PlantNurseSuperMegaMax

Utrecht university interaction technology project

Installation

First download the zip or clone the project to your machine.

Setting up the NodeRed server

1. Run npm install in /NodeRedPlantNurse
2. To Run the NodeRed server, run npm run dev
3. Open the NodeRed interface on http://127.0.0.1:1880/admin
4. Set up a your MQTT broker by clicking on a MQTT node and adding a new server config

Setting up the device

1. When the device runs the code for the first time, it will create a WiFi you can connect to. The name and password will be displayed on the OLED screen.
2. Upon connection you have to "log in" to the access point and it will open a web interface in which you can fill in the credentials for your WiFi network
3. After saving the credentials, the device will attempt to connect to the Wifi network you configured and shut down the access point or reboot and show the access point again if that fails.

Setting up the gesture recognition

1. Download face dataset (pictures and corresponding masks) from https://surfdrive.surf.nl/files/index.php/s/7d2647093409cec90119a0bea4fcc779
2. Place images and masks in a folder called 'Face_Dataset' in the same folder as the .exe (if you cloned this repo, this should be in {project-root}/SkinAndGestures/x64/{Debug|Release}).
3. Put images and masks in separate folders named 'Photos' (skin color photo's) and 'Masks' respectively.
4. Open the NodeRed interface on http://127.0.0.1:1880/admin
5. At the top change the path of the wfwatch node to the absolute path to {project-root}/SkinAndGestures/SkinAndGestures/commands on your system.
6. Change the value of the Search for field in the Remove filepath node that is connected to the wfwatch node to the same path as above.

Setting up IoT MQTT Panel on Android

1. Download IoT MQTT Panel on your device
2. Add the credentials for your MQTT broker to the .json file (you might also have to play with connection_id, if you already have other connections configured)
3. Copy the IoT MQTT Panel export.json to your device
4. Import json file in the app

Pin mapping

Pin	Use
D0
D1	Servo
D2	Selector pin for Amux-board
D3	Built in Flash button
D4	Built in LED
D5	I2C SDA
D6	I2C SCL
D7	Reservoir empty led
D8
A0	Multiplexed output for soil moisture sensor and light sensor.