10120625AI Mapper: Generative AI-powered Accessible TfL Journey Planner For Disabled PeopleClosedCollaborative R&DInnovate UKA recent report from the TfL (Transport for London) highlights that people with disability use public transportation less frequently than non-disabled people due to various disruption factors such as overcrowding, cramped conditions and service delays, which are often linked to reduced social participation and limited access to essential services. Here, we aim to fast-track the development and testing of AI Mapper, a novel generative AI-powered accessible journey planner that leverages a wealth of real-time data from the TfL for disabled people, particularly visually impaired people and wheelchair users to complete the journey comfortably and safely. Generative AI has been changing the market landscape rapidly, partly led by advanced large-language models (e.g., GPT-4 from Open AI) which have shown great capabilities in natural language understanding and generation. Our AI Mapper will offer not only novel solutions fully utilising natural language understandings, but also more accessible, disabled-user-friendly interfaces through human-centred approaches scientifically practiced in accessibility contexts. We will co-create AI Mapper's user interface through identifying user needs and developing design insights with disability specialist groups and stakeholders. In parallel, we will develop its backend engine that enables multiple transformer-based language models to efficiently retrieve the public TfL data, particularly disruption and crowdedness information (through TfL API) and performs sophisticated decision making for a more accessible and comfortable journey. This will be followed by our evaluation phase where we thoroughly test AI Mapper's functionalities, performance and accessibility through expert reviews and field studies. Upon successful completion, the AI Mapper will be released and publicly available as a beta service. This 10-month project will be led by the Global Disability Innovation (GDI) - WHO Collaborating Centre for Research on Assistive Technology (AT) which sits in the department of Computer Science at University College London. During the course of this project, we will actively engage with the wider public and stakeholder groups through public events, capacity-building and online presence.