Sustainable travel inspiration powered by AI

Purpose and goal:

The aim of this project was to change people´s travel habits and reduce long-haul travel by testing a recommendation engine that displays points of interests nearby that looks like places far away.

The project can be considered successful as we have built a recommendation engine that can recommend places close to the user that visually look like places far away. In tests, it has been used by a significant proportion of users, especially in Stockholm where we have been able to show more nearby alternatives, which indicates that it has the potential to change travel habits.

Expected results and effects:

The solution has been tested via an app for potential travelers looking for inspiration or planning a trip. The tests show that the recommendations have been used by a large part of the active users, and to a greater extent in Stockholm, which has been the focus of the project.

With this test, we have been able to evaluate the potential and performance of the technology and received indications that we have the opportunity to change travel behaviour. Long-term, the recommendation engine can be scaled to more destinations which would have further positive effects for the climate.

Approach and implementation: The project was conducted from October 2021 to November 2022 in four phases where we

1. Created a database of locations in Stockholm and collected photos from locations all over the world for the purpose of creating matches 2. Created an AI-based recommendation engine that can visually match places in the database 3. Built an app based demo and brought in test users 4. Evaluated and reported results and outcomes

The project took longer than expected, but could be largely executed as planned. The phases we had planned turned out to be both feasible and suitable to implement.