Like it or not, grocery shopping is a major unavoidable chore. Especially in these circumstances, shopping can be a pain and we decided to automate the tediousness of checking expiry dates, whats in the fridge/pantry and also whether we should wear a mask indoors (sometimes we should, especially if there is company) the workshop on Saturday inspired us to build a cheaper and better quality air sensor which can even detect cough and sneeze droplets indoors

What it does

there are two main sets of features: 1: shopping list generation and expiry date detection. All you have to do is take pictures of what’s in your fridge and kitchen, FridgeRemindr takes care of the rest. The app detects the products present in the images using machine learning and generates a list. It lets you add essential items to a list along with their expiry dates. The app keeps track of the dates and generates your grocery shopping list for you. FridgeRemindr also generates a shopping list for you based on your cravings. All you have to do is name the dish you want to prepare. You could exploit this feature to help you prepare for a potluck, Thanksgiving dinner or any special occasion. FridgeRemindr compares the list of ingredients required with the items you have at home and generates the list for you. 2: active air quality monitor the price and functionality of the purple air sensor was shockingly high, so we "borrowed" a sensirion sps30 sensor which is probably the most accurate retail available air quality sensor which is less than 45$ retail. paired with a DHT11 (less than 3$ retail) and a cheap microcontroller with wifi (esp32, less than 7$ retail) it is possible to build our sensor for a hardware price of under 55$. This setup senses particles from pm10 levels all the way down to pm0.5 levels. This is important because respiratory droplets, airborne pathogens and micro pollutants can be detected at this level of granularity (we actually tested if a cough or sneeze can be detected, and our sensor can do this: video here - ) our setup uses a dht11 for temperature and humidity but we can also use a better sensor like a bme280 . basic video hardware video 2:

How we built it

hard work and perseverance and no sleep

Machine learning: used models from GCP cloud vision and clarifai OCR - used google cloud and tesseract Servers: used python flask and GCP functions React-Native to build a cross platform app Adobe Illustrator for designing the logo, assets, UI/UX Hardware : sensirion sp30 particle sensor, dht11 temperature and humidity sensor, ARduino, raspberry-pi (didnt make the UI, ran out of time :( ) Ngrok tunnels everywhere nodejs/express : push notifications server

Scientific references


Challenges we ran into

coding for the sps30 was challenging (had to adapt some things directly from datasheet).

Accomplishments that we're proud of

the things all work.

What we learned

its hard work integrating everything .

What's next for FridgeRemindr

.better integration of components more robust hardware and casing better notification system Placing orders automatically for essential groceries

Share this project: