Food Wastage Reduction Management Android App

eBin: An Automated Food Wastage Tracking System for Dormitory Student’s Mess

The system eBin consists of four sub systems. (i) Food Waste weighing system (ii) People Counting system (iii) LED Display system (iv) Online Web Portal All these components are connected to each other via central entity which acts as data store cum data provider.

Food Wastage Reduction Management Android App

A. ebin System Components

Block of weight machine measures the waste dropped in the waste bin under which the weight machine is kept. The weight is measured every second. The values read by the machine are sent over the network to central data store.

People counting mechanism counts the number of mess cards dropped inside. Here the assumption is that number of mess cards is equal to the number of the people who have line come to take their meals. The data is logged in the data store on every event, event being card drop. The number of dropped cards is relayed to the central data store via being wireless networks. LED display shows the measurements read by the weight machine and People counting blocks. The LED display module pulls the data from the data store every 3-seconds and writes data into the IC buffers which are then displayed on the matrix. Online Web Portal block also pulls the data from the central entity and displays statistics on the web page.

B. Sub-System Design 1) WeightMachine and Card Coutning System

The weight machine, figure 2, is a compact system which consists of raspberry-pi[3], arduino-pro-micro[4], hx711 IC[7], NOVA weighing machine[6], Wi-Fi Dongle. The card counting system consist of raspberry pi, arduino-pro-micro, sharp-Y210A21-IR sensors [5] and wifi dongle. The card counting system is a drop box which has IR sensor fitted into it. There is a slit on top of the box. The mess cards are dropped through this slit and it cuts the IR sensor. IR sensor which was calibrated to read a constant distance reads a new obstacle distance. This obstacle distance is conveyed to the Arduino. Arduino communicates the data to the Raspberry Pi over serial port which works the same way as Raspberry Pi described in the above section.

2) LED display matrix The LED Display Matrix consists of Raspberry Pi, Arduino Pro Micro, AT to PS2 converter, WiFi Dongle. This is one reverse engineered part of the system. The LED matrix earlier was getting static data from the AT compatible keyboard. Arduino was used to emulate this keyboard. A wrapper is written on top of a PS2 library to send data in needed format to the display matrix. Raspberry Pi collects the data from the data store and passes it to the Arduino. Arduino wrapper then converts the data into HEX codes and send it to the LED matrix, which is then displayed on the matrix.

HOW DO ACTUAL DELIVERABLES LOOK LIKE

1) Compact Weighing Scale The weighing scale, Fig 4(a), is covered completely with plastic sheet to protect it from the water. All hardware is fitted inside the box with the strain gauges.

https://codeshoppy.com/shop/product/a-food-wastage-reduction-mobile-application/

Only power cable comes out of the system. The hardware box is attached to the drop box which when powered ON gains life. Again only power cable comes out. Even the power cable can be fit into the box if we use a power bank which is available at low cost.

Why styrofoam? We used styrofoam because it’s easy to use; quick prototypes can be made and that sufficed our purpose. LED matrix needs separate power supply to burn all those 64 x 16 LED`s. Rest of the emulated keyboard ts in a small box and connects to WiFi when powered up.

TESTING STRATEGY AND DATA

To ensure accurate reading from weight-mat and to prevent any false triggering of card counter we had to do lot of manual verification of the data given to us by the two systems. A. Testing weight-mat The weight-mat works by connecting several strain-gauges in a wheat stone bridge formation. The voltage change produced by imbalance in the bridge, when weight is put on the mat varies from one weight-mat to another. To ensure that it gives correct value, its output was compared against the readings of a weight mat which was known to work correctly. The auto-calibration mechanism of the weight-mat must detect bin-removal for bins of different weights. B. Testing card counter After each meal, we manually verified the count and tweaked the system whenever we found the count given by the system wasn’t correct. This required both structural changes and changes in program’s logic. Determining the distance between the walls was critical to the correct working of IR sensor. We also had to make adjustments to ensure that cards fall properly.

Pros of the System 1.To ease testing and facilitate independent operation of different components we modularized the entire system. This also ensures that in the future we can upgrade different modules of the system without hindering the operation of other parts. 2.The system as a whole requires little user intervention for operation. This ensures its acceptance by mess workers. They need not undergo any form of training to use the system. 3.Keeping in mind the elements the weight mat will be exposed to, we made it entirely self contained, robust and capable of being operated in the watery environment of the kitchen. 4.Different modules of the system talk to each other over RF, this precludes setting up of complex wiring, prevents any discomfort to the people in the mess and increases system’s adaptability to different hostels

Cons:Places where the system couldn’t fare well Network connectivity: The signal strength of wifi availablein the kitchen varies considerably from one hostel to another, because of which the system sometimes fails to connect to the WiFi. A high gain WiFi antenna might improve the reliability but such an antennae is quite long, hence it will affect the compact design of our system. Manual sorting of cards: The mess authorities keep the cards sorted by their numbers to enable the students to find their cards easily. As card counter subsystem has only one slit and a single compartment because of which all the cards get stacked on top of one another, this required us to sort the cards manually to ensure least unease of the people. To overcome this limitation, a box with multiple slits might help. Learn More