Stocksky

Inspiration

One day, Yuval was hungry. When he came back home, he dreamed about the amazing sandwich he was about to eat, but this is where the plot thickens....

Yuval went home, opened his fridge and was shocked to find only a rotten cucumber and some cheese.

If only Yuval had a way to know he had nothing to eat, he could buy all the required groceries on his way home.

Another story happened to Roman (me). My girlfriend knew we needed to buy groceries, and she asked me to go and buy them, When I asked, what do we need to buy? It ended up with buying a lot of unnecessary groceries, and forgetting the important one's.

If only I had a way to know what actually goes on in my fridge, it could save me money.

How it works

You put a sensor on the side of fridge's door (from the inside), this sensor is being used to scan and identify the products you put inside / take outside of the fridge.

The final solution is using RFID and some quick way of identifying the products, until then, we'll use the old school barcode scanners.

When the fridge device scans the product, it submits the action which happened (put inside / outside) to the cloud.

The user has a mobile up to view what's going on in realtime in his fridge. The user can also set up his shopping list, which contains the products and amounts he is used to buy, and when the cloud recognises the amount is lower than the required amount, it will notify the user about buying the product.

After the notification the app will recommend the user about the best places to buy the product at (location & price wise).

This way, the user never misses anything in the fridge and always knows the fastest and cheapest place to buy his products.

Challenges I ran into

The HARDWARE - we had no devices to use for the scanning process (had no camera for the Arduino), so we used a mini iPad to function as a barcode scanner and notify the cloud about the data.
The Product list - even though all markets in Israel were required to submit their product lists and prices, they did it in a very nasty and uncomfortable way, we created some code to auto-fetch the records, but its not that precise. Hopefully they will publish it in a better way later.
The product-barcode connection - Even though we had the products, we couldn't connect the product to the barcode, after going over a lot (A LOT!!) of bad data posted by super markets, we found one which also gave out the original barcodes (with some changes from time to time and some different standards), and then we uploaded and stripped it to the cloud.
The cloud - due to a lack of time, we decided to use Parse - which didn't gave us the exact solution we wanted, most of the logic remained in the mobile client side, and its very hard to scale / transfer to other platforms. We will replace it of course later on with a dedicated server.

Accomplishments that I'm proud of

We scan a barcode - we get the actual product.
The app gets the actual products in the fridge.
The above was accomplishes in less than 24 hours.

What I learned

We learned how to use Parse and integrate it (including weird configurations).
We learned to plan before working (it helped us with the timeline).
We learned that the big firms, don't actually care about any rules or regulations they are required to perform, and that when we read "all must give out their products and prices" it doesn't actually mean you'll get a RESTful API / nice XLS file with the data, it just means the data might be somewhere on the web in impossible-to-read format.
11 year old kids are very creative and can be very surprising.
You can write tons of code in 36 hours.

What's next for Stocksky

The next stages are using a scale on each shelf to actually know the usage of the products and to give the user usage graphs - give him advices about purchasing other related products, letting the user know he better buy more milk / cereals because he uses one more than the other.

More phases - connecting the products in your fridge with recipes, and offering the user things he can make himself from what he has - to prevent hunger.

Realtime family shopping list integration - notifying the person closest to the supermarket to keep the fridge full.

Built With

Submitted to

Hacking Generation Y Israel

Created by

I designed the architecture, and worked with Alon on the DB architecture, and planned the timeline and milestones. I developed both mobile apps (sensor prototype and client app), "designed" the general UI with help and tips from the other team mates and helped with creating the product database of the big firms.

Roman Slutsky
I worked on the backend side, it was my first time using Parse.com service so although i managed to create the data and understand how to get it with rest it was challenging to deploy code of your own for some logic of our app. I am satisfied with the way i dealt with the docs and finished the job

Alon Levy
I worked on the backend side. I made a python script which we used to parse and convert the horrible xml files of the big shopping firms into usable Json for our Parse.com infrastructure. I wrote some of the cloud code on the server.
At one point, we thought the internet will never return to us, so I started to create the server side using Django. But then it came back, and we continued working with Parse.com

yuval 10
Avihay Shem Tov