8. SOFT ROBOTICS¶

This class introduces the emerging field of Soft Robotics, where robots are made from flexible, compliant materials instead of rigid structures.
(mostly) Inspired by nature, soft robots mimic the movement & adaptability of living organisms.
We explore bio-inspired design through the creation of soft actuators, sensors, & grippers. The focus is on understanding material behavior, pneumatic control, & fabrication techniques using silicone, TPU, &/or other felxible materials.
Applications range from wearables & rehabilitation devices to surgical tools, rescue systems, & environmental exploration.
REFERENCES & KEY TOPICS¶
- Biomimicry – learning from nature to inform design
- Pneumatics – principles of air pressure & control in actuation
- Flexures – compliant mechanisms for motion & deformation
- Locomotion – movement strategies inspired by living organisms
- Molding & Casting – fabrication techniques for soft structures
INSPIRATION & REFERENCES¶

MOTORSKINS¶
Motorskins is a studio working in the field of soft robotics, developing textile-based pneumatic systems that move like artificial muscles based in Berlin, Germany. Their work merges design, technology, & biomechanics to create wearable, adaptive structures—garments that actively respond to the body & its movements.

SASKIA HELINSKA¶
Saskia Helinska’s final project explores soft robotics in wearables, focusing on inflatable textile structures that provide visual & haptic feedback. Through material-driven experimentation, she investigates how soft, flexible materials can create movement & tactile interaction on the body. web: helinska
PREVIOUS WORKS¶
During the Fabricademy Bootcamp in May 20205 in Brussels, I gained my first hands-on experience with soft robotics while @ the Fabricademy stand during the Maker Faire Brussels.
We experimented with thermo-adhesive vinyl by cutting shapes from parchment paper & sandwiching them between 2 layers of vinyl. Using a household iron, we fused the layers together — the areas covered by parchment paper remained unsealed, creating air chambers that could be inflated.
ASSIGNMENT: WEEK 08: SOFT ROBOTICS¶
PUSH SHRIMP¶
My goal of the “shrimp project” was to create a small weeble-like object that could stand upright & return to its position.
= push puppet / string puppet toy a toy that folds up when pressed & returns to its original shape when released.
References:¶
- youtube: Make a Push Puppet (3D printed) - Ynze van der Spek
- youtube: Lego Push Puppet Fun Build - Lego Maniacs
- youtube: Thumb push puppet man on a chair - Roarke's toys and scale models

For this part of my soft robotics assignment, I used a vacation photo of a small shrimp on the beach, which a friend had sent me. I first translated the shrimp into a 3D model using Studio Tripo

STL FILE¶
I had to separate the shrimp into individual parts. I first attempted this in Rhino CAD, but I quickly reached technical limits—for example, after splitting the geometry, the parts were often open surfaces instead of closed volumes.
So I tried a different strategy: I manually cut the image of the 3D shrimp in Photoshop & then re-generated the seperate segments in Studio Tripo, which gave me clean individual pieces I could continue working with.

Preparing the Shrimp Segments for the Push-Puppet Mechanism¶
The next step was to prepare each shrimp segment in Rhino CAD so that a functioning push-puppet mechanism could be integrated. I added small internal channels (thin hollow tubes) along the inside of the segments to guide the string system that will later allow the figure to collapse & rebound. These channels ensure that the thread can run smoothly through the body without becoming visible or interfering with the movement.

In addition to this, I also designed the base, the central push-rod, & the spring mechanism directly in Rhino.
The pedestal houses the vertical tension rod & the internal string routing, allowing the shrimp to behave like a classic push puppet: collapsing when pressed & popping back into shape when released.

Slicing & Material Choice for 3D Printing¶
After preparing the individual shrimp segments in Rhino, I imported all parts into Ultimaker Cura to slice & prepare them for 3D printing.
For this project, I selected a WOOD Filament
Ultimaker Cura – Print Settings for Wood Filament (PLA-based)¶
| Parameter | Setting | Notes |
|---|---|---|
| Material type | Wood filament (PLA-based, RoHS compliant) | Contains wood fibers; behaves similar to PLA |
| Filament diameter | 1.75 mm | Manufacturer specification |
| Nozzle diameter | 0.4 mm | Larger nozzles (≥0.5 mm) recommended to reduce clogging |
| Layer height | 0.2 mm | Balanced detail and print stability |
| Printing temperature | 200 °C | Within recommended range (190–220 °C) |
| Build plate temperature | 50–60 °C | Improves adhesion; optional for PLA-based materials |
| Print speed | 40–50 mm/s | Reduced speed improves surface quality |
| Retraction | Enabled | Short retraction distance to avoid clogging |
| Cooling fan | 100 % after first layers | Ensures clean details |
| Infill density | 100 % | Suitable for non-structural parts |
| Infill pattern | Grid | Even material distribution |
| Build plate adhesion | Brim | Prevents warping and improves first-layer stability |
| Supports | Normal | Depends on geometry |
| Slicing software | Ultimaker Cura | Used for G-code generation |
The WOOD Firmalent visual appearance & tactile quality closely imitate traditional wooden toys.
(Actually, it smells very pleasant while printing: like warm wood & funnily enough, a little like popcorn)

