PlantRoute

Front page with the map
The leaderboard dashboard
Showing the images.
Detail map
The Final Page of the Itinerary
The carbon alternative

Inspiration

We all love travelling. However, during the fun, we often overlook the damage it causes to the environment. According to data from Sustainable Travel International, tourism accounts for approximately 8% of the world’s carbon emissions today. The plane would burn enough fuel to melt down 3 tons of ice. The hotel would use excessive energy for AC and general equipment to please the customers. Despite the situation being hard to change quickly, with the most advanced technologies, we can guide travelers to more climate-conscious choices while delivering the best memories. Therefore, we developed PlantRoute, an AI-powered platform that recommends ideal activities to users, generates a travel plan based on their personal needs, and provides the most climate-friendly travel options and hotels. By explicitly showing the effects of their decisions, we would encourage them to step away from heavy carbon-burning options while still creating unforgettable experiences.

What it does

PlantRoute guides users to quickly design their next travel plan that matches their preferences and choose environmentally friendly options. 🌿

A global 3D map depicting the detail locations with search bar that could easily and visually access any location with visual previews 🌍
Pick activities for the users based on their preferences. 🕺
Display the best hotels and planes at a location based on carbon missions. ✈️
Generated a daily schedule for the users while still able to change them easily. 🗓️
Gamification with a leaderboard that ranks the user with the lowest average $CO_2$ per trip. 🎮
Able to redirect straight to Expedia to implement their travel plans. 🌐
An AI Assistant Chatbot powered by Google Gemini. 🤖

How we built it

Next.js that based on TypeScript
Prisma Database
A trained LLM predicted withuser preferences and past travels based on our database.
Get the hotels and plane data with Amadaeus API and Google Place AI.
Calculate the carbon emission with our algorithm.
Save/retrieve preferences and trip memory with DB and Supermemory integration.
Selected the hotel and plane option with Google Gemini API.
NextAuth API for authentication.
Mapbox API for the 3D map.
Inference service powered by Modal API for modeling and predictions.
Python for Backend.

Challenges we ran into

We have encountered a variety of issues.

How to retrieve data from the Amadeus API and integrate with Google Gemini for recommendations.
Attempting to find a mature database for carbon emissions.
How to correctly connect the Gemini API to our project for the Chatbot.
How to calculate the carbon emission data.

Accomplishments that we're proud of

Connect all the various APIs for functions.
Able to be actually useful to guide users to more environmentally friendly choices.
Deeply incorporate modern AI functions within our project.
A beautiful and easy-to-use UI.
With mostly first and second-time beginner hackers, we create an applicable, satisfying project.
Teamworking with everyone contributing to their specialities.

What we learned

Quickly found various APIs for our needs and learned to use them.
How to start quickly from brainstream to real implementations in a matter of hours.
Designing Strategies that achieve our goal and guide our users.

What's next for PlantRoute

Switch to multi-functional APIs, as we currently use a decent amount.
Convert the point system that the user gains throughout the process to a reward system.