Skin Electronics¶
RESEARCH PART¶
Skin electronics, which is also called e-skin, is an emerging trend in technology. It is a soft and stretchy electronic material that can be worn on the human body like a second skin. Now we go to more easy-to-wear and easy-to-use devices that can help us in our body movement and healthy control, use for communication, when we are ill, or in anything else. Now every global company is working and adding new and new technologies.
In skin electronics they usually use stretching and flexible materials with electronic sensors. Some e-skin materials also incorporate nanotechnology, such as carbon nanotubes, to improve their conductivity and sensitivity.
In the last years we see more and more skin electronics and wearables in sport, not only in professional sport but also in the usual person’s life who loves sport. And in any case a lot of people love to use this kind of electronics because of the interesting and unique design, like an accessory.
The most interesting and unbelievable part is the medical use of skin electronics. A sense of touch can be given to robots, which can help perform medical surgeries even without human contact. Or doctors can make operations from a distance with a skin-electronics mechanism — when the doctor moves the hand, the robot does the same, but **more correct, without human shaking or small movements, to be more precise in action.
E-Skin Systems in Human-to-Robot Control¶
A great example of this could be a project from 2022 called “Electronic Skin Lets Humans Feel What Robots Do — And Vice Versa.”
What they say: “An integration of soft materials, sensors and flexible electronics is bringing robotic ‘skin’ closer than ever to reality.”
Humans have millions of nerve endings that sense heat and touch. This helps us feel and investigate the world around us. For around 40 years, scientists and engineers have tried to create something similar to the human body with all its functionality and integrate it into robots, but it is still impossible. Now the work is going the other way — the way of e-skin.
When we attach sensors to the human body, they read our movements and reproduce them in robots. It is a really strong and promising development branch that can give us robots able to do very hard tasks, and in the future allow robots to learn from human movements — how something should be done and in which moment the most common or correct action is.
By doing one thing, we develop two important directions in this field: robot control at a distance, and robot training.
Touch and temperature sensors were the first to be developed for this kind of system. Wei Gao, a biomedical engineer at the California Institute of Technology, decided to try combining these sensors with ones that could detect chemicals.
Gao’s lab used an inkjet printer to layer a specialized ink made of nanomaterials—mixtures of microscopic bits of metals, carbon or other compounds—within a soft hydrogel base.
Main Types of Skin Electronics¶
We spoke about one tupe of e-skin but we have some uther points which are not less interesting they the previous one. Lets see and speeak about them.
Skin electronics main types:
- Stretchable sensor skins
- Conductive elastic materials
- Electronic tattoo skins
- Self-healing e-skin
- Hybrid e-skin
Stretching sensor is a very ՐՐflexible oneՐՐ. You attach it to the skin and it stays there, moving together with your body. It can stretch, bend, twist, and still read body signals. The previous one is a great example of this type. These sensors are mostly used for health and sport-tracking purposes.
Conductive elastic materials are made from soft, skin-like materials with added conductive elements. They can easily transfer signals and do not change their flexibility. This type is mostly used when high sensitivity is needed or when e-skin is applied on soft robots.
Electronic tattoo skins look like temporary tattoos. They are the most comfortable ones to use — ultra thin and very light. They sit directly on the skin and can read muscle activity, skin temperature, breathing, or anything else. They feel so natural that people don’t even notice them, just like a real tattoo.
Self-healing e-skin is the only one that is “reproductive.” It can have small cuts or cracks, but the material sticks back together and continues working. This type is very common for long-term wear or for robots working in risky environments.
Hybrid e-skin is a mixed type. It can include micro-electronic parts together with conductive materials or other elements. It gives both high durability and very high accuracy. It can be used for reading tiny body movements or for robotic tasks.
WHAT INSPIRES ME¶
In cooperation with the Electronics Research Institute of Shizuoka University (Japan), Yamaha Corporation (Hamamatsu, Shizuoka, Japan) is investigating the use of millimeter-long multi-walled carbon nanotubes in elastomeric strain sensors for gloves.
The musical instrument manufacturer is not only looking at virtual reality applications where lightweight, soft data gloves could replace bulky, rigid alternatives, but they would also naturally tailor the data gloves for musicians, whose finger movements could be better monitored with a thin, stretchable, sensor-based fabric.
They used a sheet of unidirectionally aligned MWCNTs impregnated with an elastomer (urethane resin), making electrodes with a conductive paste at each end to form sensors a few centimeters long that could be stretched to twice their original length. Only a few micrometers thick (up to 200µm with an additional layer to support the device's elasticity), the resulting strain sensors had a response time of less than 15ms with a measurement factor greater than 10 (high sensitivity).
The researchers then developed a prototype sleeve incorporating the CNT strain sensor connected via conductive silver synthetic fibers within an elastic compression fabri. They also devised a soft data glove capable of detecting fine movements of the finger joints without hindering the user's movements.
“The data gloves introduced in our article were specifically designed to detect finger movements in a musician’s dystonia (a neurological movement disorder) under a joint study with a Japanese university.”
I took this information from the Music & Market site. Check it if this project interests you.
WHAT WE MUST KNOW BEFOR STARTING THE PROJECT¶
Sensor Types¶
Hear are the main types of sensores.
The most common sensors are pressure sensors, light sensors, motion sensors, thermometers, and gas sensors, which we often see in our homes or everyday life but usually don’t notice at all. In lab projects we usually use some of these sensors, plus capacitive sensors, magnetic sensors, Hall-effect sensors, and others — but not as often.
Every sensor is interesting to use and to make, and you must choose one by thinking first about which field you will use it in and what exactly you need it for. The easiest and most usable sensor to make is a pressure sensor.
Let’s make the pressure sensor together.
How it works — spoiler: in a very simple way.
What happens when you press the sensor:
- The two conductive layers get closer to each other,
- the resistance decreases,
- and the microcontroller (Arduino, Adafruit Flora, etc.) reads a different value.
More pressure → less resistance → higher sensor reading.
What you should have ideally:
- Wire striprers
- scissors
- tape
- 2 jumper wires
- 2 pieces of copper tape
- fabric
- piece of velostat
How to make it — you can see everything in this video:
IT'S TIME TO DIVE INTO PROCESS¶
Electronic Tattoo Skin¶
The first idea I had was about stick-on “tattoos” that people can use when they’re ill and need to control their body temperature during the day. I was thinking: what if we create a tattoo film that changes color depending on body temperature, starting from 36°C up to 39°C?
Sometimes people can’t properly detect their temperature because of illness, or they’re simply too busy to check it. But it’s very important to understand how your body feels and what it needs. For example, sometimes we have 37°C and don’t even notice it — but that’s already a sign that our body is fighting a virus. Knowing this can help us rest on time, take vitamins, adjust our diet, and support our body to become stronger and recover faster.
First I cheak how to create conductive ink that changes color with body temperature. I found some instructions how to create. hear is some of them ⬇
-
Which thermochromic systems work at 35–40 °C
-
Leuco-dye microcapsules
Work mechanism: it is composed of three components — the leuco dye, the developer, and the co-solvent. When these components are mixed together, the material changes its color or the intensity of the color. The change usually happens between 31°C and 39°C, somewhere in the middle of this range.
Typical behavior: the working temperature range is about 3 to 6°C, and you can arrange the temperature shift yourself — for example, make it change at 31–37°C or 36–40°C.
-
-
Thermotropic Liquid Crystal films or pigments. They have vivid spectrum shifts. This material changes color from red to green and then blue as the temperature changes. You can set this reaction to work in the range of 35–40°C.
-
Which option I recommend for you. The range is from 35 to 40 °C:
I think the best way is a two-layer system. It is easy, works good, and you can control the colors very well.
-
Base layer: Make the conductive line first. You can use PEDOT:PSS if you want a soft and wearable trace, or carbon/silver ink if you need it to be more strong and with lower resistance. PEDOT:PSS is good because when the body gets warmer its conductivity also changes a little, so even when you put a color layer on top, it still works normally.
-
Top layer: On the top you add a very thin clear layer (really thin, like 10–50 microns) mixed with thermochromic microcapsules. Choose the ones that change color exactly in the 35–38 °C range. If you want not only color change but also rainbow shift, then use liquid-crystal pigment instead.
That’s all — simple two layers, one for the signal, one for the color. It reacts fast because the topcoat is thin, and the colors look clean and bright.
-
But all these materials are made from ingredients we don’t have in the lab, and ordering them from places under the city takes too long. Because of that, I tried to search if I could do this with materials we already have. The only thing that is conductive and can change with temperature like a tattoo is copper. Me and Anush Arshakyan tried heating copper to see how it changes and at what temperature it reacts.
What result we have:
On the sample you can clearly see the darker and orange areas — these are the parts that changed color after heating. But the problem is that this reaction didn’t happen in the temperature range we actually need (36–40°C). The color shift only appeared when the copper was heated to around 400°C, which is extremely high for the human body and obviously cannot be used as a wearable or tattoo-like sensor in daily life.
Singing Stretch¶
Resistor¶
About the resistor — how to find the correct one visually?
This part is actually easy when you know the meaning of the color bands. When you understand the code, you can pick the resistor that fits your needs and double-check it to be fully sure.
