FridgePal

website

Inspiration

What it does

fridgepal is always trying to find the newest recipes for you, utilizing force sensors to detect something new in your fridge and swiftly aim it's cameras toward it. Following that, our custom-built image detection model will identify what exactly was put into your fridge, adding it to your virtual fridge catalogue. Whenever you're craving something to eat, you can use our website to quickly create a personalized recipe based on what is in your fridge. You can also specify whether or not you have certain allergies, whether this is breakfast, lunch, or dinner, and around how many calories you're aiming for.

How we built it

Challenges we ran into

The largest challenge we encountered was the issue of addressing hardware issues and integrating a fully software and hardware based project into a product. A large majority of the time was spent debugging motors and troubleshooting working with sensor that interact in a system with other components. From a software perspective, some challenges were faced in the use of ML and AI, where the cloud-vision database did not provide sufficiently accurate results. This forced us into taking a more tedious approach of training our own model to sufficiently identify given data items. Overall, the challenges faced in this project make it even more rewarding to see it become functional.

Accomplishments that we're proud of

Some things the group is proud of is being able to add elements from AI and ML from a python script to interact with hardware components like force sensors and stepper motors. The detection system with a webcam and the methodology of obtaining relevant recipes based on a user's fridge inventory were ideas that were desired when undertaking this project- making the final result a huge success from a software perspective.

What we learned

A key takeaway from this hackathon in particular was to be more prepared in terms of hardware since lots of issues could have been avoided with the use of sufficient components and preparation beforehand. Unfortunately, this served as a learning experience since there were times when the tools at our disposal did not seem sufficient. Overall, the group developed immensely in all domains of hardware and software and this serves as a valuable experience for future hackathons/makethons.