A Minnerva project for Transport for London developed the specification for an advanced and detailed methodology, using Bayesian and Maximum Likelihood statistical methods, to enhance the quality of transport survey data

When large surveys are created from a number of component surveys, in this case the LATS 2001 survey, it is inevitable that there will be inconsistencies between the component parts. Often these inconsistencies are ignored but they can sometimes be embarrassingly large, even with good surveying techniques, on account of the variations to which transport systems are so often subject

The methodology also allows new data to be introduced to maintain the currency of the survey data. This provides an important means of protecting the large investment associated with major surveys

This work on metadata was part of a larger programme of work for TfL, to which Minnerva contributed and provided specialist inputs, to enhance the manner in which transport information should be handled within TfL and its related organisations, including the Greater London Authority and the London Boroughs.

(Contact: Miles Logie)