JournAI: Revolutionizing Urban Mobility with AI

The CMU Corporate Startup Lab capstone project focused on leveraging large language models (LLMs) and machine learning (ML) to enhance urban mobility analysis. By utilizing the Four-Square NYC mobility dataset, the team applied LLMs such as GPT-3.5 and GPT-4 to gain deeper insights into commuter behavior and lifestyle patterns. This innovative approach enabled the identification of key commuter clusters, providing detailed personas and addressing their specific pain points. The project culminated in the development of JournAI, a tool designed to generate, validate, and pitch mobility services tailored to user behavior and travel patterns.

The problem statement guiding this project asked how LLMs and ML could be used to analyze urban mobility data to improve the understanding and prediction of commuter behaviors and needs. This challenge drove the team to explore the potential of advanced AI technologies in augmenting traditional data analysis methods. The strategic use of AI in this context set a new standard for future mobility services, showcasing the transformative power of these technologies.

Throughout the project, the team, consisting of four students, engaged in extensive research and development over a 16-week period. They utilized a range of technologies, including GPT-3.5, GPT-4, LLAMA, Vixtral, and Streamlit, to achieve their objectives. The project was marked by a series of engagement highlights, including the creation of a comprehensive journey map outlining phased tasks, emotional responses, and improvement opportunities.

The team's journey towards developing a Minimum Viable Product (MVP) involved hundreds of prompt iterations and extensive user testing, leveraging multiple LLMs and numerous team meetings. This iterative process was critical in refining the JournAI tool, ensuring it met the needs of users and stakeholders effectively. The project also included a detailed interface design for the JournAI Streamlit application, which facilitated the exploration of movement trends and mobility data for improved planning.

In conclusion, the CMU Corporate Startup Lab capstone project demonstrated the significant potential of AI in transforming urban mobility analysis. By harnessing the power of LLMs and ML, the team successfully developed a tool that offers enhanced insights into commuter behaviors, setting a benchmark for future mobility services. The project's comprehensive approach and innovative use of technology highlight the strategic value of AI in addressing complex urban mobility challenges.