During the start-up phase of the company, our consultancy areas started with the analysis and forecasts of transport demand data. Common behaviour orientated demand models are based on work which goes back to our founding stages, for example as part of the master planning for the greater Nurnberg area or the land-usage and transport strategy for Munich during the nineteen 70s and 80s.
We soon discovered the possibilities, but also the boundaries, of such transport models. These models are able to explain traffic behaviour with conventional factors (e.g. car availability/ownership, costs, available income etc.) and predict changes due to situational changes.
The boundaries of such demand models are reached when it comes to modelling the flow of traffic to a very high level of detail. This is vital for credible evaluations of infrastructure projects or alternative service concepts for regional public transport networks. We have seen, that it is not possible to attain the required level of precision without sourcing further primary data.
Suitable primary data for short-distance public transport comes from surveys carried out by the transport networks for the purpose of income division. For long-distance rail transport, this information is the electronic data retrieved from ticket sales, which for the most part contain information on the origin and destination of the journeys.
A household census is usually only able to deliver behaviour orientated data. Hence it would require too large a sample, to be able to extrapolate to an accurate Origin-Destination-Matrix, the effort of which would outweigh any potential benefits gained.
Our project experience extends to the creation of both regional/local and international/national demand matrices as well as the analysis and quality control of transport data. Our service spectrum even reaches to the forecast of car park requirements for Park&Ride services and demand in the area of event specific traffic.
Due to the fact that the raw data is very different in the different transport areas, it is necessary to vary the algorithms to suit the available data. As such, we tend to use our own software, which we can easily adapt to fit the circumstances of an individual project, rather than relying on software available on the market.