Bill Turnbull's assessment of utilising capacity is a good one. If any system claims to provide travel for the future, it should make best use of its inherent capability. This is not achieved using single vehicles at large headways.

To argue that one might build more length of an under-utilised system later, to make up for its capacity deficit, does not make economic sense.

Further, to dismiss platooning as yielding unacceptably high guideway loads is to miss a vital structural design case. Should a "brick wall" stop happen on the guideway, there will be a significant length of vehicles stopped in close proximity with each other. There is no guarantee where this would happen. The support structure would thus have to be designed to withstand a live load consisting of a continuous train of vehicles loading various spans of guideway, with a substantial factor of safety built in.

Coupled platooning offers the opportunity of providing high throughput while complying with "brick wall" stopping criteria, as probably will be required of the system. Merely reducing the headway of individual vehicles creates a more dangerous scenario for the crash stop.

The reduction in drag offered by platooning is also only effective when the vehicles are in very close proximity. Drag may actually increase if the following vehicle is a little distance behind the leader, in the trailing vortex of airflow. No matter how slippery the individual vehicle, a train will produce less drag than the sum of individual components. Indeed, because there is often a space trade-off in producing good aerodynamic shapes, platooning can permit more boxy individual shapes to be used while still providing an overall low-drag shape.

Whether it is practical to implement platooning depends on the system configuration. It may not be possible to merge and dismantle platoons within an urban environment, where intersections may be closely spaced. Certainly it should be borne in mind when routing the network. Ed Anderson’s comment that network layout is an artform in itself should not be taken lightly.

This is primarily because on dissolving a platoon, control is handed off to the human driver. If one considers that the driver was a passenger up to that moment, then demanding the level of human attention necessary at the precise moment of handoff is asking for a crash. It is the Achilles Heel of the AHS effort, and to my mind, has never been properly addressed there.

Having stationary forming and dissolving of platoons at first glance seems to be a giant step backward. But consider:

Being dual-mode, a vehicle is driven from its origin to a staging area. On entering, control is handed off, and the driver is immediately relieved of the driving task. The time taken to dispatch a 20-car platoon is about 45 seconds. This is the maximum wait; the average wait is around half this time.

If one zooms the magnification back a little, this looks no different from any other automated dispatch system: all that changes is that packets replace singletons. This in effect acts as a systems control multiplier - for the same level of system complexity there is a twenty-fold benefit in capacity.

Fully automated rail–mounted systems have the further advantage that local interactive control is also possible between individuals, which may permit make-and-break on the fly. FlexiTrain cannot do this, but has the advantage that its destination may be altered in real time to suit traffic conditions. (At the endpoint, the platoon simply dissolves).