💻 computer science

Multi-material Direct Ink Writing and Embroidery for Stretchable Wearable Sensors

This paper presents a scalable textile-compatible fabrication workflow that integrates multi-material direct ink writing with automated embroidery to create durable, stretchable strain sensors embedded directly into garments for accurate motion tracking in sports and rehabilitation applications.

Original authors: Lukas Cha, Ryman Hashem, Ria Prakash, Tanguy Declety, Wenze Zhang, Liang He

Published 2026-03-20

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Lukas Cha, Ryman Hashem, Ria Prakash, Tanguy Declety, Wenze Zhang, Liang He

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you want to build a "smart shirt" that can track your movements while you run, play tennis, or recover from an injury. The problem is, most electronic sensors are like tiny, rigid bricks. If you sew a brick onto a stretchy t-shirt, the brick will either pop off when you move, or it will stop the shirt from stretching, making it uncomfortable.

This paper presents a clever new way to solve that problem by combining 3D printing with embroidery. Think of it as teaching a sewing machine to "draw" with soft, squishy materials and then stitch them right onto your clothes in one go.

Here is a simple breakdown of how they did it and what they found:

1. The "Soft Sandwich" (The Sensor)

Instead of using hard plastic, the team printed a tiny sensor that looks like a soft sandwich:

The Bread: Two layers of soft, stretchy silicone (like a very squishy rubber band).
The Filling: A layer of conductive carbon grease (a thick, black goo that conducts electricity).

They printed this "sandwich" directly onto a piece of fabric using a modified 3D printer. Because the materials are soft, the sensor can stretch and bend just like your skin.

2. The "Magic Stitch" (The Integration)

Usually, after you print a sensor, you have to glue it or sew it on manually, which is messy and weak. This team used an automated embroidery machine to do the heavy lifting.

Think of the embroidery machine as a super-fast, super-precise needle that does two jobs at once:

Job A (The Anchor): It uses regular thread to stitch the edges of the soft sensor tightly to the fabric. This is like using a strong safety pin to hold a patch in place so it doesn't fly off when you stretch.
Job B (The Wire): It uses special conductive thread (thread that carries electricity) to stitch a path from the sensor to the outside world. This connects the sensor to a computer without needing any extra wires or glue.

It's like the machine is simultaneously sewing the patch on and plugging it into the power outlet in one smooth motion.

3. How Well Does It Work?

The team tested this "printed-and-stitched" sensor by stretching it, pulling it, and wearing it on elbows and knees.

Stretchiness: The sensor could stretch to 120% of its original length (more than double its size!) without breaking. That's like stretching a rubber band until it's twice as long, and it still works.
Accuracy: When stretched up to 60%, the sensor was very predictable. If you stretched it a little, the electrical signal went up a little. It was almost a perfect straight line (99% accuracy in that range).
Durability: They pulled it back and forth 80 times. It held up well, though the signal drifted slightly (like a clock that gains a second every day).
The "Tug-of-War" Test: When they pulled the fabric until it ripped, the fabric broke first, not the stitches or the sensor. This is a huge win! It means the sensor is attached so well that the clothes themselves are the weak link, not the technology.

4. Real-World Test: The "Smart Sleeves"

They sewed these sensors onto elastic sleeves for elbows and knees. As a person bent their arm or leg, the sensor tracked the movement.

The Result: The sensor could tell the computer exactly how much the joint was bending.
The Catch: It wasn't perfect. The error rate was about 17%. Why? Because when you bend your knee really far, the sensor stretches beyond its "comfort zone" (the 60% linear limit), and the signal gets a bit wobbly. Also, the sensor was sensitive to temperature changes, like how a rubber band gets stiffer in the cold.

The Big Picture

This paper isn't about creating the perfect sensor yet (it still has some quirks). Instead, it's about proving a new way to make them.

The Analogy:
Imagine trying to put a solar panel on a wetsuit.

Old Way: Glue a hard solar panel on. It cracks when you swim, or the glue fails.
This New Way: Print a soft, flexible solar panel directly onto the suit, and then use a sewing machine to stitch it down and wire it up instantly.

Why it matters:
This method opens the door for mass-producing "smart clothes." Instead of hand-sewing tiny electronics into every shirt, factories could use robots to print and stitch sensors directly into sportswear, medical gear, or soft robots, making them comfortable, durable, and ready for the real world.

