Let's Make Robots!

The TROLLER 1D - New Prototype Test Video

A robot that can crawl or turn into a circle and roll itself around for locomotion.

The TROLLER 1D is my newest robot project.  The concept uses a series of simple 'links' to make up the TROLLER.  Its locomotion can be similar to that of a caterpillar or snake while laying flat, but the novel concept is that it can roll itself into a large circle and still propel itself (like tank treads) by shifting links in the proper sequencing.  

Right Top: Video demonstrating the TROLLER 1D prototype.

 To the right is a video of the initial servo controls test.  I plan to verify and finalize the design before joining the three units together and developing the rolling algorithms.


Latest Updates here:  http://trobotown.com/index_files/troller.htm

The other video (below right) is of the original prototype testing from earlier this spring.

Below is drawing showing the newest CAD layout and the design I'm putting together now.  

Below are the first 4 links, This is the Master Control Link Group.  It consists of a battery back, 4 servos, a Pololu Maestro controller, and the Arduino Mega master controller with an MPU-6050 Initerial Measurement Unit.  I'm also trying out a Wixel Pair for wireless communication and programming of the Arduino.  

 

 

Currently I have assembled acomplete second iteration of the prototype (shown above).

The goal is to refine the crude prototype design into a functional robotic product kit.  The first version will be the TROLLER 1D.  This version of TROLLER will only be capable of moving along 1 dimension.  By limiting the focus of this project to a single dimension, I will be able to perfect the initial two concepts --'Propulsion by Self Rolling' and 'Crawling'.

Propulsion by Self Rolling works by shifting the links in sequences to rotate itself around the perimeter.  To refine this concept the TROLLER 1D will use motion sensing capabilities like the sensors found in a Wii remote (accelerometers  and gyroscopic sensors).  These sensors will allow the control programming to learn how to roll at different speeds, how to change directions, and how to transition into crawling mode.  Crawling propulsion will be achieved by mimicking the movements of a caterpillar.  Proper sequencing of the servos will allow smooth transitioning. 


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

I think mainly those servos didn't have enough power to really move it.  I have a new TROLLER about 2/3rds of the way together now.  The new servos have over twice the torque.  The initial tests are looking good,  I’ll be posting some new videos very soon.

will the plastic parts stl files released so some of us can build something simmilar or the same and develop it on their needs ?

I'm not sure yet.  I've had some inquires about possible commercialization but I still have much work to do on the design to make it work reliably before I could continue that discussion or openly release them.  I’ll figure that out as I get furthur along.  

Wow, this is rather impressive. There are many people who feel that they are mechanically inclined enough to make something like this work who can truly learn a lot from you. Thank you for sharing.

I wonder how you are able to tell those servos to co-orporate all at once!

Is this guy going to be trolling the forums? :P

beauty!!!

Nice concept. Is it already working? Do you have videos to show us the rolling or crawling?

 

I'll try to address several comments here.

I did get it to roll all the way around once, it wasn’t very smooth and hit kinda hard and then stripped out a servo.  It really needs some motors with a little more torque (which is why I just started a Kickstarter project for it).

The design isn’t quite ready to be shared; it is going to be drastically revised.  I learned a lot building and testing it so far.  It needs to be wider so it’s more stable, better wire routing (with places to zip tie), distributed locations for battery packs (to balance the weight), and probably add a WiFi interface to control it.

 

Thank you for the kind words, I'm looking forward to refining the design and figuring out how to program it.

When I read your comment about stripping a servo the first thing that came to mind was these magnetic servo clutches. Not sure if you can get them in the size required to suit the servos you already have, but it might be worth looking into.

Fantastic project, really interested to follow the continued development =)