10103768Understanding Passenger Behaviour: From insight to impactClosedCollaborative R&DInnovate UKThe transport sector faces significant challenges in understanding their passengers. Across rail, bus and metro, providers are required to perform a broader set of tasks with passenger feedback than ever before. * The consumer landscape in public transport is in a state of flux. COVID, industrial action and the possible impact of GBR are regularly changing public opinion. How can providers understand and prioritise the needs of customers, when customer opinion is constantly changing? * There is an acute balancing act between financial imperatives and public service, as we have seen recently with the consultation on the closure of rail ticket offices. Is it possible to quantify a level of inconvenience against a financial saving? * Public transport providers are also charged with influencing passenger behaviour, from getting commuters out of their cars, to taking advantage of savings and opportunities. Can customer behaviour be understood, and influenced, at scale? The flexibility of the Wordnerds platform allows for very exciting new approaches to these challenges. We plan to harness the power of Large Language Models within a framework of methodologies, from the fields of behavioural science, economics, and linguistics. We will train a set of AI categorisation models specifically on UK public transport customer feedback, which will automate key learnings from these different fields of innovation, and create an unprecedented picture of passengers requirements and actions that can be taken. In the course of our work in developing our SaaS model, we have worked with a range of UK Train Operating companies, and witnessed fascinating and innovative approaches to the above challenges. These approaches range from the COM-B model for behaviour change, to pareto reporting to prioritise maximum impact of actions taken, to new ways of reporting back to the ORR. This project will embed these innovations into a robust framework, then train small AI models which will be used to make data-driven decisions, predictions, and suggestions for action. As a result of this grant, Wordnerds will allow transport organisations to understand their customers more fully, react to change more readily, uncover new depths in passenger behaviour, and pinpoint and deliver action. All of these elements, where they are attempted at all, are extremely time-consuming for providers, and so this project will make enormous strides in provider productivity.