KCA (Key Connection Analysis) method consists of four stages:

Stage 1: Gathering data on journeys made by public and individual transport using quantitative (traffic model, traffic survey, smart card data) and qualitative (interviews with local experts, spacial analysis) sources.

Stage 2: Analysis of public and individual transport offer for the above journeys, with emphasis on identifying missing routes and duplication in the existing rail-based and/or bus network, considering the whole city as a system. This would include, for example, an assessment of whether buses are competing against or feeding to a rail, metro or tram network.

Stage 3: Finding the Key Connections, i.e. high frequency direct routes; the rest of the network can be served by supporting routes of potentially a lower frequency . The aim is a better match of the transport offer to an unevenly spread demand for transport.

Stage 4: Recommended route changes which would increase efficiency of a network, bearing in mind the need to strike a balance between:

• Direct connections from any point to any other point (seemingly ideal from passenger’s perspective but unachievable in practice) and a highly efficient network (transfers are a fundamental feature of such networks) – my suggested solution is a limited number of convenient transfers.

• Changes required in order to increase network’s efficiency and passengers’ habits (any change, even for the better, can create resistance) – that is why I concentrate on cases of low efficiency where a change would give the largest benefit. I tend to leave existing solutions of medium efficiency unchanged in order to avoid revolutionary changes which could discourage existing passengers.

I use my own model which quantifies the relationship between route length, frequency and fleet size. As a result, I can easily perform a network optimisation with constraints for any transport system, e.g. reduce fleet size and increase frequency at the same time.

Example:

Suppose that the Key Connection Analysis (KCA) have shown that there are two routes linking areas A and B and no route links areas A and C. Simply speaking, two solutions increasing the network efficiency are possible:

a) Growth mode: redirecting one of the routes to link areas A and C.

• Operational costs remain unchanged, network’s attractiveness grows twofold: due to reduced duplication and a new direct connection being offered between A and C. This leads to an increase in passenger numbers and ticket revenue.

b) Savings mode: cancelling one of the routes between A and B.

• Operational costs drop due to reduced vehicle usage, network’s attractiveness increases due to reduced duplication. Passenger numbers and ticket revenue remain unchanged.

The choice of either of the above solutions or a combination of both depends on priorities of a public transport authority in each case.