Experiments with Raspberry Pi Pico W and Redis

Introduction

This repository contains MicroPython and Node.js code to support my series of livestream videos where I build out a system that sends data from sensors into Redis for processing. I'll add more to this README as the project progresses!

Videos

Episode 1, introducing the Pi Pico, installing MicroPython, creating a Redis Stack instance in the cloud and sending data to it with Redis Streams. Visualizing data with RedisInsight.
Episode 2, adding headers and a Grove shield to the Raspberry Pi Pico W, modifying the MicroPython code to read temperature and humidity values from a sensor and send them to Redis, reading those values from Redis using a Node.js application.
Episode 3, added a light sensor, looked at how to modify the Node stream reading code so that it remembers where it was in the Stream when restarted and demonstrated how to use a Sorted Set to model 1:many relationships. We ended by storing sensor data per room in JSON documents in Redis.
Episode 4, dealt with the issue of trimming the incoming data Stream, added last modified timestamps and sorted out some data type mismatches then added a RediSearch index and demonstrated some queries over the data stored in JSON documents.
Episode 5: the final episode for this project. Added a second Raspberry Pi Pico W with a fan and display attached, used a Redis List to sent it commands to use the fan to lower the room temperature after the temperature has exceeded a threshold for a certain time.

Getting Started

To use this repository you will need:

Raspberry Pi Pico W. This project uses two of these.
Headers for each Raspberry Pi Pico W - you'll need to solder these onto the devices in order to be able to attach the Grove shield.
Your wifi credentials (SSID and password).
MicroPython runtime for Pico W (install this on the Pico W using the drag and drop method).
Node.js.
A cloud Redis Stack instance (free). Alternatively, you'll need to be able to run Redis Stack somewhere on your local network that both your development machine and a Raspberry Pi Pico W can reach it via an IPv4 address or hostname. It's easiest to just do this with Redis Cloud :)
The code in this repository.
You can use the Thonny IDE to edit the code and load it on the Pi Pico W, personally I prefer VSCode with this extension to manage the Pico W. This also allows you to use a single IDE to work with both the MicroPython and Node.js codebases.
If you want to use the exact sensor kit and Grove shield that I am using, you'll need the Grove Starter Kit for Raspberry Pi Pico from Seeed Studio. This project uses two Grove shields, one for each Raspberry Pi Pico W device.

Create the Initial Data Fixtures

This project relies on some data setup tasks that need to be completed before running the Node.js code. Follow the instructions in create_initial_data_fixtures.md before starting the code.

MicroPython

The MicroPython code requires some secrets management... don't commit secrets (passwords etc) to source control, instead create a file secrets.py in the micropython/temp-humidity-light-sensor and micropython/fan-controller folders... it should look like this:

WIFI_SSID = "wifi network name here"
WIFI_PASSWORD = "wifi password here"
REDIS_HOST = "hostname.of.your.redis.server"
REDIS_PORT = 6379 (or whatever port you have Redis running on)
REDIS_PASSWORD = "redis password here"

The Redis client used in this project is in the file picoredis.py. It is PicoRedis by Christopher Arndt and is MIT licensed.

Watch the first video to see how to get the code for the temperature / humidity / light sensor up and running on the device, and how to create a Redis Stack instance in the cloud. In the second video, you'll learn how to add a temperature/humidity sensor to the device and send data from it to a Redis Stream. Setup and code for the fan controller component is covered in the 5th video.

I'm using Seed Studio's examples from their wiki for the code that talks to the various Grove sensors in their kit.

Node.js

The Node.js code also uses the same secrets, but gets their values from a .env file Create a file that looks like this and save it in the node folder as .env:

REDIS_HOST = "hostname.of.your.redis.server"
REDIS_PORT = 6379 (or whatever port you have Redis running on)
REDIS_PASSWORD = "redis password here"

Optionally, if you're using a user name and password to connect to Redis, you can also set:

REDIS_USER = "your.redis.user.name"

To set up and run the Node.js Stream consumer code you'll first need to install the dependencies (this code uses the Node-Redis client and dotenv package) and finally start it:

cd node
npm install
npm run consumer

The code connects to Redis, and reads data from the stream that the MicroPython code is writing to. See the second and third videos for more details. It also instructs the fan component to turn on the fan when a room is above a certain temperature for a while. This is covered in the fifth video and uses a Redis List.

To trim the Stream periodically, deleting entries that are more than an hour old, run the trimmer component like this:

cd node
npm install
npm run trimmer

The code connects to Redis, and uses the XTRIM command to remove stream entries that are over an hour old. Note that to do this is uses the MINID trimming strategy which became available in Redis 6.2. See the fourth video for more details.

simonprickett / raspberry-pi-pico-redis