Project 1: Auto-Graph


A few of my initial ideas involved iron filings and ferrofluid, as well as magnets attached to a clock motor.

Some initial ideas

However, I had to scrap these ideas upon realising that iron filings and ferrofluid are difficult to obtain in Singapore.

Sketches for magnets rolling idea

My next idea involved a tilted board with elevated surfaces upon which magnets could roll down. On top of this would be another tilted board with magnetic paint/metal items on top of paint blobs, which would be the method to make marks as the magnets rolled down beneath.

Above is the prototype for the board beneath, designed to have multiple magnets rolling and knocking into each other to create a chain reaction so as to create a more interesting piece on the surface layer. I used coins for the time being to test if the prototype’s lanes would be feasible.

However, after much research and testing, I discovered that this idea would not work as the force of the magnets was either not strong enough to pull the metal items along the top as they rolled down, or if the force was strong enough then the magnets would lift off the board entirely and stick to the metal items on the surface instead of rolling down.

Hence, I had to change my idea entirely. While researching on the patterns I could make for the board where the magnets would roll down, I came across a pinball machine surface and was inspired to make a pinball machine instead, with the pinball as the main tool to make the drawing.

The biggest problem would be getting the paper onto the machine (to get the drawing) while still having all the usual obstacles that the pinball would bounce against. As these obstacles are usually set on the surface of the machine, I decided to make an extra cover on which the obstacles would be placed so that paper could still be placed on the machine and removed after the drawing was made.


I used this video as reference to make the flippers. I decided to go with using cardboard boxes to make the machine instead of wood as I was unsure whether I would be able to get the exact diameter of the chopsticks through the sides using a drill. I had the impression that using cardboard would be more precise as I was poking the chopsticks straight through.

Flipper components

I started making the components on the bottom with styrofoam as these had a sturdier form than cardboard, which I would have to cut and paste to achieve the thickness required, and would easily fold under pressure. I worried that styrofoam may be the same where it might crumble if too much pressure was applied, but it maintained its structure and so I decided to go ahead with using it.

In progress of setting up the flipper mechanism


Flipper mechanism sans rubber bands

The biggest problem I had with the mechanism was that the styrofoam kept getting ripped from the box despite being superglued or hot-glued down to the box. It was not strong enough to withstand the force of the rubber band, hence I decided to get wood pieces instead to replace this component.

Another problem I had was getting the rubber bands to work – I realised that I had bought the wrong kind of rubber bands – they were too big for the mechanism. To combat this, I wrapped one end around the chopstick several times and it worked.

Close-up of Push Mechanism
Top View of WIP


Front View 

I used a translucent polypropylene sheet for the cover as I wanted the obstacles and drawing to be visible so that the player could see what they were doing and making as they played on the machine. I used this same material to form the sides, back and surface of the machine to keep the box tilted at an angle.

Side 1
Side 2
Top View with Cover
Top View without cover, after drawing machine has been used

This was the first test, in which I put the purple paint on the paper for the pinball to pick up as it traversed the machine. The results were unsatisfactory as it only made dot-like marks and it did not really track the path of the ball. I added water to the surface of the paper in an attempt to help the ball move around with less friction, but it did not work and only caused the paper to wrinkle.

I tried this again but with pink paint at the side of the machine itself and not on the paper. It worked better as the ball picked up the paint more easily and its path was trackable as can be seen from the pink lines in the photo.

Acrylic paint mixed with a little water

The medium I used to make the painting was acrylic paint mixed with a little bit of water, as acrylic paint alone was too thick and dried too quickly, causing the pinball to stick to the paper/machine instead of rolling down.

This was the first successful piece in which the ball’s movements were tracked using the paint and the areas where the flippers hit it or where the ball knocked against the obstacles are visible. The most common path is also demarcated by the thickest paint lines.

This was the result of the demonstration and try-out done on presentation day – not many marks as people only tried it out for a few seconds.

This piece was made after adjustments to the machine to direct the ball better and the flippers were fixed to be fully functional.

This piece was made on tracing paper, unlike the rest which were made on normal opaque paper.

This is an orthographic plan view drawing of the machine.

This is a combination of the drawing made by the machine and the drawing of the drawing machine itself.

Final Thoughts

I am glad that it worked as I spent a lot of time agonising over the flippers and whether or not they would have enough power to propel the ball upwards into the playing field.

On hindsight, perhaps I should have used wood after all as that would give the machine a more complete finish – or used opaque polypropylene sheets instead of translucent ones. However, I am satisfied with the end product and have learnt much about form and mark-making using different mediums and methods.