My progress to date has been a series on tests, both with and without the robot, to determine the controls I will use for the easel, canvas, and paint balloons.

 

For these tests with the robot, I set up an easel, each time in the same location, that held a canvas with a balloon stapled on from the top.  I used cable-ties to secure an exacto knife to the tool and jogged the robot to cut the balloons on the canvas.

 

 

 

Test 1 – Downward Cut:

– robot moved tool-head to top portion of balloon and moved down through balloon

– didn’t take into account the slant of the canvas on the easel; cut through the canvas

– downward streaks, through somewhat spread out from balloon itself

– gives the most diversity of paint movement of all the robot tests

 

 

Test 2 – Poke:

– robot moved tool-head from a point just in front of the center of the balloon forward into the balloon

– paint came out more in one blob and less splattered

– again, movement on paint was mainly straightdown

 

 

 

 

 

Test 3 – Poke & Twist:

– robot moved tool-head from just in front on balloon to just inside, then turned its 6th joint to robot the blade in the balloon

– paint came out in one blob and came out the heaviest on all the tests

– minimal splatter

– broke tool-head

 

 

 

Test 3 seemed to cause the tool-head too much stress because on the twist into the balloon and broke the tool-head off:

 

Off the robot, I mainly experimented with created smaller punctures in the balloons and applying pressure to the balloon in various directions to get varied splatter. This was quite messy and unpredictable as it caused a lot of the paint to land elsewhere from the canvas, but when it landed on the canvas it had a nice effect. For this reason, moving forward I want to combine a version of this test with test 1 on the robot: the robot will use the downward cut to open the balloons up and have a way to move the balloon around the canvas to splatter the extra un-escaped paint still in the balloon.

 

The next steps on my project will involve building a two-headed tool (if I can get someone with experience to help me as I have never done this before) with one head able to hold an exacto blade, and the other head a smoothed, thick end so as not to pop the balloon but move it around. In code I will program the movements on the robot based on the position of the balloon and be sure to take into account the slope of the canvas.