Introduction
As cities transition toward sustainable transport, multimodal journeys—those combining different modes such as public transit, shared mobility, cycling, and walking—are becoming more common. In this evolving landscape, AI-powered travel assistants are expected to play a crucial role, not only guiding travelers but also making decisions on their behalf.
How can the interaction between travelers and AI travel assistants be designed to support shared decision-making and enrich multimodal journey planning?
Current journey-planning apps, such as Google Maps, already use algorithms to provide route recommendations based on factors like travel time, congestion, and public transport schedules. However, these systems primarily focus on efficiency and simplification, often neglecting personalized travel behaviors, user preferences, and contextual factors that influence human decision-making.
This thesis explores how multimodal travelers plan and make journey decisions and proposes design principles for a personal travel assistant, Travel Buddy. Unlike existing systems, Travel Buddy does not merely suggest optimal routes—it adapts to user behavior over time, integrates with personal data (such as calendar events and travel history), and allows for customization based on user-defined priorities.
By fostering a collaborative decision-making process, Travel Buddy balances automation and user control, rather than simply optimizing for speed and convenience. The final prototype envisions a future scenario where travelers and their smart assistants co-plan journeys together. This research concludes with a design manifesto, outlining key principles for creating future intelligent mobility assistants that support personalized travel while preserving user autonomy.
*It might take some time to load images and contents. Thank you for your patience.