Saturday, 22 May 2021

The Five Principles

Earlier this week I gave a talk to Cyclox about what I thought made good cycle routes. The talk covered the five principles for cycling infrastructure with a round up on Low Traffic Neighbourhoods.

Now, I have covered LTNs many times in this blog and so this week I thought it might be interesting to talk about the five principles. I am a fan of the details, but a little step back to look at principles is always a good idea as it helps us understand how things fit together. This post is essentially the long-hand version of the slides I used in my talk which you can watch here.

The five principles crop up in all sorts of UK cycling planning and design policy and guidance - even in the most dire of the genre! It is no surprise that they pop up in the new English design guidance, LTN1/20 Cycle Infrastructure Design, and they are covered in some detail in Chapter 4. It's also worth stating that in fact they equally apply to planning and designing for walking. The reason for this is that these principles major on the human experience of self-propelled travel.

I was probably vaguely aware of them in recent years, but it wasn't until I got myself of a copy of the Dutch "Design Manual for Bicycle Traffic" (often called the CROW manual - above) that I really started to understand how this all fits together. It also pushed me to imagine network level design being as important (possibly more important) than route design. In other words, a problem issue on a cycle route may be more easily addressed by looking at what is happening at the network level - something I talked about in by post about Lea Bridge Road in Waltham Forest where separate walking, motoring and cycling networks have been considered (below), it's just that they are on the same route for Lea Bridge Road itself.

The five principles are as follows;
  • Coherence (Dutch use Cohesion)
  • Directness
  • Safety
  • Comfort
  • Attractiveness
The CROW manual states;

In general it holds that if the minimum level cannot (or can no longer) be met for one (or more) of the five main requirements, then the infrastructure will need to be modified.

This is where we discuss what a compromise might look like, but first, we need to have a look at the five principles. The CROW Manual uses them first to define what a main cycling network should be (maybe over a 300m - 500m grid in urban areas), but as we will see, they can apply at both the macro and micro level.

LTN1/20 reminds us that accessibility should also flow through the five principles. Whilst that is absolutely true and a good thing to point out, accessibility should be implicit and integral to planning and design.

Coherence - the cycling infrastructure forms a coherent whole and links all origins and destinations.
People travel because they have certain needs to satisfy. It's obvious when you think about it, but so much cycling provision in the UK has been developed in such a way as to tuck cycling out of the way as a problem to be managed, rather than a mode to be embraced.

People need to get to work, go shopping, drop the kids off at school (or get to school themselves), get to a GP appointment or provide care assistance to others. They also want to travel for leisure or entertainment or maybe they just want to get out in the fresh air. Those reasons for travelling therefore make it an absolute requirement to be able to get from home to all of these destinations, or to link trips together. This means that we are going to have to make many roads and streets useful for cycling.

However, the cycling network is going to be different to the motoring network and the class of road or street shouldn't really matter, we need to be providing a seamless experience. 

Grove Road in Stratford (above) takes cycle traffic from the side road onto a main road. The side road has people cycling on the carriageway and the main road has a cycle track. The transition is pretty seamless because it forms a coherent route. People can cycle from quieter side streets onto a main road which features the places people want to get to. 

This includes integration with other transport modes, because people sometimes need to use more that one mode for longer trips. The S-Train in Copenhagen (below) has cycle carriages to help suburban dwellers to cycle to their local station and then around the city. The Dutch have integrated their rail system with cycle hire which allows people to conduct their destination business before dropping the cycle back at the station.

Gaps in a network reduce coherence. In a situation where cycling is well-developed, it's not a huge problem, but where small numbers of routes are relied on, then gaps can be fundamental fails in provision. One of the implications is that barriers need to be tackled such as crossing large highways, railways and watercourses.

This extends to planned and unplanned events. A well-developed network is far more resilient than a single route because it gives options.

Above is a closure for utility works on the CS2 route in Central London. There is little parallel or network provision in the area and such a closure both ruins coherence, might discourage people from cycling in the future and some people simply cannot dismount and use the footway.

Directness - the cycling infrastructure always offers the cyclist as direct a route as possible (detours kept to a minimum).
Cycling (and walking) is about people getting around under their own steam. Sometimes people have a little assistance and sometimes they need more assistance, but they are moving around their worlds at a human scale which is an entirely different proposition to travelling by motor vehicle.

Directness is about minimising the distance or time people need to travel and so can apply at both the macro and micro levels. Cycling absolutely needs to be quicker (in time rather than absolute point speed) than the car for it to be competitive and the way we design streets fundamentally affects this. 

Judd Street in Camden (above) is filtered at Euston Road which provides a direct access across to Midland Road and a pair of cycle tracks. The estate behind the filter is pretty quiet with motor traffic having to take routes on main roads. so the local cycling network is direct in the general area and specifically through the junction.

Contrast this with the A12 Colchester Road in Romford where walking and cycling share tight space and have to use a staggered toucan crossing. There is no wider network and so people are forced through an indirect route through this junction in both space and time. Of course, it's entirely possible that a slightly longer route avoiding a large junction might be longer in distance, but shorter in time. It's also worth remembering that traffic signals are a product of motorisation.

