Horizontal Travel Robot Arm Iterations
Thanks for reading up to this point. This blog talks about how I got started with this project, and what I learned across the multiple iterations.
Feel free to leave a comment if you have any questions and would like to learn more.
The Source of Inspiration and the Iteration Process
Learnings from Iteration 1
It took me roughly 2 weeks to get the first iteration right, and it was a straight attempt to replicate the industrial picking robot 1-to-1.
This iteration could only travel in the X and Y axis, and the grip as extremely limited (off the shelf purchase).
I was using 2020 T-Slot extrusion for the axis (just like the robot I saw), but the Nema motors proved to be extremely under powered.
This is the point where I started to realize the challenge at hand. I had underestimated the technical advancement of those robots and quickly realized that building a 2 meter robot in my home was not going to be practical.
I learned and changed my goal. Now, my objective was to build a robot that could do a “light-industrial” motion on a table top. This is what I envisioned:
Learnings from Iteration 2
I started researching, and it seems that a robot arm could be a good solution to have flexible Z and Y movements. Taking a look at my 3D printer, I then realized that I could use rails for the X motion and achieve my objective.
I am new to robotics and kinematics, so my iteration 2 was a miniature that had the objective of helping me figure out the mechanics.
I wanted to achieve my objective with the least amount of motors, and realized that having stabilizer arms for the grip would save me a lot of work, at the same time that it would work for pick and place motions.
Learnings from Iteration 3
Iteration 3 was now a real sized prototype, but it never so action.
It became evident with this iteration, that just using bolts on the 3d printer parts forced me to “loosen” the screw tightness, which introduced “play” in every joint.
The problem I found with this, is that each joint would have a little “play” or “wobble” and it would get amplified a lot at the end of the arm. You can see this effect in the Inspiration and Iteration video.
Learnings from Iteration 4
For iteration 4, we reached “production”. This design was very reliable and started to enable me to work on the software.
It could roughly move 30 cm (11.8 inches) and I realized that I wanted sometime more practical, I tested 50 cm (19.6 inches), but ended up setting my goal on 100 cm (39 inches) so that I could rest comfortable it could have some real world usage.
The challenge with this iteration, is that the lateral stability was really poor. You can see this effect int he Iteration and Inspiration video, and that triggered the need to redesign the plastic parts.
Two major changes happened here:
- The introduction of bearings significantly improved the arm performance
- My kids started to get involved, and they brought with them toys to test.
Iteration 5 — The “Good Enough” Design
I still have ambitions to do an Iteration 6, but this one started to work really well.
I redesigned the 3d parts, and you can see the difference in the picture, there are three main improvements here:
- I made the arms longer, from 100mm to 130mm.
- I added more lateral support for the main arms (the ones that look like ladder).
- And I also build a semi-modular system, that could make changing servos very easily.
This last point is crucial, because Iteration 3 and 4, had the equivalent of “hardcoded” servo mounts, which would make it a challenge if you ever wanted to try this out with servos I didn’t use.
Remember… the reason for this, is that all this will be available for you to replicate, from hardware to software.
So, now we are talking, this thing can now perform what I catalogue as “practical” movements (per my original goals, of having 1 long axis).