Evaluating Print Quality & Switching to Resin¶
Unfortunately, the wood-filament print came out rough & imprecise, especially along the small internal tubes that were meant to guide the string. The uneven surface & slight inaccuracies prevented the thread from moving smoothly through the channels.
Because of this, I decided to reprint the entire model in resin, which offers much higher precision and crisp, clean internal geometries—ideal for the delicate mechanics of a push-puppet system.
For this, I had to reconfigure all individual parts using a different slicer specifically designed for resin printers. I used Anycubic Photon Workshop, as resin printing requires a fundamentally different preparation process compared to FDM printing. In particular, the orientation of each part & the correct selection & placement of support structures are crucial.
This is because resin prints are formed layer by layer while being pulled out of the resin vat, meaning that insufficient support or poor orientation can lead to deformation, failed prints, or damaged fine details. Proper alignment helps reduce suction forces, preserves delicate geometries, & ensures dimensional accuracy—especially important for moving parts that rely on tight tolerances & smooth interaction.

For the print, I used Anycubic Standard 3D Printer Resin - grey in combination with the ANYCUBIC Photon Mono M5s resin printer.
The following settings were used for this print:
| Parameter | Setting |
|---|---|
| Resin | Anycubic Standard 3D Printer Resin – Grey |
| Printer | ANYCUBIC Photon Mono M5s |
| Layer Thickness | 0.05 mm |
| Normal Exposure Time | 2.8 s |
| Light-Off Time | 0.5 s |
| Bottom Exposure Time | 25 s |
| Bottom Layers | 5 |
| Z-Lift Distance | 8 mm |
| Z-Lift Speed | 6 mm/s |
| Z-Retract Speed | 6 mm/s |
| Anti-Aliasing Level | 1 |

Issues Encountered With the Resin Print¶
Unfortunately, I also ran into several problems during the resin printing process. Some segments were incomplete, others printed only halfway, & many edges came out slightly rounded, which created unwanted gaps between the pieces.
These issues were likely caused by the fact that the resin tank had not been cleaned properly, & the protective film showed small punctures. As a result, resin was able to seep underneath the film, which interfered with the exposure process & led to inaccurate, inconsistent prints.
Due to the limited time, I decided not to continue developing this short project for now & instead focus on the next SOFT ROBOTICS short project.
INFLATABLES¶
SHRIMP¶
For the second project of this assignment, I created a small inflatable vinyl shrimp based on the same geometry.
To do this, I transferred the shape of the shrimp onto white baking paper by holding it against my laptop screen & tracing the outer contours. The shape was then cut out, & the paper shrimp silhouette was placed between 2 layers of thermo-adhesive vinyl.
It is important to maintain sufficient spacing along all edges, leaving a minimum margin of 1 cm around the perimeter.
Additionally, an inlet channel was created using a 2. piece of baking paper, positioned only on the lower side of the template. This channel measured approximately 2–3 cm in length & 5 mm in width, allowing a tube or straw to be inserted after sealing in order to inflate the air cushion.
The thermo vinyl has 2. different sides. The more glossy sides were placed facing each other, allowing them to fuse together when heat was applied. The baking paper inserted between the vinyl layers prevented bonding in those areas, resulting in air pockets wherever the paper was present and thus creating internal cavities.
The layers were heat-sealed using an industrial iron at 150 °C for 15 seconds. The sealing time needed to be long enough for the vinyl layers to bond & form a clean, continuous edge around the baking paper template, but short enough to prevent damage to the vinyl & the formation of air bubbles.
The milky, matte appearance of the outer vinyl surfaces comes from a protective layer, which was removed after the heating process &, if necessary, after further cutting.
The remaining unsealed areas formed narrow inflatable air channels, while the sealed surfaces remained relatively large & flat.

