1. Department of Electrical and Computer Engineering MDR Presentation Team Frij

2. Department of Electrical and Computer Engineering FRIJ Andrew Paisner - EE Carlton Jones - EE Adviser: Prof. David Irwin Shravan Nayak -CSE Amrit Khalsa – CSE

3. 3 Department of Electrical and Computer Engineering Improving the Refrigerator  Currently a “dumb” appliance, with no interface with the food it holds.  Owners are responsible for knowing what food is or is not in the refrigerator.  Most people rely on handwritten lists for shopping.  Goal: Create a smart refrigerator that knows what products it contains and can use that data to make prediction.

4. 4 Department of Electrical and Computer Engineering The Frij – Goals Reiterated  RFID and weight sensors keep status of food items in the Frij.  Accurate and easy product identification via interface with UPC (barcode) database.  Server/Database stores inventory for easy access and predicts grocery needs.  App displays current inventory and accesses database-generated shopping list.

5. 5 Department of Electrical and Computer Engineering Block Diagram from PDR Refrigerator Barcode scanner RFID sensor Weight sensors Raspberry Pi Food Item (with RFID Tag) Server Product recognition UPC Database Pattern recognition Mobile App Database Block 1: Carlton Jones - EE Block 2: Andrew Paisner - EE Block 3: Shravan Nayak -CSE Block 4: Amrit Khalsa - CSE User Interface

6. 6 Department of Electrical and Computer Engineering Updated Block Diagram Refrigerator Barcode scanner RFID sensor USB scale Raspberry Pi Food Item (with RFID Tag) Server Product recognition UPC Database Pattern recognition Mobile App Database Little change Only the “Weight Sensors” block needs to be replaced User Interface

7. 7 Department of Electrical and Computer Engineering Block 1: Carlton Jones – Weight Sensors  Ordered 50kg rated weight sensors.  Created Wheatstone Bridge and Instrumentation Amplifier.  Tuned amplifier outputs to measurable range.  With Andrew built ADC interface for reading via Raspberry Pi.  For CDR: Scrap this setup in favor of a pre-built scale because of accuracy and tuning issues.

8. 8 Department of Electrical and Computer Engineering Block 1: Carlton Jones – RFID  Ordered UHF RFID reader.  Ordered small adhesive UHF tags.  With Amrit obtained the software that allows it to read the tags.  Tested sensor inside refrigerator and with multiple and various tags.  For CDR: Interface the RFID reader with Raspberry Pi (can’t run manufacturer’s program on Linux).

9. 9 Department of Electrical and Computer Engineering Block 2: Andrew Paisner – Raspberry Pi  Ordered and set up Raspberry Pi.  Created program to read and send inputs from USB barcode reader.  Added functionality to read and send voltages from weight sensors via ADC.  With Amrit designed function to read and send serial data from RFID reader.  Program sends barcode and weight data to server after it is scanned (not RFID as of now).  For CDR: Connect to server via Wi-Fi.

10. 10 Department of Electrical and Computer Engineering Block 2: Andrew Paisner – Barcodes/UPC  Ordered USB barcode reader.  Created program to read UPC tags on Pi.  Researched publically available UPC databases and their API interfaces  Using UPCDatabase.org API created lookup function that runs on server.  For CDR: Add more precise product identification for better UI. And search multiple databases.

11. 11 Department of Electrical and Computer Engineering Block 3: Amrit Khalsa – Server  Server that will store and remove food items Server receives information from Raspberry PI via HTTP POST Stores translated item from barcode lookup into database Sends and receives JSON objects Has functionality to remove and link data to items  PostgreSQL database Two tables: frijdb.items and frijdb.systeminfo Java code to connect to and update information in database

12. 12 Department of Electrical and Computer Engineering Block 3: Amrit Khalsa – Server  Communication Web Server JSON Object in HTTP POST HTTP GET JSON Array  Server responds to different types of data from Raspberry Pi: Barcode numbers RFID numbers Weight sensor data App

13. 13 Department of Electrical and Computer Engineering Block 3: Amrit Khalsa – Database

14. 14 Department of Electrical and Computer Engineering Block 4: Shravan Nayak – App  Communicates with server App sends an HTTP GET request to the url of Heroku server Heroku server sends a JSON array containing the JSON object to the app which contains the product info. App can also send data to the server using HTTP POST Message  JSON Parser App parses the JSON array and retrieves the name and weight of the product The information is stored and displayed on the using the TextView class

15. 15 Department of Electrical and Computer Engineering Block 4: Shravan Nayak – App

16. 16 Department of Electrical and Computer Engineering Block 4: Shravan Nayak – Communication HTTP GET JSON Array SERVER CLIENT

17. 17 Department of Electrical and Computer Engineering MDR Deliverables – Proposed  RFID reader reading tags in refrigerator.  RFID reader and weight sensors communicating with Raspberry Pi.  Barcode scanner with UPC API lookup.  Server that will store and remove food items, weights, and their barcode/RFID data.  Develop the UI for the app and communicate with the server.

18. 18 Department of Electrical and Computer Engineering MDR Deliverables – All Accomplished  RFID reader reads tags inside refrigerator.  RFID reader, weight sensors, and barcode scanner communicate with Raspberry Pi.  Server does UPC API lookup of barcodes.  Server database stores and remove food items, weights, and their barcode/RFID data.  App has UI and receives database updates from server and can send items to database.

19. 19 Department of Electrical and Computer Engineering  Hardware features completed.  Installed in refrigerator unobtrusively.  USB (or other) weighing scale.  Accurate weight sensing (can be compared to UPC).  Integrated power supplies.  Pi sends RFID data and is connected via Wifi.  Accurate product description from UPC lookup.  Server and database with product history and predictive algorithms.  Android UI with complete functionalities.  Incorporating shopping list generated by server. Proposed CDR Deliverables