Tinkering; Like Guntering for the Connected Generation.

I added a touch screen in the Kitchen, with access to Home Assistant to monitor the House and Farm.

I like the Wavin pipe mod, simple but effective. Did you consider beam breaks? or just nothing available for Tuya eco system?

The reason why I started using Home Assistant is its ability to mix devices from different IOT eco systems which gives far more choice , also in the case of Tuya with some third part integrations its possible to control the devices locally so they still work if the cloud servers or your internet connection go down.

I've been using some image processing from my Hikvision DVR to send alerts to Home Assistant, but I'm using fairly basic cameras and it only supports smart detection on one camera and basic area detection on the others. I've the smart detection setup for the front gate and it rarely has false alarms, but the area detection on other cameras can be triggered by leaves on a windy day. I'm considering using Frigate with a dedicated processor for proper person and object detection.

Having said that, sometimes you just can't beat simple, I have a few spare PIR's lying around so might give your idea to narrow the field ago.

plan B went up in smoke, due to under rated Resistors 🔥 😧

Plans A and B will be revisited later but here is Plan C. I was avoiding this one as I thought it might not be accurate enough but initial tests are promising.

A non contact EMF sensor, surprisingly easy to make with just a resistor and a short length of wire.

Sensitivity could probably do with some adjustment, but its just a matter of using different resistor values.

Adafruit lora32u4, Has built in LoRa Module and a LiPo battery socket with USB input and battery charging circuit, I'm hoping it will be a simple job to add a small Solar panel like the micro USB ones for Cameras like Reolink use.

#define MY_DEBUG
#define MY_RADIO_RFM95
#define MY_DEBUG_VERBOSE_RFM95
#define MY_RFM95_RST_PIN 4
#define MY_RFM95_CS_PIN 8
#define MY_RFM95_IRQ_PIN 7
#define MY_RFM95_IRQ_NUM digitalPinToInterrupt(7)
#define MY_RFM95_MODEM_CONFIGRUATION RFM95_BW125CR45SF128
#define MY_RFM95_FREQUENCY (RFM95_868MHZ)


#define MY_NODE_ID 84


#include <MySensors.h>
#include <Arduino.h>




uint32_t SLEEP_TIME = 30000; // Sleep time between reports (in milliseconds)


#define CHILD_ID_LIGHT 0
#define LIGHT_SENSOR_ANALOG_PIN 0


MyMessage msg(CHILD_ID_LIGHT, V_LIGHT_LEVEL);

// EMF


int aVal=0;
int VL=0;








void presentation()
{
    // Send the sketch version information to the gateway and Controller
    sendSketchInfo("EMF Sensor", "1.0");


    // Register all sensors to gateway (they will be created as child devices)
    present(CHILD_ID_LIGHT, S_LIGHT_LEVEL);
}


void loop()
{
    aVal=analogRead(LIGHT_SENSOR_ANALOG_PIN);
   // if (aVal > 1 ) {
        VL=map(aVal,1,100,1,255);
        Serial.println(VL);
        send(msg.set(VL));
  //  } 


    


    // Sleep until interrupt comes in on motion sensor. Send update every two minute.
   //sleep(SLEEP_TIME);
}

Initial Arduino Sketch for testing. It is setup a a light sensor, as Home Assistant and MySensors don't have an EMF option.

Possible maybe, but probably not very simple or practical either. By the time you add required safety functionality for the high torque motors and sensors suitable for that sort of environment you'd be as well off building from scratch. There may well be open source projects that would provide a more suitable platform for a machine of that scale.

😁

I've worked on a few Joz robotic scrappers, they are a simple enough machine mechanically.

Building a decent DIY one though probably wouldn't be too cheap by the time you build the chassis, aquire appropriate motors and build the charging station, the electronics wouldn't actually be the hardest part, although software could be tricky but ROS would probably be a good starting place.

https://www.ros.org/blog/getting-started/#

I'd love to spend some time building a robot of similar scale but I think unless I find some sort of cheap donner machine for parts/chassis it's not likely for a while. I was thinking of a field robot for spot spraying weeds using openCV to recognize weeds like ragwort.

I don't think it would be a major issue to fabricate, but think it will be costly enough by the time all the mechanical parts are fitted.

Not sure a drone would really be more practical in practice.

Licensing for Autonomous drones is more complicated for a start.

A slow moving land vehicle for pasture weeds, would have better battery life possibly extended with solar panels, be able to carry more chemical and waste less due to better identification of weeds and better shielding due to drift. Drones probably better suited to tall craps and a slightly wider broadcast of chemical targeting problem areas of a field but probably not individual plants.

Your Automated world will still take a fair bit of administration and maintenance, I like to imagine it would leave more time for family. It could also provide more time for actually looking after the animals too, but I suppose it could go either way on many farms. I like the idea of devices for animal health monitoring as an aid rather than replacement for regularly checking and handling animals. Having recently given up working off farm, I'm beginning to wonder how I ever found time to farm.