For a bus system to be truly accessible to elderly and disabled people, and to be attractive to car drivers, more is required than simply purchasing low-floor buses. The vehicle is only part of the system and the whole journey, from door to door, must be accessible and attractive.
The principle elements of the bus journey are:
Bus design has received much attention (not before time). Access to bus stops has in practical terms seen little work. However, the bus stop and its interface with the bus are beginning to be tackled, and it is this aspect which is dealt with by the CILT survey on which this paper is based. A number of innovative schemes and trials are in progress around the U.K. which seek to improve the stop environment and ease boarding and alighting from low-floor buses, often as part of Quality Bus Partnerships between local authorities and bus operators. The aims of these schemes include:
This paper describes a number of the trial schemes and their performance, particularly from the perspective of user needs. The five examples considered are in Birmingham, Liverpool, London, Northampton and Norwich. A number of the trials are in their early stages and detailed analysis of their performance will form a later part of the study. All the route enhancement projects represent a relative step change for local public transport, but they vary in their absolute levels of access, convenience and attractiveness.
In order for a bus system to be accessible, the interface between the bus and its stop must be as smooth as possible: "seamless" in the current jargon. Currently, various problems prevent the full realisation of this goal. These can include:
Clearly, some of these factors could be dealt with by management means, but by no means all; if it is impossible to align a bus parallel with the kerb at a given stop, no amount of management enthusiasm will alter the fact. Furthermore, there is currently an increasing realisation of the need to make bus services more attractive, particularly to motorists, and indeed to make them more like trams. The key factors here are:
A desire to improve bus stops and provide guaranteed level-boarding is not new. In 1979, before low-floor buses were considered feasible, the Swedish town of Halmstad began to build platforms at stops, to the height of the floors of the town's buses. Access to these platforms was by steps and ramps. Correct alignment of the bus with the stop is crucial for step-free access, but especially so with raised platforms as the potential for injury from falls is greater than with conventional kerbs. The original means of achieving this alignment in Halmstad was by means of wires buried in the road surface at stops. The buses were guided by the electromagnetic field produced by these wires into a correct position relative to the platform.
Unfortunately, the bus drivers tended to hold on to their steering wheels as they approached stops, lest the guidance system not function properly, which resulted in strain and shoulder problems. As a result, the system was soon abandoned in favour of manual positioning, which appears to have worked satisfactorily because the buses were also fitted with hydraulic bridging plates which reached to the platforms. These plates were needed because the buses retained conventional steps for use on routes without raised platforms. There were, however, functional problems with the plates, particularly in winter, and the cost of the specially adapted Volvo buses and of providing platforms meant that progress with the rolling programme of constructing platforms at all stops was slow. In 1991, with the latest special buses only recently arrived, the decision was taken to abandon the system in favour of low-floor buses.
Curitiba (Brazil) uses high-platform level-boarding of double-articulated buses, where ramps extend automatically from under the bus doors. This operates in a very different climate to that of Halmstad, where cold winters affected the hydraulic system. Boarding and alighting in Curitiba is also speeded up by passengers purchasing their tickets from a member of staff seated at each stop. This person also operates a lift to allow disabled access to the platform (URBS, 1995).
Clearly guidance systems offer the most effective docking tool. Guided systems can be designed to give very precise alignment quickly, whereas relying on the extension of a ramp (or even a manual ramp) can have a time penalty and may not provide true step-free access for people who do not appear to the driver to be 'obviously' disabled, where use of the ramp is at the drivers' discretion.
The potential is today being used to its full with most newly built guided public transport routes. Trams are guided by their rails and can use either ramps (e.g. Grenoble, France) or flexible suspension (e.g. Paris) to ensure a constant step-free access despite variations in the height of the vehicle (due to passenger loadings and wheel wear). Discussion of the issues surrounding tram access can be found in Wood (1994) and Wood & Peck (1995).
Guided bus technologies also allow for close alignment. These fall into three categories:
Some of these technologies have potential for assisting bus-stop docking on non-guided systems. The use of electromagnetic guidance (and its problems where not used on the entire route) in Halmstad is detailed above, and a lateral guide-wheel forms part of one of the U.K. trials, that in Northampton (see below).
