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.
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, enriching their journey with knowledge and keeping the travel spirit alive with engaging narratives.
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.
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:
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.
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.
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: