What do you get if you combine Python and and MCUs?
Project description
reticul8
Remotely articulated MCU endpoints for Python
reticul8 allows you to use Python to remotely control a compatible microcontroller such as an Arduino or ESP32.
On the Python side, it uses Python 3.5+ and asyncio
On the microcontroller side it uses PJON and PJON-cython to communicate with the micro controller - anything uC that can run PJON should be compatible.
It also uses protocol buffers to encapsulate the RPC messages.
For example, you could use the following setup to wirelessly control an ESP32 using ESPNOW
"HUB" (Running Python) <--Serial/UART--> NODE 1 (ESP32)
^--ESPNOW --> NODE 2 (ESP32)
^--ESPNOW --> NODE n (ESP32)
Rationale
reticul8 is designed to meet the following requirements :-
- The system should be able to run "complex application logic" and be "internet connected"
- Nodes in the system should be able to connect to the hub using a variety of mediums (wired and wireless)
- Nodes should be able to run on common MCU hardware (Arduino and ESP32 targeted initially)
- Nodes should be fast and reliable, but don't need to be "smart" - application logic can live elsewhere
- Communication between nodes and controller should be fast and reliable (ie not over the internet!)
Notice that one key requirement is the absence of internet connectivity. What happens to your home automation system when the internet goes down?
reticul8 is designed for a home automation system where the nodes are not (necessarily) directly connected to the internet. This also has the benefit of making communication between the controller/hub and the nodes much faster than something like pub/sub (<70ms rtt for a two node setup with ESPNOW and ThroughSerial).
Building on PJON as the communication medium between the nodes allows for plenty of options.
reticul8 is designed to be part of a home automation system - specifically it allows nodes (eg an ESP32 or Arduino) to operate as dumb remote endpoints controlled by a smart controller (eg Python running on RaspberryPi).
Competing projects include :-
- Mongoose OS - An open source Operating System for the Internet of Things
- MicroPython - Python for microcontrollers
- Zerynth - The Middleware for IoT using Python on Microcontrollers
But when I looked at the features I required, none of these seemed like a good fit. MicroPython and Zerynth seemed to be too "resource heavy" to run a simple dumb endpoint. Mongoose OS was a pretty close fit but it still assumes your nodes are on the internet.
Arduino-like API:
The nodes (endpoints) are controlled using Remote Procedure Calls (RPC) defined with protocolbuf.
An Arduino-like API is provided :-
import asyncio
import uvloop
from reticul8 import rpc, pjon_strategies
from reticul8.arduino import *
class Node(rpc.Node):
async def notify_startup(self):
print("Received startup message from {}".format(self.device_id))
with self:
# schedule the inbuilt LED to blink 10 times
with Schedule(count=10, after_ms=100, every_ms=500):
await digitalWrite(22, LOW)
with Schedule(count=10, after_ms=600, every_ms=500):
await digitalWrite(22, HIGH)
await asyncio.sleep(10)
#manually blink the LED
await pinMode(22, OUTPUT)
for i in range(5):
await digitalWrite(22, HIGH)
await sleep(.1)
await digitalWrite(22, LOW)
await sleep(.1)
#read the value of the pin
await pinMode(19, INPUT_PULLUP)
value = await digitalRead(19)
print("HIGH" if value == HIGH else "LOW")
#ping the remote node
for i in range(10):
await ping()
#an ESP32 feature - built in PWM
await PWM_config(22)
while True:
await PWM_fade(pin=22, duty=0, fade_ms=500)
await sleep(1)
await PWM_fade(pin=22, duty=8192, fade_ms=500)
await sleep(1)
class PJON(pjon_strategies.SerialAsyncio):
def notify_connection_made(self):
print("ESP32 connected")
def notify_connection_lost(self):
asyncio.get_event_loop().stop()
asyncio.set_event_loop_policy(uvloop.EventLoopPolicy())
loop = asyncio.get_event_loop()
transport = PJON(device_id=10, url="/dev/ttyUSB0", baudrate=115200)
Node(remote_device_id=11, transport=PJON)
loop.run_forever()
loop.close()
Supported RPCs
GPIO
- pinMode()
- digitalRead()
- digitalWrite()
- INPUT -> Watch pin for changes with callback on change, debounce
I2C
- i2c_read
- i2c_write
ESP32 specific features:
- PWM (ledc)
- OTA Update
reticul8 helpers
- schedule commands to run repeatedly
- run multiple commands
Planned features
- Analog output
- Analog input
- Touch sensor (ESP32)
- Pulse counter (ESP32)
Building an ESP-IDF component node
Create a new ESP-IDF project, and add the Arduino component.
Add reticul8 as a component :-
cd components
git clone https://github.com/xlfe/reticul8
Your main.cpp
just needs to setup your PJON buses, and pass these to the reticul8 class. Call setup and loop as per
the arduino functions.
// Define Wifi config for ESPNOW
#include "esp_wifi_types.h"
static wifi_country_t wifi_country = {
cc: "AU",
schan: 1,
nchan: 14,
max_tx_power: 80, // Level 10
policy: WIFI_COUNTRY_POLICY_MANUAL
};
#include "Arduino.h"
# PJON defines
#define PJON_INCLUDE_ANY
#define PJON_INCLUDE_TSA
#define PJON_INCLUDE_EN
#define TSA_RESPONSE_TIME_OUT 100000
#include <reticul8.h>
PJON<Any> *bus = NULL;
RETICUL8 *r8 = NULL;
void loop() {
r8->loop();
}
void setup() {
Serial.begin(115200);
Serial.flush();
//EPSNOW
StrategyLink <ESPNOW> *link_esp = new StrategyLink<ESPNOW>;
PJON<Any> *bus_esp = new PJON<Any>();
bus_esp->set_asynchronous_acknowledge(false);
bus_esp->set_synchronous_acknowledge(true);
bus_esp->set_packet_id(true);
bus_esp->set_crc_32(true);
bus_esp->strategy.set_link(link_esp);
//Uncomment the line below to make a single bus device (eg leaf)
// otherwise the device is initialised as a bridge between esp-now and serial
// r8 = new RETICUL8(bus_esp, 10); /*
//Serial
StrategyLink <ThroughSerialAsync> *link_tsa = new StrategyLink<ThroughSerialAsync>;
link_tsa->strategy.set_serial(&Serial);
bus = new PJON<Any>(11);
bus->strategy.set_link(link_tsa);
bus->set_asynchronous_acknowledge(false);
bus->set_synchronous_acknowledge(false);
bus->set_packet_id(false);
bus->set_crc_32(false);
PJON<Any> *secondary[1] = {bus_esp};
r8 = new RETICUL8(bus, 10, secondary, 1);
//*/
r8->begin();
}
Finally, make sure your component.mk
(in same directory as main.cpp) includes the following :-
COMPONENT_DEPENDS += reticul8
#Used for build timestamp
CPPFLAGS += -D"__COMPILE_TIME__ =`date '+%s'`"
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
File details
Details for the file reticul8-0.1.tar.gz
.
File metadata
- Download URL: reticul8-0.1.tar.gz
- Upload date:
- Size: 19.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/1.12.1 pkginfo/1.4.2 requests/2.20.0 setuptools/39.1.0 requests-toolbelt/0.8.0 tqdm/4.27.0 CPython/3.7.0
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 3fc7aae8c662c3053d7592ca1375dc594f80f919fc328a4e4b7f5152c41b0968 |
|
MD5 | bfea16933b517e0661dd7c76f74ddc36 |
|
BLAKE2b-256 | fe72ef604987e94e3a0737b1138bed194baffef763adf92d4911102aed055242 |