There is guidance on what is an acceptable diversion which might range from 10% to 25% extra distance compared with driving, but I would caution the application of this because it's easy to end up designing routes which are slower cycling than driving.

Attractiveness - the cycling infrastructure has been designed and fitted in with its surroundings in such a way that it is appealing or attractive.
This is probably the most subjective and difficult to define of the principles, but it is still very important. Take a look at Adinda-Flemmich-Straße, Freiburg im Breisgau, Germany (below). This is a street in the Vauban neighbourhood which is famously low traffic.

I would suggest that most people looking at this street would say it is attractive, but working out why might be a little more subjective. However, the attractiveness has been designed in from the start. The low traffic nature of the street is obvious and the designation as a "home zone" can pretty much be ignored. 

There are some nice materials on the street which don't dominate the scene. There is also lush planting which is nice to look at as well as changing the visual nature of the street to the few drivers needing access. The street is also given a sense of enclosure by the scale of buildings to give pleasing proportionality to the street. It's a place to cycle, but it is not cycling infrastructure as such, but the surroundings are appealing.

A smooth asphalt surface is attractive (above). I don't mean the colour, I mean the fact that there is no vibration imparted to the rider as they move (although colour can be part of the puzzle which makes something attractive through the "look and feel" of a place). Vibration is an assault on the senses and it detracts from the sights and sounds of the immediate world.

Attractiveness extends to having legible junctions which allows someone to make decisions before the decision point so that their experience is flowing. As soon as things become disjointed or confusing, navigation starts to become an irritation. It's also worth noting that attractiveness extends to the end of the journey. Can one cycle up (or into) the destination to park or does one have to search for a place to park which wastes valuable time?

Attractiveness is about escaping monotony. Cycling somewhere which has life (and can be lively), where there are things to look at be it the landscape or architecture. Personally, there are some post-industrial areas in my part of London which are interesting, if not aesthetically attractive. Marine Parade in Great Yarmouth (above) has a wonderful cycle track which has plenty of interest which draws one's mind away from thinking about the mechanics of cycling into what's going on in the immediate environment.

Safety - The cycling infrastructure guarantees the road safety and health of cyclists and other road users.
Safety is both objective with collision risk, designing out interactions with fast or heavy traffic and it is subjective in terms of how safe something feels both day and night (also called social safety).

Cedar Road in Romford (above) was filtered as an experiment. It runs parallel to the A12 and was used as a rat run by people trying to avoid traffic signals on the main road. At the time of the experiment, there were 1,920 vehicles a day using the street which dropped to 403 after filtering.

The street has parking on both sides and from a cycling point of view, having to face off with drivers who's attention was on beating the A12 was subjectively unsafe. In addition, having fewer vehicle interactions has improved the objective safety. The street now feels perfectly safe to cycle along without any protection and pollution exposure is far less than it would be cycling along the side of the A12. 

Safety is also about the details and not providing layouts which create risk to people cycling. The photo above is a cycle track at Main Road in Romford where a forgiving kerb has been provided which won't throw someone off if it is clipped and it won't catch pedals either.

Junctions are the highest risk in a traffic system because it's where different modes interact or end up in conflict. Good design helps us ensure that junctions are legible so that everyone understands the behaviours expected of them when passing through.

The roundabout on Provincialeweg, Vogelwaarde, Netherlands (above) has been designed to make crossing by cycle (and foot) simple and safe. The crossing point is set back on a decent width and rectangular island, crossing is in two parts and people only cross one traffic lane at a time. The roundabout is designed to promote low entry and exit speeds, so drivers have a bit more time to see people crossing. It's a safe design.

Let's compare it with this this roundabout on Buckden Road in Cambridgeshire. The crossing point is close to the roundabout. The crossing island is triangular and narrow at one end to help create a flare at the roundabout which promotes high entry and exit speeds. The crossing is in two parts, but it's over two lanes on the vehicle entry side and over a wide lane on the exit. This is an unsafe design; even if there are no collisions being recorded with people walking and cycling involved, it *feels* unsafe in use.

Comfort - the cycling infrastructure ensures that cyclists experience minimal nuisance.
Having to interact with motor traffic is a nuisance for people cycling. To be fair, the same could be said of people cycling from the perspective of someone driving. I don't mean that cyclists are a nuisance, it's just that in some situations, giving people their own clear space makes life far comfortable for everyone.

Gaasperpark, Amsterdam is the street which runs past Gaasperplas Station (above). It's somewhere which is away from traffic and where people walking and cycling have clear space. Out of shot, there is a drop off point, but there is no interaction. It's a comfortable place to walk and cycle with no traffic nuisance.

Having to mix with traffic is tiring. One has to be on constant alert and in many cases, this ends up having to try and think what a driver might be doing. This is not a comfortable experience and so minimising interactions with traffic either with cycle tracks or by building low traffic places is key. 

