A distinctive advantage of proposed advanced transport systems such as PRT is that they will operate “on-demand” rather than “by-schedule”. Current mass transit systems such as bus and rail are by-schedule and are run according to a timetable regardless of passenger load. Taxis provide an on-demand service, expensive and of limited range, but not a mass transit.
Do we have to wait for novel technology or is it possible to have an on-demand mass transit now with existing technology?
How it would work
Typically, journeys would involve changes of vehicle so that passengers can be accumulated into larger and therefore more economic vehicles for the longer segments of their journey. Passengers would be equipped with a “Passenger Interface” which could be mobile phone containing GPS navigation and a custom app. There would be a “Journey Provider” which would negotiate with various “Transport operators”, co-ordinates all the journeys and instructs passengers via their Passenger Interfaces.
There would be strategically located interchanges where passengers would change from one vehicle to another, guided by their Passenger Interface. The Journey Provider will know where every passenger and every vehicle, is at any time. It will know where a particular passenger is relative to their delegated vehicle, use the Passenger Interface guide them to it and know that they are aboard before instructing the vehicle to proceed.
The Passenger Interface could be primed with the mobility characteristics of individual passengers. Most will be happy to walk for a few 100m, particularly if this reduces the fare for the total trip.
Journey example 1
A commute to work the other side of the town, would be initiated by the passenger perhaps 30 minutes before the start time. The passenger would already have the journey details, destination and required time of arrival, stored on their Passenger Interface. The Journey Provider has a fleet of different sized vehicles at its disposal, maybe: 4 seat taxis, minibuses with around 12 seats, buses with 20 to 30 seats and coaches with 40 to 60 seats. Most will already be on the move and the Journey Provider will know exactly where they are, where they are going and which seats are filled. At the request of the passenger the Journey Provider finds seats for one or more journey segments with scheduled interchanges. Typically the passenger is directed to a nearby pickup point where a number of passengers going to the same general area board. There would be numerous pickup points with one within 500m of every house, generally there would have space for a mini-bus. These are the preferred start and end points for journeys, however, for an extra fare, vehicles could stop at any front door. Passengers are taken to an interchange where they transfer to other vehicles which take them the rest of the way.
Journey example 2
Journeys can be spontaneous, for example, it is a sunny day and I decide to go for a walk in the countryside, about an hours journey time from my town home. I book the journey to the starting point of my walk and am given a time for pickup at one of the nearby pickup points. My Passenger Interface gives me a count-down to when the vehicle will arrive, so I can put on my boots and walk to the pickup point arriving comfortably just before the vehicle arrives.
Just as I am about to leave, emergency! Maybe it is an urgent phone call or my neighbour asks me to rescue their cat. I request a postponement of my journey and when the emergency has been dealt with I request a recommencement. I am given a new start time and off I go. The first section of the journey is on a mini-bus from which I transfer to a coach for most of the distance. All the rest of the passengers on the coach are going to another town, but I have been given a seat near the door and the coach stops at a village on the way where I alight. A taxi will take me from here to the start of my walk. I have accepted an offer of a pause in my journey in exchange for a reduction in the fare because I know there is a café in the village where I can have a coffee. This delay allows the journey provider to take another passenger part of the way.
The end of my walk is at a village. Starting a walk in an unfrequented place is not too much of a problem but finishing in such a place makes for difficulties in timing. Also, I can have a pint at the village pub while I wait for a vehicle that is being filled efficiently.
The longest section of the return journey is on a bus and, maybe as a result of that pint, I need to visit a toilet. I request this on my Passenger Interface and am dropped off soon after at a village where there is a public toilet. The bus does not wait, but I am allocated a seat on a bus that would not otherwise have stopped. This bus may be going on a different route, but the rest of my journey is rescheduled to join up.
Note that I was not restricted to a circular walk as is the case using ones own car.
A market for journey segments
An on-demand system would complement by-schedule transport, for example, by acting as a feeder to railways.
Mainline railways would be very difficult to run without timetables, but this often means that, on the more remote segments of line and at off-peak times, trains run severely under capacity or with a very low frequency of service.
On-demand operators would buy empty seats from the railway operator at the last minute and include them as segments on journeys that have been requested or even already started.
We have inherited railway systems lines that fan out from terminuses in major cities. The result is that the sections of line closest to the terminals are overloaded whereas the more remote sections are underused. Also, the networks do not join up efficiently, frequently requiring passengers to go into one terminal and out from another. This is inconvenient for the passenger and demands transport in the area of the city where it is most expensive to provide.
