Wk9&10: 2016/03/30-04/06

The assignments for wk9&10 were as below;

1. Make a machine, including the end effector, build the passive parts and operate it manually.
2. Automate your machine.
3. Document the group project and your individual contribution.

These two weeks, we, Fab Academy students in Fablab Amsterdam, decided to make 'Random Answer Machine' with a programmable arm. If people ask a question and the arm provides a random answer through a push motion. This project consistted of the three main parts.

• The mechanism which will move the arm after programming
• The programmable arm
• The card holder

'Random Answer Machine'

Here is our group page;

• Group project documentation

And these are the source files of the project:

• Structure design files - Solidworks
• Answer sheet sticker files - Illustrator
• Placeholder fiber rods - Solidworks
• Top plate - Rhinoceros 3D
• Hand - Rhinoceros 3D
• Machine Files for Download (Code for manual control)
• Machine Files for Download (Code for answer selector)

Actually, thsese two weeks was a bit challenging for me because I have to work with Fablab Amsterdam members, almost all of who are the professional designers or artists in the field of product design or so. I managed to add value to this team in any way.

Initial Test

I had to be absent for the class at first because the break time of Fab Academy had moved forward 1 week, then I had already reserved my easter trip. After coming back to Amsterdam, other menmbers shared the progress and told that we decided to make 'random answering machine', which I though was quite intersting and fabulous.

First, we had to decide the whole structure of our project. Regarding the main body, what was the issue was what structure was the best. After combining the idea and functions, we got the idea of using glassfibre rods which will be moved up and down by the supplied lead screw motors. The rods will form an arm which maybe can push up or out some form of answer cards.

Concept sketch of the machine structure

We had two options on the structure of the arm;

• 3-axis: more flexible but less stable, and controralbe less precisely
• 4-axis: less flexible but more stable, and controralbe more precisely

Each options seemed each advantages/disadvantages, however, it was just theoretical idea, so we decided to make both first and examine their motion. Let's trial and error.

3-axis arm test

cutting of 4-axis parts

4-axis test

From these test, we found that 4-axis rods model was too stable to move smoothly, actually it was a bit slow. On the other hand, 3-axis model was not so unstable and more precisely, contraly our expectations. In this test process, I mainly took the responsibility of the designning, cutting and assembling the 4-axis arm.

From this test, we were able to choose the 3-axis model.

Cutting&Assembling the frame and arch

In these two weeks, some of our members were absent at Fablab Amsterdam because some were a bit busy and others livded far from Amsterdam. Paul and Sebastian, who were the professional indastrial designers make the cutting desing of the frame and arch of the project. In this construction parts, I mainly cut parts and assemble them because I was faminilar with cutting various parts quickly by laser cutter as a internship in Fablab Amsterdam, where I taugh visitors who had various backgrounds how to cut the parts correctly. Actually, I cut almost all parts of the project.

We used 6mm poplar plywood, which Sander bought. The wood sheet was a bit wide, so I examine the several setting of the power/speed of cutting and found the appropriate setting for cutting them.

Assembling the frame

Cutting parts

Assembling the arch

Test Arm - 3D print

After fixing the structure of the frame and arch, we had to decide the detail part of the top hand. We discussed and came up with two idea on it:

• Using 3D printed hand
• Flat polywood hand

Still we had a time to test, I decided to make both of it. First, I tried the way using 3D printed hand. I downloaded free 3D hand data from internet, and added the hole to attach to our existing arm part as below.

Adding hole to 3D hand

After preparing the file, I started to print. However, it took long time because of its big size, 20cm heigh or so....

Print setting on Cura

3D printing the hand

After finishing printing, we had discussion on the hand again. The 3D printed hand seemed to be a bit fragile to push the sticker part on our arch. Also, the printed hand was unstable because it was too light to push other things. Considering the result of this test, we decided to reject the idea of using 3D printed hand.

Polywood palm/hand

Then, we started to make polywood hand part. At first, we made very simple structure as below. Just downloading free hand data from internet and cutting it on polywood sheet, then putting the hand part into the end plate.

First trial

However, I found the problem that the connection between the hand and the end plate was a bit unstable for pushing the arch. So I had to make them more stable. To solve this problem, I changed the design; added the support parts as the image below:

Drawing support parts on Inkscape

Fixed hand part

And the production was as the image above, and I was satisfied with the output. The hand parts become enough stable to be functionable for our project.

The arm end plate onto which the palm/hand

During the construction and motion test of the frame, we were faced with other problem. It was the lack of the strengthness.

Several parts broke during the test as below images, especially the end plate and the rods holder.

broken end plate

To strenghen the end plate, I decided to change the material from polywood to acrylic plate because much pressure would be added to the point during the motion of the machine. After changing the material, we got enough strengthness to move.

Acrylic end plate

The fiberglass rods holder

Same prombem as above happend on the fiberglass rods holder.

To fix the fiverglass rods, which support the motion of 3-axis arm, we made the rods holder as below and made it of 4mm polywood at first.

fiberglass rods holder

However, the parts received much pressure during the machine moving, the first part had broken as the image below.

Broken part

To solve this image, I make up the solution using two layer of polywood for the rods holder. As a result, the parts become stronger enought to fix the fiver glass rods.

Double layer rods holder

Direction of the answers

The point where I was not satisfied with the machine is that the directions of answers are differennt from each other, so users have to move around the machine after send their questions to the machine.

In this project, we decided to make the answer board by arch, considering the constitutional limit of each parts, especially the reachness of the arm and hand. If I can improve this machine, I would like to change the structure as below, which means all answer card face in the same direction, so that users are able to see the answer from same postion without moving around.

User can see answers from one side
because all answer cards face in the same direction