John Lindsay, Reader in Information Systems design, Kingston University, UK
The British Government has recently published a consultative document on an integrated transport strategy with the intention of preparing a White Paper . By the time this paper is published that will have seen the light of day. It seems appropriate therefore to ask the question "in what sense might one talk of a transport system?"
It is easy to understand a motor car. It can be decomposed and assembled, it is manufactured and serviced, it has a life cycle, it fits into the world of its owner, personal and public. To scale from the motor car to its environment and structure that as a system requires a change in granularity, abstraction, scale and boundary. There is clearly a network of roads, a highway code, licensing, insurance, ambulances,police, yellow lines, yellow vans, taxes and service stations. But is it useful to consider this a system?
It is less easy to understand a bus. From a motor car to a taxi to a matatu to a minibus to an omnibus is a progression of changes in public and private spaces, of organisation and individuality, of competition and co-operation. There needs to be a move to systematisation. Routes,route numbers, origins, destinations, stops, timetables, interchanges, prices, users, (or customers or passengers). A bus within a bus company is an investment to produce a return on capital employed. It has a sense of a market and a plan. It is a system in an environment. It is engaged in competition and co-operation. It might be engaged in competitive tendering, service level agreements and contracts.
To understand a railway we have to see it as a system. Even the vehicle is a system. A train consists of an engine, carriages or trucks, running on tracks with signals. The there is the addition of routes, route numbers,origins, destinations, stations, timetables, interchanges, prices and users. The physical reality is less mutable. Overtaking requires very early planning.
To see a transport system we have to move to a third level of abstraction for now we must be considering the relation between land economics, employment, housing, capital investment and transportation, the whatness as well as the whereness. Cargo and people.
The Engineering Council, through the Institution of Civil Engineers, has produced a Transport Vision 20:20  as a mechanism for enhancing the profession of engineering but also to promote the role of engineers in policy formulation. Partly this paper is an engagement with the vision which they present.
So is there any useful sense of talking of a transport system? I'd like to show there is, but I'd like to go a stage further. I'd like to suggest that the private motor vehicle is not a sustainable medium of transport, but that a combination of walking and public transport is, what I call green transport, though I'm prepared to extend it to include cycling and even roller-blading. I'd like to suggest that the significant advantage of the motor car is that all you need is the A-Z and the AA Atlas and thereafter you are a self organising resource which can reschedule tasks to fit changed circumstances. The weakness of the walker with public transport is predominately one of access to information: if we can re-engineer the information we can re-engineer the system. To change the word of Negroponte , moving bits is cheaper than moving atoms, changing information is cheaper than changing reality.
It seems indisputable that the railway was a transforming technology of the 19th century. It is therefore worthwhile wondering whether its failure in the 20th was the consequence of being superseded by a superior technology, the motor car, and one should simply accept that; or whether a combination of political, economic and management factors was responsible for a defeat which need not have happened and which needs to be understood. If the latter may reasonably be shown to be the case then the possibility of restoring rail to an integrated system might be demonstrated. Simultaneously if the consequences of the triumph of the motor car arise from a set of political rather than economic or management decisions then it is possible to reverse engineer those political processes.
The Beeching report provides the opportunity to see the system model which resulted from the previous 100 years of development and argument on its consequences allow for evaluation. It could be of particular interest to information systems designers as he makes special reference to the absence of information "* and to method for obtaining it *. It is possible to run the film of history backwards before that in order to determine the roots of the argument, but we must limit ourselves to only two. In 1934 London Transport was formed with responsibility for buses, underground but not overground rail. In 1948 was formed the British Transport Commission, with responsibility for the whole of transportation, but the nationalised railways remained in a separate executive.
For an integrated transport strategy the point seems clear: the whole might be seen as a thing, but is then too complex to be managed. That might be; but it also might be that the political vision and perspective of the participants simply stops them from seeing what an integrative strategy might look like.
I'd like to suggest that the key is scale: the distances involved. A taxonomy might be inter-regional; inter-city; intra-urban. (There has been a key shift since the 1950s in the level of world trade and the movement of people that one cannot see this damp little island as a boundary.) Then it becomes clear that for the first air and rail with a minority of road becomes an option; for the second rail and road; for the third mass rapid transit.
There have been two impediments to integration: inter-modal transfer and interchange engineering. In both cases there is an interesting inter-relationship between physical reality, thinginess, and information. My suggestion is that by appropriately re-engineering the latter it is possible to affect the behaviour of the former without the same level of cost and that this demonstrates the system properties we need to develop.
Let us consider in detail one example, the representation of public passenger transportation in London [4a]. It enables us to start with one of the classics of design, the Beck underground diagram (see figure 1). By transforming mapping into representation he gave a shape to an emerging London Underground which was partly the consequence of the work of an American, Yerkes, thirty years earlier in electrification, frequent service, frequent stops, and names lines, colour coded with standardised logo and typeface. Which lines came into the system and which were left out was almost a matter of accident. His combination of horizontal, vertical and only a single degree of angle, of ticks for stations and circles for interchanges, the removal of all spatial referencing except the Thames made for an object which became immediately popular.
Figure 1: London Underground diagram of 1991, still based on the original design by Beck - scanned from the back of a Nicholson London Streetfinder
Now almost unknown is another representation of the same time. The whole of London's transport, rail, buses, trams, underground, country coaches, with public open spaces and parks, using a scale of 11/2" to 1 mile at the centre and 1/2" to 1 mile at the boundary with an x and y co-ordinate squeeze that produces a fisheye model stretching 35 miles from Charing X.
What is striking is that since Beck there has been no step forward in visualisation or representation of transportation. We are left with the question whether his construction of the Underground was the limitation of what he saw in graphic design, the codes, structures and notations simply would not allow for a more complex system, or whether it was a representation of the political and organisational vision of the time. Strikingly the attempt to integrated produced by Bernhard Slater on the basis of Green's formulation of Network Southeast uses a notation of section ownership rather than any attempt to build on Beck, and puts London Underground in black and white, while the London Transport version has the whole of British Rail as black and white with no attempt to show anything extra. Even within the Southern region sections, Network South Central represented its system in single colour with no attempt to show what went where, or intersections with Network South East, and Network South East did the same. Since privatisation the pattern continues except that they are owned by the same company, Connex.
Recently there was an attempt to badge London South Metro with a diagram giving the number of trains an hour passing through a station, but no indication of where they were going, and still only the journeys of the old Network South Central, which is of course the old London, Brighton and South Coast. Network South East is the old London, Chatham and Dover. In other words, despite grouping between the 1880s and the 1920s, nationalisation in the 1940s and then privatisation in the 1990s, the patterns of service and interconnection, of systematicity, are those of their foundation .
A private telephone conversation with the marketing department elicited that this representation was a logo, not an information system! So for the first time we have a marketing gimmick which takes the form of an information system, but is intended to disinform.
A development noticed during the writing of this note which might progress in an interesting direction is a new set of posters from what is now called Connex Metro. These show the direct connections from a particular station, for example Balham, with the number of trains an hour, but still only for the old London Brighton and South Coast. A private communication with the Managing Director assures me that the plan is to integrate Connex Metro. We will see.
Balham is an interesting starting point for it shows the disharmony of system design. If you go down from the overground to the Underground, you are presented with the Beck representation and can see the system as a whole. Though of course none of this would work without the A-Z, without which you wouldn't know where anything was on the surface or which station to travel to. But if for some reason which is hard to imagine, you were to arrive at Balham from the Underground and move overground, how on earth would you be supposed to know anything about where anything above ground went? There is a list of stations in alphabetical order, but a small list, giving train times, but the only indication of which train is a small headcode number which doesn't appear on the train or on the train indicator board. You have to listen to an announcement or read all indicator boards (which are of course on different platforms) to chance upon your station. It goes without saying that the station will tell you nothing about the buses which go near it or where they go.
So I'm left with the question, is the failure of public transport above all else the failure of its information system? And is the failure of the information a consequence of politics, not of representation? Might it be possible by re-engineering the representation of the system to change its perception and then its behaviour?
I started from Beck's diagram and wondered why it had not been extended using a similar notation. But in going back before Beck to see what it had looked like on the ground, I was struck by another point. On the ground, Kenton and Northwick Park look like an interchange, they are only a couple of minutes walk apart. Of course in 1934 they were owned by different railway companies. London Transport hadn't at that stage achieved that degree of integration. How many other places were there which could be represented as interchanges which were not? Seventy three was my conclusion. How would the behaviour of the system change if the interchange density increased by this number? But what is an interchange? A one minute walk? One train an hour?
My arrival at 73 needs a little elaboration. If you start from the A-Z  you will have some difficulty finding stations, and know nothing about what travels through them. If you start from the Ordnance Survey maps, you will find the overground stations and lines, and underground stations and lines when they are overground, but nothing else. Not even the names of stations. So the task was to find the names of stations for the map (mainly from the A-Z), the routes that travelled through them (from the Timetable ) by moving backwards and forwards between the alphabetical list of stations and tables and headcode numbers. Sometime after I'd started this, I found Jowett, book and map [7a] which gave historic representations of parts of the combinations, the latter at 1:50,00 and so an excellent fit with the Ordnance Survey. Had I found it a couple of years earlier part of this would have been simpler.
But remember that what I'm trying to do is prove the feasibility of green transport. So I had also mapped as I found them possible walking routes which would be pleasant, and features which are notable. This raises a difficulty: my utopia is your distopia. Three focus group meetings indicated that I was making reasonable presumptions. It also raises the whatness of where. The set I built, which satisfied my particularity, would not be generalisable, but that is where the significance of new information and communication technology will emerge and I'll return to this point. It might also be worth throwing in the environmental enthusiasms for birds and butterflies, unitary development plans, national cycle network and the London Walkers' Forum just to show that there is a real complexity here which is not entirely of my making.