Exhibition Road in Kensington & Chelsea (above) comes from the opposite school of thought which essentially throws everyone together and expects the need to interact to be the controlling features. Unfortunately this is nonsense. Exhibition Road is one of London's "Quietway Links" (whatever that really means). In fact, in 2018, the street carried 8,757 vehicles per 12-hours (daytime) during the week and 7,316 at the weekend.

Despite the posted speed limit being 20mph, the 85th percentile speed in 2018 was 27mph. Looking at Figure 4.1 in LTN1/20, Exhibition Road is well into the area where some sort of protection for people cycling is required (or the traffic needs to be significantly reduced. The CROW manual in Table 5.3 is even more stringent - this street should have cycle tracks.

One useful snippet from the CROW Manual is that every stop someone cycling has to make means they need to use the effort equivalent to cycling between 75m and 100m to get going again. Go and count the average number times you have to stop on a shared path route with side streets and signals - that's your comfort being wiped out a stop at a time! You can also throw in minimising turns (on main cycle routes) because not only does that slow people down requiring energy to get going again, the chances of getting lost increase.

Storgatan, Malmö is within a central area of the city where motor traffic is either filtered out or heavily controlled. There are shops and other destinations that people cycling will want to visit and in doing so, a "cycling mind map" can be formed. Getting lost is uncomfortable and so good wayfinding is essential and this could be both using signs and streets which provide legible routes.

Smooth all-weather surfaces (which properly drain) enable everyone who wants to cycle to be able to do so all year round and they are therefore comfortable to use. Cobbles, unbound surfaces and hand-laid surfacing are not comfortable to ride over. 

The Five Principles – the balance
The quotation from the CROW manual at the start talked about the minimum requirements and these will be found in more detailed guidance. However, I think it's worth looking at the balance of the requirements in the round and it's worth acknowledging that some may be more important in some circumstances and compromises are not always a significant issues.

Gunnels Wood Road in Stevenage (above) is part of the towns famous and now rather worn separate network. It is objectively safe, although maybe subjectively unsafe at night. It is comfortable with smooth surfaces and gentle gradients. The problem is, though, that the network is not especially direct or coherent because apart from it taking ages to get anywhere (or at least it feels like that) and cycling is banned in the town centre. It's an attractive network to a certain extent, but it is so boring to cycle around.

Trying to deal with the balance or the compromise is the stock in trade for planners and engineers. This is pushed by space, time and very often politics. If we drop the bar for one or more of the five principles too much, we create gaps in the network, maybe not physically, but in how they feel in use.

A short section of reduced comfort and attractiveness might be acceptable if it remains safe, direct and coherent. The reduction in comfort might be a bit of a narrowing or even a section of shared path. However, constant compromise will start to become tiring and off-putting to use. If people can understand and see the end of the compromise, then it's going to be more acceptable. CS2 on Royal Mint Street (above) had building works over it for months, but the cycle track was kept open. It wasn't an attractive layout and the space was pinched a bit, but the extent of the compromise was both obvious in spatial terms and that people realised it was only for a finite length of time.

How above a layout which creates a slightly longer cycling route to avoid a large junction which reduces directness, but in being away from traffic and complication, the other principles are enhanced.

Sacrificing safety will discourage people. If we are dealing with a route then compromised safety will create a weak link with no work around (whether that's through design or a short term issues such as street works). People walking should not have to be drawn into compromises. If something has to give, it should be space for motors and then cycling space (subject to the other considerations). 

Bradford Street (above), had a pop-up protected cycle lane built last year as an emergency response to Covid where people leaving public transport needed alternatives. It's a pretty good scheme, but the bus stops are a compromise which impact on bus passengers. This is recognised by the City and it is hoped to be upgraded to a more permanent layout.

If there is something which doesn’t work at the local level, the solution is at the network level. In my talk, I mentioned some of the narrower streets radiating from the centre of Oxford. It may be that there are solutions where some roads are for general traffic and others are heavily filtered to prioritise cycling and bus access. Delivered alongside a network of Low Traffic Neighbourhoods, the City could help to do something about the 60% of trips which are driven into the City.

Conclusion - the acid test
In concluding my talk, I suggested five things which might be indicators that a cycle route is doing well and which are perhaps harder to capture than in numbers alone;
  • People of all ages cycling.
  • People using non-standard and adapted cycles.
  • People cycling side by side having a chat.
  • Mistakes by all road users are forgiven.
  • Children being children in perfect safety!
Looking at the photo of the kid cycling along the Embankment cycleway in Central London immediately tells you that the five principles have been well-considered. The route is coherent as it's obvious where it goes, the junctions are pretty good and it helps people get from A to B. This section is a direct line from the edge of the East End to the City and then Westminster.

The kid is objectively safe and he feels subjectively safe enough to be able to pop a wheelie. The surface is machine-laid, traffic is physically separated and the path is wide - it is comfortable. It's also an attractive route because there are famous sites to see including the popularly known "Big Ben" clock tower before the route swings to Parliament Square.


  1. Thank you Mark, that was interesting (and the YouTube talk). Cheers for your hard work, Sarah