[image reference is broken]
[image reference is broken]

National Crime Mapping Conference 2007

[image reference is broken]

Developing aoristic network analysis upon London’s transport system

Andy Gill, Transport for London

The Greater London Authority aim is to ensure every part of London’s transport users’ journeys is easy, comfortable, safe and reliable; and that all of London’s communities have equal access and opportunity to use the network. Transport for London’s Transport Policing Enforcement Directorate (TPED) is taking the lead and has developed partnerships and strategies facilitating collaboration between all the stakeholders responsible for policing the various transport modes including the Metropolitan Police Service (MPS), the British Transport Police (BTP), TfL and its associated transport providers. A major difficulty of analysing crime upon transport networks is that events may not be static (e.g. an assault on a bus somewhere along the route) thus making it challenging to apply traditional spatial analysis techniques that are normally applied to static events (e.g. an assault at a specific place in the street). The challenge is exacerbated by the complex nature of London’s transport network as the average commuter journey involves three different transport modes. A small scale pilot study is being undertaken on one transport mode with the intention to extend to other modes. The focus is the problem of theft upon the Underground. The main drivers are the imprecise knowledge of where thefts occur, the high volume of theft offences and the BTP targets for crime reduction and improved detections (There were almost 8,000 recorded theft offences on the London Underground during 2005/06 – approximately 150 per week, accounting for 40% of all Underground crime. Targeting high volume crime, such as theft, is critical to achieving the BTP aim of reducing crime by 20%). The analysis should also assist the BTP in officer deployments for disruption and prevention initiatives. The main problem is that a high volume of theft offences do not have a fixed location. For example, a theft victim may not realise they had property stolen until reaching their end of journey destination (which would not necessarily reflect where the offence occurred or where pickpockets operate). The solution being developed is a form of aoristic probability whereby if the start, end and any interchange stations used are known; a weighted value is allocated to sections of the route (a section is hereby defined as the Underground network between two adjacent stations.). With multiple theft offences analysed simultaneously, it is anticipated a ‘picture’ can be created of where theft may be occurring the most, or in other words the identification of potential ‘hot sections’ of the Underground. The appropriate visualisation of these hot sections will also be explored. The inspiration for this form of analysis has derived from the existing temporal aoristic analytical technique developed by Jerry Ratcliffe. Additional contributory factors will be incorporated to improve the quality of analysis including passenger volume, direction of travel, and the date and time of travel (particularly relevant due to the differing transport patterns throughout a typical day or week).


Session 3A Andy Gill [2.14 Mb]

Back to conference programme

Page last modified on 02 dec 10 18:51