-guard and other vehicle restraint system (VRS) solutions. They can be installed on existing roads quickly and efficiently or incorporated into road upgrade schemes. From the low incidence of braking across the trial sites, this would seem indicate that motorcyclists are generally set up well for these bends but that other effects on position and speed enhance safety further.

Across these strategic pillars PRIMEs have the potential to provide a new and unique contribution to a ‘Safe System’ approach. There is clear evidence from the research conducted over the last 3-years that PRIMEs influence rider behaviour and it is important to begin planning for an implementation phase of work and address further research questions that will underpin the roll-out of PRIMEs more widely.

The project consortium have identified representatives from a Local Authority in Scotland with an interest in installing PRIMEs on their roads. This would provide an opportunity to widen the scope of PRIMEs in Scotland while also providing also ideal testbed for trialling a PRIMEs installation process (i.e. a user guide for authorities and councils so they can install PRIMEs without the need for expensive research).