With this diagram I ejected from the board the 1 Mohm and 500 kohm resistors for future use. You can see the 1 MΩ resistor in the previous photo. I kept them because they can help later when I want to tune the sensitivity of the stretch sensor or test different movement ranges.
This resistor was originally for the pressure sensor, because I wanted to read every small movement and use it later in my project. But after some experiments I changed my direction and started making a stretch sensor instead. The main reason was that I already made a pressure sensor before, and this time I wanted to study something new for myself — something I never tried. That’s why I moved from pressure sensing to stretch sensing and continued the work with the elastic fabric and conductive thread.
Inspiration¶
After the previous fail I started to work with a new idea. At that time I found some interesting hair-design concepts with braids, and I really liked them. I guess everyone has seen Game of Thrones and knows Daenerys — the main character. In every season she had perfect, unique hairstyles.
So I started thinking: what if I make a touch-sensor that a person can activate just by touching the hair? When they touch it, the buzzer starts to “sing.” I can make a Daenerys-style braid and hide the sensor and the buzzer inside it, so every time you touch it, the famous opening sound of the show plays.
It makes the hairstyle more unique, more interactive, and honestly, more fun.
In the working process of this idea, I started to find interesting and traditional hairstyles. Many of them use fabric pieces, accessories, or just have very unique architecture. And inside those styles we can also hide our stretch-sensor and play traditional songs, cultural sounds, or any special music during dances or other activities.
I really like the idea of mixing the sensor with traditions. But this week I didn’t have enough time to explore more traditional hairstyles or find sounds that will be harmonic together. Still, I am totally sure that this project can have a very interesting and continued effect if developed further.
Work process starts¶
I took an elastic fabric and started sewing on it in a zig-zag pattern with conductive thread. Try to place them both at a considerable distance and close to each other.
How it was and what we have ⬇
This construction is elastic enough, but not too conductive as i think. I made some experiments with it, tried how it works, and started searching for the right resistor to stabilize the movement.
Then I made another fabric piece (more tiny and close sewing), but it didn’t give the effect I needed, so I went back to my first sample and continued working on that one.
And after that I started making the connections to see if all this together could work or not. Spoiler: it can, but understanding it and finding the point where it will work normally was really hard.
First, I took the code from my previous project, and Anush Arshakyan also sent me some of her codes to check, but none of them worked.
The second thing I did was change the strong 1 Mohm resistor to a 100 kohm one — or even a 68 kohm, which also works.
These have less resistance, and because of that my signal could finally pass through the circuit after the resistor connection. With the 1 Mohm the signal was basically too weak and got lost, but with the lower values it started to show the real movement.
After that reading the numbers was okay, the sensor reacted, but the buzzer didn’t start to sing at any point. I started checking new codes, resistors, and everything I could, but it still didn’t work.
The first three codes I used were mainly for testing — to check whether the stretch sensor was working or not in the Arduino IDE program. You can read more about this program in the Wearables Week section.
N1:
// DIY Stretch Sensor on Adafruit Flora
// Connect the sensor to A9 (or any analog pin)
const int sensorPin = A9; // Flora has A7, A9, A10 as analog inputs
void setup() {
Serial.begin(9600);
while (!Serial); // For ATmega32u4 boards: wait for serial connection
}
void loop() {
int sensorValue = analogRead(sensorPin); // Read analog value (0–1023)
Serial.print("Stretch sensor value: ");
Serial.println(sensorValue);
delay(50); // small delay to make data readable
}
const int sensorPin = A9; // Analog pin
int sensorValue = 0;
int mappedValue = 0;
void setup() {
Serial.begin(9600);
while (!Serial); // wait for serial connection (Flora)
}
void loop() {
sensorValue = analogRead(sensorPin); // 0-1023
// Map the value to a more sensitive range
// Adjust the input min/max according to your sensor
mappedValue = map(sensorValue, 300, 700, 0, 1023);
// Constrain to avoid going below 0 or above 1023
mappedValue = constrain(mappedValue, 0, 1023);
Serial.print("Raw: ");
Serial.print(sensorValue);
Serial.print("\tMapped: ");
Serial.println(mappedValue);
delay(20);
}
N3:
const int sensorPin = A9; // Analog pin
int sensorValue = 0;
int mappedValue = 0;
void setup() {
Serial.begin(9600);
while (!Serial); // wait for serial connection (Flora)
}
void loop() {
sensorValue = analogRead(sensorPin); // 0-1023
// Map the analog value to 0-100
// Adjust 300 and 700 according to your sensor's min and max readings
mappedValue = map(sensorValue, 300, 700, 0, 100);
mappedValue = constrain(mappedValue, 0, 100); // keep within 0-100
Serial.print("Raw: ");
Serial.print(sensorValue);
Serial.print("\tMapped 0-100: ");
Serial.println(mappedValue);
delay(20);
}
My mistake was that I had connected a resistor with a value that was too high, so the sensor values could not be detected — the resistor "absorbed" most of the signal. Earlier I wrote about the values, but I only fully understood these differences after going through all these steps. I changed the resistor to 68ohm and continued working.