| Step | Process Stage | Description | Purpose |
|---|---|---|---|
| 1 | Template preparation | A shape is traced onto baking paper & cut out | To create a paper template. |
| 2 | Layer setup | The paper template is placed between 2 layers of thermo-adhesive vinyl | To define areas that remain unbonded. |
| 3 | Edge spacing | A minimum spacing of 1 cm is maintained around all edges | To ensure airtight and stable seams. |
| 4 | Inlet channel creation | An additional piece of baking paper is positioned at 1 edge (2–3 cm × 5 mm) | To allow later inflation using a tube or straw. |
| 5 | Material orientation | The glossy sides of the thermo vinyl are placed facing each other | To enable effective heat bonding. |
| 6 | Heat sealing | The layers are heat-sealed using an industrial iron (e.g., 150 °C for 15 s) | To permanently bond the vinyl layers while preserving internal cavities. |
| 7 | Air chamber formation | Areas separated by baking paper remain unsealed | To form internal air chambers & channels. |
| 8 | Protective layer removal | The protective backing of the vinyl is removed after heating & trimming | To reveal the final surface finish. |
| 9 | Final structure | Sealed areas remain flat while unsealed regions inflate | To create a controllable inflatable structure. |
PATTERN¶
Next I created folded paper-cut shapes that were inspired by classic origami folding patterns.
Here, the fabrication followed the same step-by-step workflow as outlined below:
Template preparation → Layer setup → Edge spacing → Inlet channel creation → Material orientation → Heat sealing → Air chamber formation → Protective layer removal → Final structure
I tested step by step & gradually removed unnecessary material to achieve cleaner airflow & more controlled inflation behavior.
PINE CONE¶


Inspired by the opening movement of a pine cone, I explored how natural mechanisms can inform soft robotic motion.
Therefore, I drew on the geometry of pine cones & generated different pattern variations with the help of ChatGPT, transferring these patterns onto parchment paper to use as displacement zones between the thermo-adhesive vinyl layers.
Interestingly, the resulting inflatable shapes looked less like pine cones & instead began to resemble leaf or feather structures, opening up a different aesthetic direction than originally planned.

3 DIMENSIONAL Inflatables¶

The next step was to investigate whether this technique could also create multi-part, 3-dimensional objects.
Surprisingly, it worked very well!
I created an additional half-leaf shape, attached it along the center line & added another layer of thermo-adhesive vinyl on top. The new “side panels” could then be folded outward step by step, allowing me to reach all areas with the iron & seal the layers cleanly.
BRANCH - DEVELOPED PATTERNS¶
For the last experiment with thermo-adhesive vinyl, I designed a sequence of progressively developed patterns.

PDF¶
The first version was a simplified “branch structure” — a stylised stem with finger-like extensions.
In the next iterations, I gradually added circles, rings & diamond shapes to observe how these interrupted geometries would influence the airflow & therefore the movement of the inflatable object.
I created the shapes / the pattern in Rhino & laser-cut the parchment paper templates.
Because the laser’s ventilation system can easily move such lightweight material, I had to weigh down the edges to keep the paper in place. I also structured the cutting sequence carefully: all inner holes were cut first, & only at the very end the outer contour. This prevented the pieces from shifting & ensured that every hole stayed aligned in the correct position.


For cutting the shapes, I used the CFL-CMA1080K CO₂ laser cutter.
| Function / Specification | Value |
|---|---|
| Laser Type | CO₂ laser |
| Power | 100 W (typical, adjustable) |
| Working Area | 1.00 × 0.80 m |
| Cutting Speed | 0–36,000 mm/min (variable) |
| Maximum Cutting Thickness | up to approx. 25 mm (material-dependent) |
| Resolution | up to 4000 DPI |
| Accuracy | 0.01 mm |
The following tables show the cutting parameters used for the baking paper:
Baking Paper / Parchment Paper¶
| Power (%) | Speed (mm/s) | Passes | Notes |
|---|---|---|---|
| 3–6 % | 500–900 mm/s | 1 | Minimal energy with high speed for clean cuts and reduced scorching. |
| 4–8 % | 400–700 mm/s | 1 | Slower speed may improve cut quality but increases risk of discoloration or burning. |


Conclusion – Final Inflatable Pattern Series¶
In this series, I worked with both white & transparent thermo-adhesive vinyl, & the material differences significantly influenced the results. The transparent vinyl was noticeably softer & more flexible.
It was also noticeably more difficult to control the exact ironing time with the transparent vinyl. Because of this, the edges became less crisp, & in some areas small air bubbles formed between the layers, affecting the overall precision of the inflatable structures.
The white vinyl, by contrast, behaved stiffer & more controlled, resulting in sharper, more defined movements.