Monitoring Water Consumption With Rasperry Pi Zero2 W

This instructable allows you to read out water meter digits and monitor your home water consumption in real time.

Values are updated every minute with an accuracy of one liter. The consumption data can be viewed on the IoT Analytics - ThingSpeak Internet of Things webpage or through the ThingView app on your mobile phone.

ThingSpeak is an IoT analytics platform service that allows you to aggregate, visualize, and analyze live data streams in the cloud. You can send data to ThingSpeak from your devices, create instant visualization of live data, and send alerts.

The hardware consists of a Raspberry Pi mini computer and its camera. The camera is positioned above the digital water meter, capturing images of the water consumption digits. These images are analyzed, and the data is sent by the Raspberry Pi to the ThingSpeak cloud. In the cloud, the values are further processed and presented in charts. This process is repeated every minute.

This setup is specifically tuned to work with Kamstrup MULTICAL 21 water meter models. See the pic.

HARDWARE

1. 1 x Raspberry Pi Zero2 W Mini Computer (with pre-soldered pin header)
2. 1 x Raspberry Pi Micro USB Power Supply
3. 1 x Raspberry Pi Fisheye Camera Module
4. 1 x Flex Cable for Raspberry Pi Camera
5. 2 x 680 ohm axial resistors
6. 2 x white radial LEDs
7. 4 x 2.54mm pin header cables, ~15cm
8. 4 x 3D printed parts
9. 1 x piece of paper
10. 4 x rubber bands

SOFTWARE

1. Raspberry Pi Zero OS
2. Python script
3. ThingSpeak Matlab code
4. ThingView mobile app

ThingSpeak cloud service is utilized in this study. ThingSpeak is an IoT analytics platform service that allows you to aggregate, visualize, and analyze live data streams in the cloud.

Steps to Create a ThingSpeak Account and Channel:

Create a ThingSpeak Account:

1. Open a web browser and navigate to IoT Analytics - ThingSpeak Internet of Things.
2. Familiarize yourself with the service by reading the introduction links and exploring examples like ThingSpeak-Examples.
3. If you don’t have an account, create one.
4. Non-commercial users can use ThingSpeak for free, though with limitations on certain functionality.
5. Commercial users can access a time-limited free evaluation.
6. To get full MATLAB analysis features, log in with an email associated with your university or organization.
7. In this exercise, the free account is sufficient for non-commercial use
8. MATLAB log in is not required

Create a ThingSpeak Channel:

1. Follow the steps to Create a new channel, using the following values:
2. Name: Water Meter
3. Field 1: Liter
4. Field 2: Cumulative Liter
5. Save the settings.

Find API Keys:

1. Navigate to the API Keys tab.
2. Locate the Write API Key, which should be added to the Python script in the next step: Step 2 Raspberry Pi Zero 2 W configuration.
3. Locate the Read API Key, which is required for the MATLAB code.

Copy and Modify the MATLAB Code:

1. Select the Public View tab and click on the MATLAB Visualization button.
2. 
3. Click Create
4. 
5. Copy and paste the attached MATLAB code.
6. Modify the code by adding your ChannelID and Read API Key.
7. 
8. Save and close the file.

View Data Visualization:

1. Return to the Channel and the Public view. You should see three charts without any data:
2. Current Liter Curve
3. Cumulative Liter Curve
4. Liter/Day Bar Chart
5. Optionally, add the liter/day bar view to the Private View tab.

Pre-configuration Steps for Raspberry Pi Zero 2 W:

Prepare the SD card:

1. Use Raspberry Pi Imager to write Raspberry Pi OS to an SD card.

Set up the hardware:

1. Insert the SD card into the Raspberry Pi.
2. Connect a keyboard, mouse, and display.
3. Plug in the power supply to boot up the Raspberry Pi.

Complete initial setup:

1. Go through the OS welcome steps.
2. Verify that the time and date settings are correct.
3. Ensure WiFi is enabled and connected to your home network.
4. ensure that your Raspberry Pi automatically connects to your WiFi network after booting up
5. Enable the camera (see the pic).

Set up the Python script:

1. Download the provided Python script and save it as /home/pi/watermeter_python_v1.py.
2. The script captures an image every minute, analyzes the digits in the image, and sends the data to an IoT cloud. There, the consumption results are further analyzed and visualized in curves.

Configure autostart for the Python script:

1. Open the terminal and navigate to the autostart directory:

1. Create and edit the myautostart.desktop file:

1. Add the following line to automatically start the script on boot:

1. Save and close the file.

Install required Python libraries:

1. Run the following commands to install the necessary libraries:
2. The libraries are crucial for the script to run successfully.

Modify the watermeter_python_v1.py

1. Use a text editor or IDE to open /home/pi/watermeter_python_v1.py
2. Navigate to the line 15
3. To find your Write API key, see the instructions from the previous step: Step1 ThingSpeak configuration
4. Replace 'YOUR_WRITE_KEY_HERE' with your actual ThingSpeak Write key
5. Once the modification is complete, save the file and close the editor

Completed.

1. Now it can go ahead and unplug the display, keyboard, and mouse. The Raspberry Pi Zero 2 W can run headlessly, and your Python script should start automatically upon boot, thanks to the autostart configuration.

Make the connections:

1. Refer to the image for guidance and ensure all components (camera, flex, resistors, LEDs, pin header cables) are connected and soldered properly.

Print the mechanical parts:

1. 3D-print all four mechanical parts as required.

Assemble the camera and LED:

1. Attach the camera and LED to the 3D-printed bracket plate.
2. Install the assembled plate along with the Raspberry Pi Zero into the housing.

Secure the side brackets:

1. Use screws to fasten the side brackets firmly.

Position the housing:

1. Place the housing securely on top of the water meter.
2. Use rubber bands to hold it in place.

Power on the Raspberry Pi:

1. Turn on the Raspberry Pi to activate the setup.

There are two options available to monitor your water consumption data:

ThingSpeak Web Page:

1. Access and analyze your water usage directly through the ThingSpeak platform in a web browser.

ThingView Mobile App (Android, iOS):

1. Download the ThingView app to conveniently monitor your water consumption on the go.
2. The app is available for both Android and iOS devices.
3. Here are some example images captured from the ThingView Free app on an iPad to give you an idea of how it looks.

Overview

The green bar on the left side of the bottom chart represents the average water consumption per day, calculated across all recorded days.

Cumulative all

By clicking on the top of the middle chart, the overall cumulative curve is displayed in full-screen mode.

Cumulative zoomed

You can zoom into the cumulative results by using the buttons located on the bottom line, which include options such as Hour, Day, Week, Month, and Year. This allows you to view the data over different time intervals effectively.

Liter zoomed

Clicked on the top of the liter curve and selected Day View to display the daily data trend effectively.

The series of operations has been performed on the captured image to enhance and analyze information. Below are the steps described as executed in the Python script:

Image Acquisition: Capturing the raw image using a camera sensor.

Segmentation: isolating specific region of interest within the image.

Grayscale Conversion: In this step, the image transitions from the RGB color format to a single-channel GRAY scale format.

Pre-processing: Adjusting brightness, contrast. Converting black to white and white to black.

Rotation Calibration: This step ensures the image is rotated to achieve horizontal alignment, which is critical for accurate data analysis. It is carried out at the start of the Python script in a for loop.

Image Analysis: This step focuses on analyzing the digits within the captured image. The digits are identified as numerical values.