Other techniques for achieving correct alignment work on the bus tyres. Caen has undertaken experiments with the use of metal and rubber tubes along 280-mm-high platform edges, against which the tyres rub (combined with a visible guidance strip). Whilst the system gave generally positive results, average horizontal gaps tended to be greater with the guidance system at the raised kerb than at an untreated kerb (Lesauvage et al., 1994). Driver concerns about impact when approaching the platforms (which is partly what the tubes are designed to prevent) should perhaps be supplemented with questions of the maintenance and safety implications of wear on the sides of the tyres. It is worthy of note that Caen is now investing in the "Guided Light Transit" form of guided duobus.
In Germany, the city of Kassel has produced a concave-section kerb stone which guides the tyre in a more subtle manner. As the tyre rides up the concave surface, gravity pulls it back down. This is now being produced under licence in the U.K. and is being employed in London and, particularly, Birmingham (see below). It is claimed that tyre wear is less with the 'Kassel kerb' than with ordinary kerbs.
Uster et al. (1994) recommend an information-based system to tell the bus driver how close the vehicle is to the kerb or platform, leaving control in human hands. This is based upon a survey of guidance technologies (including the Caen trial) and a video survey of bus alignments in bus lay-bys and at bus boarders in Grenoble. It is admitted, however, that ramps may still be necessary to bridge the gap between bus and kerb or platform.
The value of bus boarders (also known as 'bus capes' and 'extended bays'), where the pavement or a platform is extended out into the carriageway, in allowing buses easier approach and departure, and reducing the likelihood of obstruction by parked cars, is well documented (see for instance European Commission, 1995; London Bus Priority Network Steering Group, 1997). In terms of public transport priorities and ease of docking, boarders seem in general to be the best option, followed by unextended kerbsides, with lay-bys the worst of all worlds, as they fill with parked cars, make alignment more difficult, delay egress and take more space away from pedestrians and waiting passengers. Drivers often fail to pull in at all to lay-bys, and even kerb-line stops, in order to maintain a position in the road suitable for a rapid return to the traffic flow. Similarly well documented is the need for straight kerb lines at stops, so as to avoid gaps between the kerb and bus.
The layout of stops is clearly crucial to people's ability to use them. Useful guidance, with detailed dimensions, is given in Barham et al. (1994) and European Commission (1995). Key requirements are:
A number of research projects are also underway at University College London's Centre for Transport Studies aimed at finding optimum bus stop design.
Birmingham's "Showcase" Line 33 is the first of a series of 15 enhanced bus routes in the West Midlands, based upon Quality Partnerships between the West Midlands Passenger Transport Executive (Centro), local authorities (here Birmingham City Council and Walsall Metropolitan Borough Council) and bus operators (here National Express Group subsidiary Travel West Midlands). Line 33 uses 15 dedicated low-floor, kneeling buses from Volvo/Wrights. These do not have ramps fitted as the docking arrangements at stops should make them unnecessary.
The entire route, from Birmingham city centre north to Pheasey, has been specially treated, with priority at traffic signals, 10.5 km (40% of the route) of 24-hour bus lane, new stop furniture (with clear information, see below), bus stop clearways and a platform raised to 180 mm above the carriageway at every stop. Buses line up correctly at the stop by means of the special edging, formed of 'Kassel kerbing'. There is also a strip of 'lozenge' tactile paving to warn of the platform edge.
All drivers receive disability awareness and customer service training. However, a potential problem is identified in that it is the longest serving drivers (who can potentially be set in their ways) who get to drive on 'prestige' routes, such as Line 33. It is understood that drivers tend only to dock their buses fully when a wheelchair-user or someone with a buggy wishes to board or alight. CILT's observations (in the company of Centro and T.W.M. management) showed good alignment.
The 87 stops have four levels of quality, a new pole without shelter, a cantilevered shelter, a standard enclosed shelter or a large "enclosed plus" shelter. Shelters cost in the region of £5000 each and the average total expenditure per stop has been approximately £10 000. Selected stops have further features, including real-time passenger information (at 20), a bus boarder and specially designed seats for young children. The bus boarder is full-width and does not allow other traffic to pass a bus serving it. Unfortunately, pressure has been brought to bear and it will be cut back to half-width to allow cars to pass.
