HomeIO backend

Super fast and resource friendly version of HomeIO backend!

HomeIO is system for remote monitor and control of various systems. I have created it about year 2005 and have been using various technologies. This is the last iteration because it is written in C++ for maximum performance.

Performance

I am using it for small wind turbine since 2014 while I moved components written in higher level language (Ruby).

System is deployed on Raspberry Pi with average 15-20% CPU usage. Memory usage depends on measurement buffer.

Frontend is written in Go as HTTP server, and most of operation is performed client side.

How it works?

Please take a moment and look this sample video.

Scenario:

• 0:07 - list of measurements
• 0:15 - current measurements graph
• 0:53 - remote shutdown of wind turbine
• 1:20 - effects of shutdown, gradually reduce charge current
• 1:37 - start - turn of turbine brake
• 2:00 - turbine slowly starts up

Features

• IoServer - bridge between UART/RS to rest of modules via TCP socket
• Measurement fetching - fetch measurements periodically. In wind turbine example it fetches about 10 types of measurements every 100ms!
• Measurement buffer - large optimized buffer. It is much better solution that Redis because every measurement is stored as unsigned int for minimum RAM usage.
• TCP server - allow communication with frontend: current measurements, graph, history, stats, execute actions, ...
• Actions - all possible actions are predefined and can be executed by Overseer classes, by Addon classes or manually via frontend.
• Overseer - simple control rules, example: shutdown wind turbine if charge current is higher than 10A
• File storage - store measurements in CSV file format which you can easily process and store
• Addons - custom addons which can create and store advanced stats, example: hourly generated energy, time utilization. Or you can put there dedicated control algorithms.
• Ncurses UI - most crucial information are accessible from console

Possible use cases:

• Aquarium: monitor water temperature, dissolved oxygen, control light and heater
• Greenhouse: monitor air temperature and humidity, soil moisture
• Houseplant: monitor soil moisture, air temperature to analyze watering interval (second HomeIO instance is used there)

Am I missing something?

I am only author of this system so everything is abvious for me, but you may find that a lot of is missing. Please do not hesitate, post issue or email me.