HealthWay

Home
Inventory
Recipes
Pills

Inspiration

Many college students, including us, prefer preparing our own meals over subscribing to a traditional meal plan. Although meal plans offer convenience, they often come at a high cost and may lack in quality. Balancing a nutritious diet on a tight budget poses a challenge, leading to a cycle of repetitive meals. Furthermore, without families to share with, a significant amount of ingredients goes to waste. Hence, we sought a solution that could guide us in utilizing our ingredients effectively while maintaining a consistent and healthy lifestyle.

What it does

HealthWay is a versatile kitchen tool and lifestyle app designed to help users manage their food inventory in real time, offering timely updates. It starts by asking the user to take a photo of their fridge, from which it generates a list of ingredients currently in stock. As items are added or removed, HealthWay updates the inventory accordingly. Based on this updated inventory, it suggests recipes, making meal planning easier and more efficient. Additionally, HealthWay helps manage health by tracking the user's medications, using prescription labels as a reference, and sending reminders for dosages. This comprehensive approach ensures users can maintain a balanced diet and adhere to their medication schedules effortlessly.

How we built it

HealthWay is built as a Next.js React application, integrating a Python backend for its core functionality, with data storage managed through a SQL database. The frontend design is streamlined using Tailwind CSS, ensuring a responsive and aesthetically pleasing user interface. At its heart, HealthWay employs a Google Cloud visual classification model to accurately identify various types of food. Additionally, it incorporates a motion tracking YOLO (You Only Look Once) model, which plays a crucial role in determining whether food items are being added to or removed from a pantry or refrigerator. The system's inventory is meticulously maintained in a PostgreSQL database, which the backend accesses through a dedicated API. For the feature that suggests recipes based on current inventory, HealthWay leverages the SuperCook API, which crafts meal recommendations from the ingredients listed in the user's inventory, ensuring a seamless and efficient meal planning experience.

Challenges we ran into

Integrating Google Cloud for visual classification, YOLO for motion tracking, and SuperCook for recipe suggestions into one API posed a complex challenge. This integration required seamless data exchange between the models and the inventory database to update in real time as items were added or removed. Each model serves a distinct function: identifying food, tracking inventory changes, and recommending recipes based on current ingredients. The challenge was in crafting backend logic that ensures accurate, efficient, and timely updates and retrievals from the PostgreSQL database, all while maintaining system responsiveness and scalability for an optimal user experience. This demanded meticulous planning in API routing, database management, and performance optimization.

Accomplishments that we're proud of

We are proud of how we overcame the challenge of integrating the three vastly different models—Google Cloud for visual classification, YOLO for motion tracking, and SuperCook for recipe suggestions—into one cohesive product, managing to run all of it seamlessly. This achievement not only demonstrates our team's ability to navigate complex technical landscapes but also highlights our commitment to creating innovative solutions that enhance user experience and streamline kitchen management.

What we learned

Designing a system capable of integrating diverse models taught us the importance of a robust and scalable architecture. We learned to anticipate and plan for the challenges of data flow between different parts of the application.

What's next for HealthWay

The next step for HealthWay involves integrating it with physical sensors to automate inventory updates, moving beyond the need for manual video uploads through a phone. We aim to broaden the app's functionality by incorporating exercise data, offering users personalized weekly workout schedules alongside the current features like pill reminders, enhancing overall health management.