Let us return to some of the difficulties of the existing information resources. A-Z distorts space in order to get in street names. OS difficulty in navigation in urban spaces. In both there is what I call the white roads and green spaces problem. The status of public rights of way and whether a path is navigable is always difficult. Both resources simply show a white road with no indication whether it is actually walkable. Both use colour to indicate ground cover, not whether there is actual access possible. London Transport produces bus maps. We return to the organisation and politics. London Transport Buses seems incapable even of showing underground lines with the coding of the beck diagram never mind telling us much about railways. In addition we have to return to scale, for to cover London there are thirty two of these objects which you'd have to collect. There is a single one for the whole of London but it is understandably difficult to use.
There seem three possible notations which are used in different part of the world (or of London). Morrison refers to two of them but not the third. He calls them Classic and French, the former when the route number follows the road or route, the second when a different colour is used for each . He doesn't mention what I'll call the connection-oriented, when roundels are used to represent interchanges and you follow the route from roundel to roundel.
Local authorities have some responsibility for transport and differently they have produced material as they have for walking, the environment and planning according to their political and organisational interest. There is no consistency. For the area I defined to make the problem manageable for myself in addition to 32 boroughs there are 10 counties now being re-organised through the restructuring of local government with also the prospect of a strategic authority for London after a referendum. Throw in the other design variables and there are thousands of items .
I have now experimented with a variety of visual indexes (see figure 2) which might present opportunities for achieving two steps forward, as well as giving me additional material for investigating information retrieval which one of my more technical competencies but would take us beyond the issues dealt with here except in one point: the nature of words in structuring systems .
Figure 2: One example of a set of a visual indices built in Powerpoint (draft – legend to be supplied?)
Interchange engineering I would suggest is an area of work which could be established more generally. From trial work with more than one hundred undergraduate and postgraduate students it seems that routes and stops can be comprehended but that interchanges present a much greater difficulty. Various ways of representing these as well as automatically calculating them have been used  which has the additional advantage of investigating the relations between machine processed information and the information processing activities of humans.
The second is Scale. From one stop to the whole world problem I think can be done in about five levels. For my London problem I used three and produced a series of diagrammatic representations.
So to return from the detail to the big picture. I suggest that thinking transport in a systems way allows us to think: Subject - object; public - private; system - environment; competition - co-operation; market and plan.
But it also allows us to ask bigger questions about the changing nature of work and location. Patterns of network behaviour - pole centres and networks. It is to these changes that public transport and planning authorities have not adjusted, and neither have the engineers with their vision 20:20.
There is a tradition of a systems view of planning, for example Chadwick , but there seems little of a systems view of transport. Bacon  suggests the idea of a simultaneous movement system in his work in Philadelphia. Architects and planners clearly have a role in this; so too should transport engineers and systems designers.
However the engineers in their vision suggest that street signs should be elaborated with representations of public toilets, apparently not recognising that they have in general been closed. Why not point to public transport (as is in part now done in Glasgow?) But to suggest something like that as a one liner is to not recognise the need for a systematic approach to information. The same document suggest the need for spending billions of pounds for drivers, on a combination of gsm, scoot, etc which is called an intelligent *. This presents information systems engineers with a marvellous opportunity to show how an integrated transport system might be built, simply by changing the representations.
But to achieve something along the lines of green transport we'll have to go further into the politics of transport information planning. To this end I have established a list email@example.com [15a]. It is also to this end that the proposed conference supported by the UKSS, the BCS and the EARL task group is directed.
 Was: Institution of Civil Engineers — Updated link: 'Professor has 20/20 vision'. In: IC Reporter
 Negroponte, Nicholas 1996: Being digital. Coronet Books / Vintage books.
 HMSO 1963: The Reshaping of British Railways (Beeching Report)
[4a]London Transport Maps (showing the Underground) 1933-2000
 Garland, Ken 1994: Mr Beck's underground map. Capital Transport Publishing
 White (?)
 Geographers' A-Z Map Company Ltd
[7a] Jowett, Alan 1989: Railway Atlas of Great Britain and Ireland (looks like out of print, cheapest Amazon 2nd hand copy costs €390.-)
 Ordance Survey 1:50,000 series 175, 176
 British Railways 1993
Morrison, Alastair 1996: 'Alternative information technologies for the provision of spatial information to public transport passengers in France, Germany and Spain'. Transport Reviews, 1996, Vol. 16, No. 3, 243–271.
 Lindsay, J. What shall we do?
 Lindsay, J. On what appears to be a very silly problem and What a grubby little object
 Lindsay and Haskins
 Chadwick, George F. 1971: A systems view of planning. Towards a theory of the urban and regional planning process. Pergamon Press
 Bacon, Edmund N. 1976: Design of Cities (Revised Edition). Penguin Books)
[15a] Formerly firstname.lastname@example.org — the web-based list archive (since 1998) is available at http://www.jiscmail.ac.uk/lists/TRIP.html