Robot Sheep

I am attempting to build robotic sheep that follow a radio becon mounted on a shepards crook. 

Currently I have 1 failed frame, thinking that my motor speed is too high (RPM) for the steel and Lead acid mounted on top of it. Frames for the 2 larger sheep are still being produced and the powertrain is currently coming togeather. The goal with those are "Follow Me" tech that can carry a significant weight (think Cooler).

 

I am using 2 250W 24V scooter motors controled with an arduino. I have not settled in on a motor controller yet but I was thinking a sabertooth, but I am unsure of the amprage needed. In reality these cannot go faster than 5mph, and would be going more like 2.5-3 (walking pace) anyway. 

 

Eyes for one of the larger sheep

 

My debunked not working baby sheep


This is a companion discussion topic for the original entry at https://community.robotshop.com/robots/show/robot-sheep

You essentially want a type

You essentially want a type of “follow me” system. Indoors, many use infrared, but outdoors those are not very effective, so many designers switch to GPS based (or even vision) systems. A radio frequency beacon is hard to follow since tracking the signal to its source is quite hard. What have you already designed and built?

radio isolation chamber

I am not a radio frequency expert but suppose you built a pie slice shaped metal box that had 3 chambers with openings on the outside ends and a radio in each chamber. Mount this on a rotating axis. Aluminum foil over plastic may be sufficient.

To use, rotate the fixture until the signal strength is strongest in the center chamber. If the max signal strength changes to an adjacent chamber rotate as needed.

Perhaps 5 chambers or 1 chamber with some hunt and peck software would be better.

We rulled out infared

We rulled out infared because the plan is to use this out doors. Currently the set up is ultrasonic and 422 hz recivers. Proof of consept tests have been done, using the US to trigger a search for a radio signal and then finding the radio signal, but I think we are running into exactly as you mentioned. We ruled out GPS initially because of cost, but I think I may go that way (or RC) . 

 

 

GPS

Certainly true that GPS is more expensive than ultrasonic, but the price is dropping. In an outdoor environment with many obstacles, GPS gives absolute positioning. Granted, the accuracy is still not great (a big thank you to the US Department of Defense for putting up those sattelites and giving everyone access). If youre ready to do some programming, you can get a complete GPS module for $16: https://www.robotshop.com/en/gps-bee-mini-embedded-antenna.html

You could use general RF to communicate between the two: https://www.robotshop.com/en/general-rf-data-telemetry.html

RF

If you want to use RF, you can build directional antennas.  (Think RADAR)  With one antenna, you can rotate it with a servo to find the direction of the strongest signal.  With two antennas mounted at 90 degrees to each other, you can use a form of diversity reception and compare the strengths of the two to find the direction (with a little trigonometry.)  

Niether method is really as hard as it sounds.  The higher the frequency and the greater the distance between transmitter and receiver the better it will work.

If you are interested in pursuing that, I can provide you some references.