On exploratory grounds

Robotic bubble making II

An material and machine exploration through iterative design

with ABB-IRB120

Can approachability help us become active participants in emerging technology? Must innovation and technology always be rigid, technical, and hard? Or can they be sensible, soft, playful, and inviting? This independent study builds upon initial ideas explored in ITAR Fall 24.

The project looks at a robotic approach to collaborative form-finding with materials that allow for open-ended experimentation. The goal is to push the boundaries of how robotic arms, often seen as rigid and industrial tools, can be repurposed to create engaging, enticing, and approachable experiences. By embracing playfulness, parametric explorations to bring robotics closer to design and the playground for discovery.

End of the arm

The first challenge is the end of the arm. The aim is to create a reliable base tool that can be customized further for the different types of materials and different configurations of the tools.

Tool definition

The base is a 60x40x12mm plywood piece counterbored on both sides for the four M5 screws that will attach the tool to the ABB IRB 120. An additional M6 screw will attach the aluminum extrusions to extend the tool, giving additional support and standarizes the point for the parametric path design. This allows for 1/4 - 20 bits to complement the attachment method to the final piece for the tool. The bubble wands will be attached to the tool.

Algorithm Development

Simulated different movement patterns and tested various paths, including spikes, circles, and varying velocity circles along the x and y axes.

The experiments revealed that different zones affected the movement precision. A more "loose" zone allowed for freer movement, while a "tight" zone constrained movement and increased precision.

Path planning

The procedural loop for each experiment follows this overall structure:

Home position> Move near the bubble solution> Dip the wand into the solution> Wait for confirmation on the flex pendant> Move to the second pre-established position> Make the intended movement> Return to the home position

Bubbles and robotic bubbles:

Recent work in HRI (Human-Robot Interaction) and computational design is increasingly concerned with the role of interaction in shaping meaningful engagement with technology. Interaction is often framed through the end-user experience. We overlook an important space, where developers, designers, and curious minds build knowledge with the tools, through the tools.

The process...

Set up

Working loop

Clean up

This process unfolded as an iterative loop both for me, the human operator, and for the code itself. I documented my workflow as a time-lapse, capturing three main stages

“How much can play and fragility help us learn and connect?” ABB IRB 120. Photographed using Sony α7 III, f/2.8, 1/8 sec, ISO 80.

Snapshots of the ABB IRB 120 robot and the bubble enviroment

Computer Vision Explorations

processed the footage using various CV filters to gain new perspectives on the interaction between the robot, the material, and me. Some of the most promising filters included:

Canny Edge Detection, Thresholding, Depth Mapping (Sobel Gradient), Infrared Color Mapping, Negative Imaging.

Infrared Color Mapping

Depth Mapping (Sobel Gradient)

Canny Edge Detection

What could CV reveal or miss? What might Babs interpret visually, and how could that affect our interaction?

A research blog (flow)

Stack, on going research

Date: May 2025 | Location: No siteType: Research Project | Professor: Josh Bard

Google Sites

Report abuse