The service operates on a daytime frequency of eight buses per hour Monday-Saturday (two per hour after 19:30; three per hour daytime on Sundays). The timetables on Line 33 are of the new style Centro is introducing, which follows continental practice of listing departure times for each hour, with approximate journey times, rather than giving complete timings for the entire route, as is usual. This makes for more rapid orientation and thus makes finding the necessary information quicker and easier.
Surveys so far show a 27% increase in passengers on Line 33. Earlier on-bus surveys, presented in the 'West Midlands Balanced Package' bid for 1998/9 (an annual application for central Government funding and credit approval), covering periods before and after Line 33's opening, up to the summer of 1997, show that 80% of new passengers had previously used another bus service, 10% car and 3% taxi. 14% of respondents specifically cited easier access as being a factor in this. Also, a comparison of June 1997 with the same month in 1996 shows most bus routes using the corridor have had journey time reductions.
The second Showcase route opens in June, between Walsall bus station and the Mossley estate, shared between Travel West Midlands and Choice Travel. The next route is planned to link Coventry city centre and Longford, shared between T.W.M. and Stagecoach Midland red.
Merseyside is also to have a series of enhanced bus routes: the SMART bus network, being developed by the P.T.E., Merseytravel. The initial set of routes (three low-frequency daytime services (SMART 1,3 & 4), and an evening and Sunday service (SMART 5), started in 1994) is not commercial, and originally received money from the E.U. J.U.P.I.T.E.R. research programme (looking at ways of reducing energy consumption in urban transport), and is now supported by Merseytravel. The subsequent routes are commercial operations, implemented as Quality Partnerships, with funding from Central Government's Capital Challenge scheme. At the time of writing, one commercial SMART route, Merseybus' services 12 & 13 along the West Derby Road to Stockbridge Village, is in operation, with the next, on the Wirral, beginning in June. The current pace of development is approximately one route (or linked set of routes) each year. One is planned for St. Helens in 1999 and one to the north of Liverpool for 2000. Merseytravel is aware that its community consultation regarding SMART routes has not been sufficient, but is improving this starting with the Wirral, via its new Community Links Team. Merseytravel also now has a disability advisor, who is himself a wheelchair user.
All stops on a route are treated, with new, distinctive yellow furniture and, at selected stops (49 of the 78 on services 12 & 13), real-time passenger information. There is a policy of avoiding placing real-time information in the most vandalism-prone locations. Platforms are not raised, and no guidance technology is used, but the buses are fitted with driver-controlled ramps to facilitate step-free access. The driver triggers the ramp if they see someone in a wheelchair, with a buggy or who appears for some other reason to need it, or if a passenger requests it. All drivers on the SMART routes receive disability awareness training and are instructed to kneel at all stops.
Buses on the initial, supported routes are from the German manufacturer Neoplan, but those on services 12 & 13 are Scania Axcess. The latter services (which run in opposite directions on a terminal loop in Page Moss) operate on a 12 minute daytime headway, giving ten buses an hour for the trunk section (two per hour, four on the trunk section, in the evening and on Sundays).
Merseytravel's Mal Ferguson (personal communication) considers that the SMART project is a cheap way of delivering improved service and some degree of modal shift. The West Derby Road corridor cost Merseytravel £1.8 million, and Merseybus £2.4 million for the 20 buses. Further costs were paid by the highway authority, with support from the E.U.'s Objective One fund.Estimated from revenue, the SMART routes are carrying approximately one million passengers each year. Between February 1995 and November 1997, SMART 1's patronage increased from just under 10 000 passengers per four-week period, to just under 16 000. Over the same time, SMART 3 & 4 increased their joint patronage from 1000 passengers per four week period, to 4500.
Of those passengers using SMART 1-5 in early 1996, 1% were wheelchair users, 0.4% escorting wheelchair users and 1.6% had a baby in a pram or buggy. Previous forms of travel were also surveyed: 84% by bus; 3.5% by car; 1% by train; 1% by taxi; 1% by foot; 4.5% new journeys due to the new route; 5% new journeys due to the SMART features.
