Powering Edge AI for Sensor Reading with RPI and Breakout Garden (EFM, NiFi, MiNiFi Agents)

September 10, 2019

Powering Edge AI for Sensor Reading with RPI and Breakout Garden (EFM, NiFi, MiNiFi Agents)

Hardware Component List:

Raspberry Pi 3B+
BMP-280 Temperature, Pressure and Altitude
ST7735 0.96 SPI Colour LCD 160x80
MAX-30105 Oximeter and Smoke Sensor
Sony Playstation 3 EYE USB Web Camera

Software Component List:

Raspian
Python 3.5
JDK 8 Java (Soon Upgrading to JDK 11)
Apache NiFi 1.9.2
MiniFi Java Agent 0.6.0
Cloudera Edge Flow Manager
Apache Kafka 2.2

Summary

Potential Use Cases: Tracking Environment in a facility that includes webcam detection, temperature, pressure and smoke.

Our Raspberry Pi 3B+ has a Breakout Garden Hat with 2 sensors and one small display. The display is showing the capture image and is constantly updating.

We currently run via nohup, but when we go into constant use I will switch to a Linux Service to run on startup.

The Python script initializes the connections to all of the sensors and then goes into an infinite loop of reading those values and building a JSON packet that we send via MQTT over port 1883 to a MQTT Mosquitto broker. MiniFi 0.6.0 Java Agent is using ConsumeMQTT on that port to capture these messages and filter them based on alarm values. If outside of the checked parameters we send them via S2S/HTTP(s) to an Apache NiFi server.

We also have a USB WebCam (Sony Playstation 3 EYE) that is capturing images and we read those with MiniFi and send them to NiFi as well. We will incorporate TensorFlow lite models into our analysis.

The first thing we need to do is pretty easy. We need to plug in our Pimoroni Breakout Garden Hat and our 3 plugs.

You have to do the standard installation of Python 3, Java 8, MiniFi and I recommend OpenCV. Make sure you have everything plugged in securely and the correct direction before you power on the Raspberry Pi.

Download MiniFi Java Here: https://nifi.apache.org/minifi/download.html

Install Python PIP

curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py

Install Breakout Garden Library

wget https://github.com/pimoroni/breakout-garden/archive/master.zip

unzip master.zip

cd breakout-garden-master

sudo ./install.sh

NiFi Flow

We Can Query IoT Events As They Stream In

Add ExecuteProcess to Run Our Shell/Python

IoT JSON Data

IoT User

Cloudera Edge Management - Monitoring IoT Events From MiNiFi Agents on Devices

Building an IoT Flow Graphically Is Easy! It follows the Hadoop Philosophy to stitch these things together.

Configure Connection to MQTT Broker

MQTT Configuration 2

Cloudera Edge Flow Manager REST API

Let's Examine those MQTT Messages from Devices

As you see we follow the Hadoop Philosophy of keeping things open, extensible, modular, flexible, transparent, composable, using open data standards, open source, diverse and cloud friendly. In this way we can always bend to the needs of the user and adapt to any environment, any data, any cloud at any time. If we need to windowing we could easily add Storm or Flink. For other streaming use cases we can connect our Kafka topics to Spark Structured Streaming or Kafka Streams for additional processing as needed. We can public our public schemas from our schema registry as open data standards within our enterprise, our ecosystem or world wide. Data is meant for sharing and utilizing to build knowledge. Let's make it happen, from any Edge to any data store to any data cloud to any AI / ML / DS / DL model.

Source:
https://github.com/tspannhw/breakoutgardenhat-spi-minifi

Resources:

Breakout Garden Hat

https://github.com/pimoroni/bmp280-python

https://github.com/pimoroni/breakout-garden
https://shop.pimoroni.com/products/bmp280-breakout-temperature-pressure-altitude-sensor

https://shop.pimoroni.com/products/breakout-garden-hat

https://shop.pimoroni.com/products/breakout-garden-hat-i2c-spi

https://github.com/pimoroni/breakout-garden/tree/master/examples/heartbeat