Toyota EdVenture

login page

Inspiration

Our inspiration stemmed from a lengthy road trip from Illinois to Texas, which started as a boring drive and evolved into a captivating educational quest. Initially, the drive was nothing more than a tedious necessity. However, as we began to explore the stories of the areas we were passing through, the trip transformed. The shared tales became the highlight of our adventure, proving that the road could be a classroom and a storyteller. From this realization, the Toyota EdVenture—our commitment to converting routine drives into interactive and informative adventures that entertain, educate, and optimize the driving experience.

What it does

The Toyota EdVenture App leverages the power of AI to turn car rides into thrilling educational experiences. Through a mode that activates the car’s camera, the app identifies nearby landmarks and generates content, including historical facts and stories about the surrounding environment. This information is conveyed to the passengers through audio, enriching their journey with knowledge and keeping the travel spirit alive with engaging narratives. The key is to have a robust content management system that can handle the dynamic generation and delivery of content without overwhelming the user or compromising safety

How we built it

We embarked on this project by equipping ourselves with new tools and technologies. With a dedication to learning, we harnessed Flask, a Python web framework, to build a foundational web application. To populate our app with rich, generative content, we integrated OpenAI's API. The journey continued with the implementation of Google Cloud APIs for geolocation services, geocoding, service usage, text-to-speech, place information, and the training of a custom image recognition model. Our database needs were addressed through MySQL in Google Cloud. Throughout the development process, GitHub served as our nerve center, allowing us to employ GitHub Pages for deployment, Actions for CI/CD, and Projects for managing our Agile development workflow.

Challenges we ran into

The venture into AI was a leap into the unknown, as none of the team members had previous experience in these fields. Every software tool was a blank canvas, waiting for us to paint our skills upon it. Some of the challenges we faced included:

Navigating the complexities of training an image recognition model with limited data.
Ensuring seamless integration between the various APIs and our Flask application.
Balancing the load on the system to prevent the app from becoming a distraction to the driver.
Addressing privacy concerns related to image and data capture.
Developing an intuitive UI/UX that could be easily navigated by users of all ages, especially considering one of our team members was new to coding

Accomplishments that we're proud of

We take immense pride in creating an MVHP (minimal viable hacked product) from the ground up, an embodiment of our combined strengths and collective learning. The team's ability to coordinate and leverage individual skills, particularly in the face of one team member's non-technical background, stands out as a testament to our collaborative spirit and dedication to innovation.

What we learned

Throughout this process, we gained invaluable insights into the practical applications of AI. We learned the intricacies of Flask as a web development tool, the capabilities of OpenAI's API for content generation, and the robustness of Google Cloud's services for real-time data. Our understanding of creating a seamless user experience deepened, and we became adept at Agile methodologies through real-world practice.

What's next for Toyota Edventure

The road ahead for Toyota EdVenture is brimming with potential. While we have established a mode that delivers historical content based on image recognition, we aim to enhance the app with:

A diverse array of interactive storytelling features that utilize real-time environmental data and passenger input.
Expanding our Q&A functionality to be more interactive and comprehensive in its knowledge delivery.
Implementing parental controls and customization options to tailor the educational experience to each family’s needs.
Integrating with more vehicle systems to offer a seamless in-car experience.
Introducing gamified learning elements to further engage users.
Preparing the app for offline use by enabling the download of content for specified regions.
Refining our safety-first design to ensure the utmost security and non-distractive operation.
Broadening the content database to include a wider array of landmarks and points of interest.