One of the central bus stations, Queen's Square, has recently been rebuilt, with saw-tooth docking. Here, buses are required to wait if their designated bay is occupied by another bus, and only allow alighting and boarding to take place once properly at their bay. Elsewhere, this is not required, so that SMART buses do not necessarily always stop at the correct position vis à vis the stop or shelter. For instance, at the in-bound Lime Street stop, SMART buses were observed by CILT to stop at the wrong end of the shelter (away from the exit gap in the road-side wall) during one observation and to drop passengers away from the stop during another.
Trials are currently in progress in at least six London locations to test the performance of 'Kassel kerbs' (see above). The first such kerbs were brought from Germany by the London Transport Unit for Disabled Passengers and London Transport Buses, and interested Boroughs (as the highway authorities) have installed them. There is, however, no overall policy in relation to their use at present and the installations are of varied design.
The kerbs are raised to between 160 and 180 mm, usually on half-width bus boarders. Otherwise, the stops are not (at present) altered, so that pre-existing furniture remains and some have bus stop clearways and some do not. The 160 mm kerbs are preferred, as the higher version blocks the use of the ramps mounted under the floors of London's low-floor buses. 180 mm Kassel kerbs are only considered suitable for use on full-width bus boarders, where buses can pull in fully, leaving a vertical gap of c. 60 mm, with which many (but not all) wheelchair users can cope. Drivers are generally encouraged to use the kerbs properly, although some are told specifically not to when driving particular bus designs, where damage might occur. A difficulty of the London situation is that the tendering system for bus services is not used to give sanctions on driver behaviour via bus operating companies. Where used properly, the system is reported to be working satisfactorily.
CILT's observations at the installations in Walthamstow (London Borough of Waltham Forest) and West Drayton (London Borough of Hillingdon) show few drivers actually pulling in properly (except for some low-floor buses on service 222 at West Drayton), even where not excluded from doing so by bus design and bad parking. Both these cases have half-width boarders with loading/parking bays either side, vehicles occupying which prevent the buses from achieving an adequately straight run into the stop. In Walthamstow, during the random half hour observation, a car and a lorry parked alongside the Kassel kerb.
CILT's surveys of users of another London low-floor bus route (the 101 in East London) show drivers making fairly good use of the kneeling facility, and appreciation of the easier access by passengers, but the need for better kerb treatments and bus shelters along the route (Bell, 1998).
The cost of providing a boarder and kerbing, including drainage work, is estimated at approximately £10 000. L.T. Unit for Disabled Passengers' Andrew Braddock (personal communication) points out that equipping all of London's bus stops in this way would cost about £170 million, which compares extremely favourably with the £3000 million cost of the current Jubilee Underground Line Extension.
In Northampton, a Quality Partnership, involving FirstBus subsidiary Northampton Transport, Northampton Borough Council, Northamptonshire County Council and British Gas, has enhanced a town-suburb route, service nos. 21, 29 and 79, with new stop furniture, amended kerb lines, bus stop clearways and low-floor, kneeling buses (Volvo/Alexander) powered by compressed natural gas (interestingly stored in composite fibre tanks, which have recently been deemed safe, allowing greater flexibility than the traditional steel tanks). The buses are also equipped with transponders in readiness for traffic signal priority.
Three stops are of particular interest in that they have had their boarding platforms raised to 180 mm to allow level boarding. The buses are fitted with horizontal guide-wheels (as used on guided bus systems) on the near-side of their front axles, which run against the upstand of the platform edge, so aligning the bus correctly. In addition, three new Scania diesel buses have been acquired for a further route and have been fitted with guide-wheels. After some teething troubles with the guide-wheels (including catching on some kerbs of 140-150 mm height), the system is now reported to be working as it should.
According to Volvo Bus, the system can be used to full advantage at any stop, including lay-bys, where approach at the angle normally employed in a lay-by is possible; whilst full parallel approach is ideal, it is not necessary. The stop itself should have a straight kerb of 8-12 m length. If drivers are careful, it is possible to position the bus such that the horizontal gap is down to 10-20 mm. Guide-wheels add an estimated £500-600 to the cost of a bus.
