Remote Operated Gate Control System

[Greg Bernhardt]

This project began life in 2017 when the author had just begun experimentation with GPIO pins on a Raspberry Pi 3. At that point in time a greatly appreciated Christmas present was an “Explorer Hat” module which provided the aspirant programmer with 5 volt tolerant input and output channels (analogue and digital) , capacitive touch pads, built in light emitting diodes (henceforth referred to as LEDs) and a breadboard on which to build proto-type i/o systems. The module is apparently no longer stocked (not by Adafruit anyway) but was certainly a useful “training ground” for programs which have digital inputs and outputs from/to the real world.

https://www.physicsforums.com/insights/remote-operated-gate-control-system/
Author: @neilparker62

 

[Paul Colby]

Cool project. State machines can be very effective. I find the design pattern used here kinda ugly. For certain, it works, but an alternative I've found useful is to encapsulate States as functions. In a recent project I needed to parse bytes from a USB device in realtime. In c++ states were std::function<void(char c)> objects. The state functions are members of a class and there is a class member that stores the current state. Of course, this state variable is set inside state functions. What I find nice about this all decisions are local to a function call. The main loop of the state machine just calls the State member over and over.

In python, things are even easier. Functions are data just like in LISP.

 

[neilparker62]

Cool project. State machines can be very effective. I find the design pattern used here kinda ugly. For certain, it works, but an alternative I've found useful is to encapsulate States as functions. In a recent project I needed to parse bytes from a USB device in realtime. In c++ states were std::function<void(char c)> objects. The state functions are members of a class and there is a class member that stores the current state. Of course, this state variable is set inside state functions. What I find nice about this all decisions are local to a function call. The main loop of the state machine just calls the State member over and over.

In python, things are even easier. Functions are data just like in LISP.

Thanks for the comment(s). Re functions: Yes - the AI engine wanted to go that way but I got a little confused by all the "defs" it came up with and couldn't keep track. As I mention in the article it's so easy for AI to make subtle programming errors which you won't notice unless you know exactly what's going on yourself. So I stuck to the basic polling system which I had "invented" (and therefore understood) prior to AI extracting a finite state machine from it!

 

[Paul Colby]

There should be one (and only one) def per state. onStarting, onOpen, onClosed, onOpening, and, onClosing. Anyway, a working solution is a working solution.