The terminus will not be the final destination for the majority of passengers who can be taken off at stations some distance outside the city, beyond the most overloaded sections of line, and to complete their journey using the infinitely flexible on-demand system.
Another advantage to the railway operators is that they could use the Passenger Interface to provide a more efficient ticketing system for ordinary journeys. Also, when breakdowns do occur, the on-demand system can be called on to deal with stranded passengers.
As things develop, on sections of line with light traffic, it may become feasible to run optional extra unscheduled services in gaps between scheduled services. These would not appear in the timetables and would only run if enough demand can be accumulated by Journey Providers.
Journey Providers do not have to own any of the infrastructure, rather they would act as agents in a market for journey segments, buying seats from taxi operators, coach operators and railway companies. There could be a choice of Journey Providers, who would trade parts of journeys. It would be a good idea to discourage monopoly.
Implementing
So far the only new infrastructure needed is the (far from trivial) software needed by the Journey Providers. This would be a development from the sort of software that is used by parcel carriers who can deliver huge numbers of parcels within 12 hours, keeping track of every single parcel throughout its journey. This should be do-able with passengers on a shorter timescale.
Although no new hardware is need to start with, as the system gets established more specialised infrastructure can be expected to evolve.
A specialised Passenger Interface
This would be a single function device, postcard sized and a few mm thick. It would have a low energy e-ink display and possibly just two buttons. Each seat would have docking stations in front of them in which the device sits during the current segment of the journey, keeping it in view of the passenger and fully charged. Alternative versions would be available, e.g. larger for partially sighted, audio for blind, braille for blind and deaf, etc.
Optimised vehicles
Most journeys will involve transferring from one vehicle to another. The layout of the vehicles will be optimised for easy boarding and alighting, particularly the smaller ones. Taxis will be of the London cab style with extra headroom and space to sit down without contortions. Minibuses will have fewer seats and a wider aisle. It may be possible to standardise the height of the floors of all the different vehicles so that wheelchairs and wheeled luggage are easy to transfer. Of course vehicles for shorter segments will be battery powered with charging points at interchanges.
Provision for luggage
Many journeys, such as shopping expeditions and over-night visits, will involve luggage. Some seats could have an associated optional luggage spaces, laid out so that the passenger can both park their luggage and sit down and then get up from their seat, grab their luggage and move on, smoothly and quickly, with the minimum of obstruction of the passengers. (i.e. not like overhead racks on aeroplanes and trains.) Some luggage spaces could be designed to take a trolley large enough to hold the weeks shopping or a large suitcase.
Fares
Can these be lower than a private car? The fares on existing road transport depend strongly on the size of the vehicle with taxis being considerably more, and long distance coach considerably less expensive than by car.
The perceived cost of a car journey can be debatable. The entire cost of running a car works out at about £0.40 per km, so, for the first example above of a single person commuting to work, this would be a reasonable figure to use.

Fig. 1: How far you go can travel on unsubsidised transport for £1 depends on the size of the vehicle.
To match this the journey would have to be made using vehicles with an average of more than about 6 seats, which could be achieved easily if most of the journey used mini-buses.
In the other example of a one-off outing the cost of owning the car may have been justified for other reasons. If two people went on this walk, the marginal cost of such a journey would be perceived as just the cost of the fuel, roughly £0.05 per person km. To compete on simple cost, most of the journey would need to be in a 50 seat coach.
The role of PRT
It will probably be more efficient to have smaller dispersed transfer sites rather than fewer larger ones. However, where the system interfaces with existing transport nodes, such as mainline railway termini and airports, the optimum locations for transfers are likely to be spread out to where there is less congestion and good connectivity. A PRT system will be ideal in this situation where passengers might be able to alight from a train with their luggage and board a PRT vehicle which takes them directly to their coach at a stop near a main road up to a km away.
A role for MAIT?
The MAIT Cabin is a box with no motive power or wheels. Instead, it has attachment features that enable it to be moved around efficiently by fixed machines in all three dimensions. It can be thought of as a lift (elevator) that can move sideways as well as vertically. It would be particularly suited to railway stations which generally have several levels.
Compared to PRT, it should be more compact and be able to handle higher passenger flows. For longer distances Cabins can be mounted on PRT-type vehicles.