The partnership appears to have increased patronage since its opening in June 1997, but Northampton Transport awaits year-on-year figures to confirm this. It is the first of a number of corridor-based schemes and cost a total of £1.3 million, £200 000 of which covered transponders, kerb work and the few new shelters acquired. The buses cost c. £150 000 each, compared with c. £125 000 for the equivalent diesel. The partnership agreement stipulates that a ten-minute headway will be maintained at least for the first twelve months.
A Quality Partnership has been entered into by Norfolk County Council (and Norwich City Council as its highways agent as far as the City boundary) and FirstBus subsidiary Eastern Counties, covering the Dereham Road ("Western") corridor. The scheme has received funding from Central Government's Capital Challenge programme.
The first phase opened fully in February 1998, with the western half of routes 19 and 20 turned into "Premier Buses", using new low-floor, kneeling Dennis/Plaxton vehicles on a ten minute headway for the trunk suburb-city stretch (assisted by a new in-bound bus lane, 24-hour bus stop clearways at some stops and planned traffic-signal priority), with all stops treated. The treatment varies between a completely new, raised stop with capacious shelter, to just a new flag placed on existing street sign-posts (particularly in the rural fringe of the outer end of the route in Costessey).
Most stops have a strip of non-slip paving and those where the kerb height was less than 100 mm have been raised to approximately 150 mm over a distance of five metres at the head of the stop. Detailed design was undertaken in consultation with the City Council's Access Officer and bus drivers, who reportedly like the paving strip as it provides a target for them. There are some stops which cannot genuinely be called step-free. All stops remain in their original positions relative to the carriageway without modification to the kerb line, so that shelters occupy most of the width of narrow pavements at a number.
The buses are fitted with ramps triggered remotely by the driver (Plaxton does not supply manual ramps, which Eastern Counties would have preferred on grounds of cost), which fold out from the top of the floor. No guidance system is in use, and CILT's observations at the time of writing are that drivers do not always pull in properly to the kerb, nor kneel the bus if people who do not appear to 'need it' are waiting to board. Guidance to drivers is to position the bus as close to the kerb as possible and to use the kneeling facility generally, although not consistently at every stop, but rather "if in doubt, use it". At the time of writing, all drivers had received a day's general customer care training, but had not yet had their planned disability awareness training, as it was not feasible to programme in sufficient time out for the sixty drivers prior to the start of operations.
Phase two is being implemented during May and June 1998, with services 18, 21 and 22 (serving the suburb of Bowthorpe) feeding into the trunk corridor, which will have a bus every five minutes during the peak. These buses are Scania/Wrights vehicles, fitted with manual ramps, as here there was a price differential.
Later this year it is planned to introduce traffic-signal priority and real-time passenger information displays at most of the stops with shelters. The total cost of the infrastructure works is £2.5 million and an average of just under £1000 per stop has been spent on kerb works. At the time of writing, it is too early to identify the scheme's effect on patronage, although weekly patronage on the whole of service 19/20 is up 10% compared with 1997, which may well be due to the increased frequency. The aim is to increase patronage in the corridor by 30% over two years.
A further partnership has been agreed between Norfolk County Council and Eastern Counties for King's Lynn, if sufficient funding becomes available; otherwise, work will concentrate on consolidating the Norwich scheme by extending the 'Premier Bus' treatment to the other ends of the bus routes involved. Eastern Counties' Commercial Director, Mike Payne (personal communication), emphasises the importance of a formal agreement in the Quality Bus Partnership to ensure delivery in the agreed timescale. The formal agreement is good for all concerned as it is a commitment that has to be given priority.
Whilst the projects described in the case studies above are generally at an early stage, a number of points of interest emerge. Some of these are related to the detailed design of installations, some to the use which drivers make of them, and others to the suitability of the technology itself.
