The Turtlebot3 would be teleoperated by various devices.
We tested it using several wireless devices e.g. PS3, XBOX360, ROBOTIS RC100, etc.
This example is operated by ROS on Ubuntu mate 16.04 with Raspberry Pi 3 (except that it tested by LEAP Motion) and OpenCR which controlls Dynamixel XM-430.
I would like to use my upcoming Waffle as a part-time telepresence robot, controlled over the internet. This is specifically so that a housebound friend of mine can go to a science fiction convention. This couldnāt just be a standard remote-controlled robot-with-a-camera and face. Science fiction cons are usually extremely busy and even real people have problems not running into other people, especially in the dealerās rooms.
The Waffle would be modified with a few more avoidance sensors and a tall head. This should make it approximately 3.5-4 feet tall.
I am assuming that the avoidance routines would have to be done on-robot and that my friend would be able to āsteerā the Waffle with a goal and the Waffle would choose how to implement this goal.
Iād use a larger lithium battery so that it could last at least an afternoon or day.
It would go along with me (or I along with it) so that it wouldnāt get ālost.ā
Iām also assuming that I might have to run the Internet off of a cell phone rather than wifi so that it might have lower bandwidth and higher latency than normal.
Has anybody made a telepresence robot specifically for these conditions?
Thank you for your interest in TurtleBot3
Your project is really nice! We expect a lot of people like you to transform TurtleBot3 into the robot they want.
So We have already make some friends of TurtleBot3. I think that your robot is simillar as Carrier
Iāve looked at Carrier and it doesnāt seem to have the sensors that I need. It needs the capability of traveling on uneven floors as well as perhaps some grassy areas.
Maybe a combination of Carrier and Monster. It must be extremely stable because Iām sure it will be bumped a lot in a crowd. Think of any kind of crowded situation where there are small tables and thin maze-like passageways. That is an SF conās dealer rooms.
And to make human recognition more difficult, many of the people will be in costume and there will be other robots there.
Part of the reason for the head is that I can put additional sensors there to locate things that a floor-level sensor would miss. The other reason is so that people will think of this device as a person. Another reason is so that the human on the other end of the link can see through the camera on the head.
I might need to plan on making the head a variable distance high. For portability, it might need to be short, but for telepresence, it might need to be tall. Iāll have to see if the camera Iāve ordered can see in a wide enough field of vision or I might need to make the head more complex than Iād like for now.