Here’s an oldie but a goodie: [Eiki], [Mark], and [Sheraz] built a pipe crawling robot for their senior engineering project at Florida Atlantic University way back in 2004. Despite being a rather old build, its aged well and still demonstrates the clever ways the guys overcame some engineering obstacles.
The original plan for the pipe crawler was to mount three spring-loaded wheels 120° apart at the fore and aft of each robot section. Six independent wheels for each section of the robot is overly complex, and too much for a single operator to control; the team moved on to a ‘screw drive’ system where each wheel is canted forward a few degrees. This drive system propels the snakebot along by simply spinning, although it does bring in a few challenges all its own.
The robot had separate sections consisting to house a motor, camera, and electronics, so a way to pass wires through a rotating shaft was needed. This came in the form of a few pairs of incredibly small ball bearings around a hollow shaft. After the mechanical portion of the build was finished, the team moved on to the electronic part where an IMU was built out of three small gyroscope sensors mounted perpendicularly to each other.
Sadly, there are no videos of the inside of a sewage pipe from the pipe crawler’s point of view, but YouTube wasn’t launched until a year after this project was finished.
“Six independent wheels for each section of the robot is overly complex, and too much for a single operator to control”
This implies that each wheel needs its own control device, and that somehow the operator has to manage each one manually. Just build a remote that puts the control in a familiar format, and it’s simple. I have no trouble operating 6-8 channels of control at a time on R/C airplanes/copters.
it doesnt look like any of the wheels are powered, either. Just the rotation of each of them.
I many ways I’m glad there’s no video of the inside of a sewage pipe.
It’s an interesting solution to a tricky problem though.