It is clear that installations designed to ease bus alignment with the kerb at stops must be positioned to allow buses a suitably straight run-in (although guide-wheels, as in Northampton, may allow this to be eased). Whilst half-width bus-boarders improve the situation, they do not provide a truly adequate solution, as buses are frequently prevented from approaching at a shallow enough angle by the presence of parked cars and lorries. Similarly, illicit parking is not deterred (clearways aside) if the parked vehicle is clear of the main traffic stream. If a kerb has been raised to provide level-boarding, then it is all the more important that all buses serving that stop pull in properly, as the step down onto and up from the carriageway is greater, as is the risk from falling. If non-suitable buses are serving the stop prior to raising, they need to be replaced. This emphasises the need for strategic co-ordination and control, or comprehensive partnerships.
Full-width boarders actually take up less kerb length than half-width boarders, and less again than either straight kerbs or lay-bys (London Bus Priority Network Steering Group, 1997), so that more space is actually available for parking, or for longer stops to accommodate more buses. The reason for not extending boarders out to the full width of the parking lane is that other traffic will have to wait behind buses when they serve the stop; the pressure on highway authorities to avoid this is great, as is shown in the Birmingham case study. However, this attitude assumes that it is vehicle flow which is important, rather than flow of people. Buses are much more efficient people carriers than cars and speeding buses up at stops and encouraging their use actually increases the flow of people even if vehicle flow is reduced.
Linked to this is the fear of congestion and "gridlock". However, allowing buses to stop traffic flow is a form of vehicle capacity reduction, which has been shown to promote traffic reduction, rather than increased congestion (Cairns, et al., 1998). If buses are genuinely to be given priority, they must be accorded the prime position in the street, and cars will have to yield to them rather than the other way around, as is the case at present. The actual delay is reduced where boarding and alighting is rapid, as with systems using on-board conductors, boarding through multiple doors, on-board ticket machines or a high proportion of off-vehicle ticketing.
Whilst systems without raised platforms at stops (with or without bus boarders) are cheaper, they depend upon ramps to provide true step-free access. However, it would appear that ramps are still needed to provide step-free access for all even at kerbs raised to 180 mm, which suggests that kerbs raised to the same height as that of the kneeled bus floor (c. 240 mm) would be appropriate (as long as a parallel approach for the bus could be guaranteed.)
Clearly, manual ramps are likely to be used less than those which the driver can operate from their seat, especially as some passengers are likely not to want to trouble the driver. All systems where the driver's discretion is employed, to pull fully into Kassel kerbs, to kneel their bus or to extend a ramp, depend upon the driver being aware of need. Whilst a needful passenger may request the ramp, they are unlikely to be able to request the bus to pull in differently once already at the stop. The driver is likely to spot wheelchair users or prams at stops, but some ambulant disabled people still cannot cope with a large horizontal gap. In most cases with a clear run-up this is not likely to be a problem, but it will in some. Furthermore, the driver can be less aware whether or not a person who needs the ramp, or full parallel alignment, wants to alight from the bus. It is also possible for someone who needs the ramp to appear at the stop once the bus is already stopped, especially if it is waiting due to scheduled lay-over, connecting services, driver change-over or early running. Kneeling consistently at every stop has been shown to reduce rather than increase dwell times (Ståhl, 1990).
Whilst most of the case study schemes have not been operating long, it is clear that they represent a relative step-change in local public transport provision. Naturally, those responsible for the route enhancement schemes are monitoring their workability and effects on patronage, and some are considering the actual degree of improved access provided. However, it would be useful for others contemplating bus innovations to know which systems work best especially from the perspective of disabled passengers and pedestrians (passing stops). CILT will be contacting access groups in the case study cities as a later stage in this study, once the new schemes have had a chance to settle in. It would be useful for those responsible for docking guidance system installations (Kassel kerbs and guide-wheels) to survey the actual gaps obtained between bus entrances (including any centre and rear doors) and the platform edge.
From the above case studies and discussion, certain basic requisites can be identified for a truly step-free interface between the stop and the bus:
In addition, it seems clear that formal quality agreements are essential to the proper integration of facilities and the co-ordination of implementation to common standards and deadlines, especially where there are several agencies involved in providing accessible routes. This is particularly true in London, where there are two departments of London Transport, multiple Borough highway authorities, the Traffic Director for London and multiple operators likely to be involved. Similarly, management of the installed facility needs to be subject to agreement, so that, for example, carriageway resurfacing does not reduce the kerb height.