This week, we set up the LEDs, and we got the soil sensor working, and as such, all of our sensors are now functioning and sending data to the WebApp. The WebApp has been deployed on the RPi, and users can get Chrome notifications of their plant’s current health status. For plants that are not in our database, users get directed to a page to input their own ideal conditions for their plant. We continued working on integration, ensuring that the LEDs, water pump, and live streaming can be controlled via the WebApp smoothly, with no conflicts between different parts of the code. We set up the greenhouse for collecting image and sensor data for ML training. We waterproofed the greenhouse using a waterproof wood sealant and set up the sensors, LEDs, and water system to their permanent positions. The camera was mounted on a swivel case that Zara laser cut, allowing us to manually adjust the position of the camera.
Progress:
- All sensors working and sending data to the WebApp
- Displaying temperature and humidity sensor data on WebApp with charts
- Working LEDs, water pump, and live stream, all controlled through the WebApp
- Working plant identification API (not integrated with WebApp)
- Working plant health classification (not integrated with WebApp)
- Chrome notifications for plant health
- WebApp deployed on RPi
- Option to manually add plant not in database
- Set up sensor and image data collection for ML training
- Waterproofed greenhouse & physical setup
Next Steps:
- Get heater actuator working
- Get mister actuator working
- Control loops for watering, heating and misting
- Setup automatic vs manual scheduling through WebApp
- Continue collecting data
- Begin testing subsystems