1. Problem Statement

The integration of stretchable sensors into textiles for sports performance, rehabilitation, and injury prevention faces significant challenges:

Integration Limitations: Most Direct Ink Writing (DIW) sensors are printed on rigid beds and require secondary attachment (adhesives, lamination) to garments. These methods often compromise fabric breathability, comfort, and durability.
Interfacing Gaps: Existing textile sensors (e.g., conductive yarns) often lack robust electrical interfacing strategies that are mechanically stable and compatible with multi-layer printed structures.
Scalability: There is a lack of a unified, automated workflow that combines soft-material printing with industrial textile manufacturing (embroidery) to create reproducible, garment-ready sensors.

2. Methodology

The authors propose a hybrid fabrication workflow that integrates Multi-material Direct Ink Writing (DIW) with Automated Embroidery to create sensors directly embedded in textiles.

A. Sensor Fabrication (DIW)

Materials: A three-layer stack consisting of Ecoflex 00-30 (silicone substrate) and conductive carbon grease (sensing element).
Printing Process:
- Performed on a modified Ender 5 Plus 3D printer equipped with dual ViPro 3 progressive cavity pumps.
- Geometry: A 25 mm × 58 mm sensor with a 13 mm × 30 mm strain gauge pattern.
- Structure: 0.3 mm layers of silicone, carbon grease, and silicone.
- Process Control: The print bed was heated to 60°C to accelerate partial curing of the first layer, ensuring structural integrity for subsequent layers.

B. Textile Integration (Embroidery)

Hardware: A Brother PR1055X embroidery machine with custom Hatch 3 software designs.
Dual-Function Stitching Strategy:
1. Mechanical Anchoring & Electrical Interconnect (Pattern P1): A dense, triple-run line stitch at the sensor's top and bottom. This uses conductive thread to create electrical traces from the sensor pads to external contacts while mechanically fusing the silicone to the fabric.
2. Stretchability Preservation (Pattern P2): A sparse, single-run zig-zag stitch along the sensor's stretch axis. This stabilizes the sensor against out-of-plane motion without restricting the fabric's natural elasticity.

3. Key Contributions

Hybrid Workflow: First demonstration of a fully automated process combining multi-material DIW with embroidery to simultaneously achieve mechanical fixation and electrical interfacing in a single step.
Direct-to-Garment Integration: Eliminates the need for post-print adhesives or lamination, preserving the textile's breathability and comfort.
Robust Interface Design: The specific stitch patterns (dense anchoring vs. sparse zig-zag) address the trade-off between secure attachment and maintaining the stretchability of the underlying garment.

4. Experimental Results

The sensor was characterized via tensile testing (80 cycles) and stretch-to-failure tests, followed by a wearable application on elbow and knee sleeves.

Performance Metrics

Metric	Value	Significance
Linearity ( $R^2$ )	0.990	Highly linear response up to 60% strain.
Gauge Factor (Sensitivity)	31.42	High sensitivity comparable to literature standards.
Hysteresis	22.9%	Attributed to viscoelasticity of silicone and carbon network reformation.
Stretchability	120%	Sensor maintains electrical continuity up to failure.
Drift (Baseline/Peak)	0.135% / 0.236% per cycle	Moderate stability over 80 cycles.

Mechanical Behavior

Failure Mode: In stretch-to-failure tests, the textile fabric failed before the embroidered interface or the sensor itself, proving the robustness of the embroidery-based anchoring.
Force Requirements: Minimal force (<20 N) is required to stretch the sensor to 60% strain, ensuring it does not restrict wearer mobility.

Wearable Application (Elbow & Knee)

Correlation: Clear correlation observed between normalized resistance change ( $\Delta R/R$ ) and joint bending angles.
Accuracy: Mean Absolute Percentage Error (MAPE) was 17.2% (elbow) and 17.9% (knee).
Limitations: Higher error at high flexion angles (knee) occurred when strain exceeded the linear range (>60%). The sensor also showed temperature sensitivity.

5. Significance and Future Work

Scalability: This approach provides a reproducible pathway for incorporating printed soft electronics into industrial textile manufacturing, moving beyond lab-scale prototypes.
Application Potential: The system is viable for motion capture, soft robotics, and athletic performance monitoring where comfort and durability are paramount.
Future Directions:
- Investigating alternative conductive materials (e.g., conductive silicones) to reduce hysteresis, drift, and temperature dependence.
- Exploring capacitive sensing within the same framework.
- Optimizing stitch geometries and developing sensor arrays for multi-degree-of-freedom motion tracking.

In conclusion, this paper successfully demonstrates a scalable, automated method for creating high-performance, stretchable wearable sensors that are mechanically and electrically integrated directly into garments, overcoming the limitations of traditional attachment methods.