Exploring Machine Learning-based Support for Commuters in Travel Planner Apps

An exploration on how to combine “human-in-the-loop” based machine learning with user experience design

This project explores how to combine human­-in­-the­-loop based machine learning with user experience design, using commuters and similar frequent travelers of the public transport system and their travel planning needs as research setting for this exploration. The results will be placed in relation to the impact the design explorations have on perceived understanding and control of the machine learning-based agent.

Specifically, the intent is to develop an understanding for different levels and forms of manual control (e.g. ability to directly coach and configure the machine learning agent), the predictive and proactive suggestions suitable for the machine learning agent (e.g. which time frame to predict travel for, likely routes of relevance, and which contextual data is relevant to consider), as well as the implicit levels of required trust and user interaction this brings from a user experience perspective.

Application area: Smart Transport

Partners: Malmö University, Skånetrafiken

Project period: 2018–2019

Funding: The project is funded by Vinnova, Formas and Trafikverket through K2, Sweden’s national center for research and education on public transport.

Contact: The project is coordinated by Carl Magnus Olsson, IOTAP, Malmö University.

Comments are closed.