Robotics is fascinating and complex topic. I’ve always wanted to dive into it headfirst, but seeming complexity, cost in both time and money has stopped me so far. Finally I decided to dedicate some time and explore this subject by building my own robot.
Hopefully this will be a series of articles on the build progress, challenges and ideas that I came across during the build.
Without further ado, let’s take a look what is T.H.O.R. Rover project.
Idea and concept.
T.H.O.R. stands for Telepresence Home Observation Robot.
I have pets. When I’m away at work or on vacation I love to check how things are from time to time. There are several robotic telpresence platforms that I’ve tried over the years. Robots like WowWee rovio and several others. Let’s be honest they don’t work very well. These platforms built more like toys with terrible camera quality, slow response and terrible obstacle avoidance. They get stuck on thresholds, or shoe or a cat toy sometimes.
As they say if you want something done, do it yourself. So I decided to build my own robot. I wanted my robot to have following things:
- Have very good HD camera with pan/tilt control. Maybe even several cameras.
- Related to #1, it must have powerful LED headlights to see in the dark. IR LED vision sucks.
- Be able to navigate over various obstacles found in average home and maybe even climb stairs. This means powerful slow RPM motors, one per wheel. We don’t need speed.
- Reliable high bandwidth network coverage with on-board WiFi router.
- High capacity battery to power motors, on-board electronics and lights.
- Smart Docking charging station
- Combination of Raspberry Pi or similar mini-computer and Atmel based micro-controllers to run web server, scripts and control and monitor all components of robotic platform (voltage, temperature, charging, etc.)
- Optional by desirable robotic arm
Pretty ambitious, I know, but this is for best case scenario 🙂
Having almost no experience with robotics (not counting my failed attempt to build a Hexapod) or RCs I’ve spent some time deciding what kind of chassis to use.
I’ve stumbled upon a RC platform “Wild Thumper” made by Chinese company Dagu.
It has 6 wheels each with it’s own motor and suspension system and looked very promising. I went ahead and bought one along with Sabertooth dual 12A motor driver, but unfortunately I couldn’t really get it going. Even when I supplied plenty of power, motors would constantly stall with no load causing motor driver to lock up. So unable to get it going I returned my Wild Thumper and continued on with research.
Then I discovered something called “rocker-bogie” system.
You can read more about it in this great Wiki article, but in summary it’s platform that utilizes 6 wheels and can climb over amazing obstacles. In fact it was developed by NASA and is used by their Mars rovers (i.e. Pathfinder and Curiosity).
It looks astonishingly simple to build and looks like it even climb stairs, as long as it’s wheels are at least half the diameter of the stair’s step height.
Rocker-bogie system has it’s cabin suspended over central rod and to prevent it from just dropping on either side of the axle, differential system is used to keep in level. There are two types, one uses set of differential gears inside the cabin, and the other one uses differential bar mounted on top of the cabin. I decided to go with latter to free up space inside cabin for electronics. I’m also not very good with gears…
At that point I was ready to start shopping for parts and build my first prototype.
Continued in Part II…