By making the vehicles computer controlled - feasible, since the vehicles are running on elevated track, and therefore cannot collide with ground traffic - the driver costs are removed, and so is the incentive to pack as many people as possible into the one vehicle. This means that small vehicles are possible, with capacities typically ranging from one to four people. This then means that these vehicles can be individually routed directly from station to station, without any intervening stops, so that every vehicle provides an express service.
The stations are provided off the main track line, in an analogy to a bus lay-by. This enables non-stopping vehicles to continue past the station without delay. The track is laid out as one-way only, to minimise cost and visual intrusion. To go in the other direction, the vehicle is routed around the block.
The capacity of the system can be, in theory, very large. If the vehicle spacing is 0.1 seconds, then 36,000 vehicles can pass a point in an hour. If the vehicles carry four passengers each, this is a total of 144,000 passengers per hour. By comparison, a standard bus service would do well to move 600 passengers per hour.
Figure 1 shows the main features of the PRT system.

Figure 1: Main features of the PRT system - WM
The Heathrow system, using Ultra technology, which is currently being built, has 3.8km of track and 3 stops, so 1 stop every 1900 metres.
Both of these station intervals compare badly to buses, which have typical bus stop spacings of 240-310m. There is no reason why the station separations cannot be reduced to something approximating this for buses. The heathrow system is designed as it is, because it is designed to connnect discrete car parks to the main terminal building. Unlike buses, PRT can enter buildings in the same way that other monorails can.
Figure 2 shows a theoretical comparison of fuel consumption of Ultra against other vehicles.

Figure 2: Theoretical fuel consumption comparison (MJ/pass km - Source: ATS Ltd).
The Ultra system can be best seen in the following video:
There is no reason why PRT cannot be made to work. The design of PRT breaks no laws of mathematics or science.
The main idea behind using PRT in towns and cities is that it enables a solution to be found to the problems of congestion in the urban environment. In other words, instead of fixing the problem (by, for example, developing new systems of road transport which can compete with the car), a whole new infrastructure needs to be built. Supposedly, the PRT system will take a large number of passengers away from cars, although the 800m separation of stations in the Minneapolis proposal is not likely to promote the service to car drivers, who will expect the service, like their car, to be at hand. BAA's selection of PRT for the Heathrow expansion acknowledges this fact. The service runs from the car parks to the terminal building, not from the points of origin of the car drivers to the terminal building, because the PRT system could not compete with the car's convenience in bringing passengers to the airport. Indeed, if the car drivers could park closer to the terminal building, they probably would do so. PRT, thus, seems to have a niche role as a campus system.
Another group who are investigating PRT as a last mile solution are Hacienda. They also seem to have accepted that this is the correct use for PRT.
The second problem is trust. When a major development is built, the client trusts the consultants to do their best at specifying internal transit. The ownus is now on the consultant to specify something that is known to work, epsecially if the buildings and infrstructure are designed around it. The problem is that with PRT the numbers keep changing. The capacity of the system is a classic case.
The theoretical capacity of the system, of 144,000 passengers per hour along each piece of track, relies upon a very low spacing between each vehicle, and the vehicles being full. If 1.5 passengers occupy each vehicle, and if the vehicle spacing conforms to the UK Railway Inspectorate spacing of 2 seconds, then the capacity is reduced to 2,700 passengers per hour. This is still better than a standard bus, but well below what BRT, Metro or Tram can achieve. It is also below the capacity of a 3.5m bicycle lane, at 10,000 people per hour.
The PRT technology companies need to work our what their system is for, and they need to get their numbers
sorted out.