Transport Decision Making

Transport operators in the road, rail, air and maritime sectors are tasked with optimising and maintaining service levels. When it comes to real-time tactical decision making, operators need to know the state of the world now, and in the near future, to choose the most effective action to take. Often this requires data not only of conditions in their own sector, but of events in other sectors (e.g. how a road closure might mean a short-term peak in rail usage), and events outside of the transport network.

Transport users are implicitly and explicitly generating data about their position, transport plans and even their attitudes towards the transport service. They can also be an important source of other additional information about conditions around them (e.g. reporting delays or adverse weather). Already, there are examples of how this data is generated, whether that be generating data on car movements (ref) or the rather more low-tech example of listeners ringing in to offer updates for radio travel reports.

However, there are a number of challenges to integrating such data into transport decision making

• Human – transport control domains already require knowledge of many concurrent variables, and the implementation of decision-making is highly constraint-based, sometimes using legacy technology. Understanding the viability of user-data in such control domains will require new human factors models of control and control strategy.

• Organisational – understanding performance for transport domains is complex. While the traveller has an individual view of their transport experience, the operator may need to make more high level decisions to trade-off reliability in one part of the network for the benefit of all. There is a need therefore to understand how data from the individual (or the crowd) relates to the performance aims of the transport operator.
• Infrastructure – some user generated data streams are already available e.g. location data based on cell id. Other forms may be possible, but yet to be fully realised, such as contextual data abstraction from twitter feeds.

As part of the Horizon transport theme, the transport decision-making project will look to explore these challenges by identifying a number of use cases where user data could support real-time transport operations.

Through this process we aim to deliver

• An understanding of performance and control in transport domains, and the relationship between domains
• Examination of appropriate Human Factors methods (critical incident method, cognitive work analysis) to understand the domain and potential representations of user-sourced data to a decision maker
• A set of use cases for user-sourced data in transport control, with at least one use case taken forward as a demonstrator

This process will involve close collaboration with a number of transport operators, transport stakeholder and transport user representatives from across transport domains, and across the personal, commercial and freight transport sectors.

For more information contact  Robert.Houghton@nottingham.ac.uk