Still, nothing happened, so I began to think that the problem was not the resistor and that my connections might be wrong. However, I couldn’t find anything, and even after asking others for help, nobody said the wiring was incorrect. After a lot of work with Rudolf Igityan, we finally understood that I simply needed to adjust the threshold values until they correctly matched the real changes of my sensor readings.
Because my handmade stretch sensor does not produce a strong signal, my working values turned out to be around 62. In the normal state the value was about 64, sometimes 63, but when I stretched it, the value dropped to 62–61. I changed this parameter and tested again.
The parameter is changed here:
// Trigger melody only on mappedValue == 4
if (mappedValue <= 62) {
playMelody();
delay(300); // short delay to prevent retrigger
}
<= 62 defines the threshold: when the sensor reading becomes less than or equal to 62, the buzzer starts playing.
With this adjustment, the code finally worked the way I wanted: it starts to play when I stretch the sensor and stops when I release it.
Final Code:¶
Lilit, [12/5/2025 6:40 PM]
We mast chang in the mapValue part
Lilit, [12/5/2025 7:11 PM]
#include "pitches.h"
// ---- Pins ----
const int sensorPin = A9; // Stretch sensor
const int buzzerPin = A11; // Use digital PWM pin
// ---- Variables ----
int sensorValue = 0;
int mappedValue = 0;
// ---- Melody ----
int melody[] = {
NOTE_G4, NOTE_C4, NOTE_DS4, NOTE_F4, NOTE_G4, NOTE_C4, NOTE_DS4, NOTE_F4,
NOTE_G4, NOTE_C4, NOTE_DS4, NOTE_F4, NOTE_G4, NOTE_C4, NOTE_DS4, NOTE_F4
// you can shorten the melody for testing first
};
int durations[] = {
8,8,16,16,8,8,16,16,
8,8,16,16,8,8,16,16
};
// -------------------------
void setup() {
Serial.begin(9600);
while (!Serial);
pinMode(buzzerPin, OUTPUT);
}
// -------------------------
void loop() {
sensorValue = analogRead(sensorPin);
mappedValue = map(sensorValue, 0, 1023, 0, 100);
// mappedValue = constrain(mappedValue, 0, 20);
Serial.print("Raw: ");
Serial.print(sensorValue);
Serial.print(" Mapped: ");
Serial.println(mappedValue);
// Trigger melody only on mappedValue == 4
if (mappedValue <= 62) {
playMelody();
delay(300); // short delay to prevent retrigger
}
delay(20);
}
// -------------------------
void playMelody() {
int size = sizeof(durations)/sizeof(int);
for(int i=0; i<size; i++){
int noteDuration = 1000 / durations[i];
tone(buzzerPin, melody[i], noteDuration);
delay(noteDuration * 1.2);
noTone(buzzerPin);
}
}
After I replaced the alligator clips with jumper threads and tested again, everything worked fine.
It would not be comfortable to wear the device in this form, so I sewed a small fabric pouch for it. The idea was to place it inside the clothing, hide the structure from others, and make the process appear a bit more magical.
Now it was time to style the hair, attach the stretch sensor, and test the entire construction in action.
How it went:
When the stretch sensor was woven into the hair, it stretched slightly, and the value of 62 stopped working correctly. I changed the threshold to ≥ 61, and after that it started working properly.
The final code looked like this:
#include "pitches.h"
// ---- Pins ----
const int sensorPin = A9; // Stretch sensor
const int buzzerPin = A11; // Use digital PWM pin
// ---- Variables ----
int sensorValue = 0;
int mappedValue = 0;
// ---- Melody ----
int melody[] = {
NOTE_G4, NOTE_C4, NOTE_DS4, NOTE_F4, NOTE_G4, NOTE_C4, NOTE_DS4, NOTE_F4,
NOTE_G4, NOTE_C4, NOTE_DS4, NOTE_F4, NOTE_G4, NOTE_C4, NOTE_DS4, NOTE_F4
// you can shorten the melody for testing first
};
int durations[] = {
8,8,16,16,8,8,16,16,
8,8,16,16,8,8,16,16
};
// -------------------------
void setup() {
Serial.begin(9600);
while (!Serial);
pinMode(buzzerPin, OUTPUT);
}
// -------------------------
void loop() {
sensorValue = analogRead(sensorPin);
mappedValue = map(sensorValue, 0, 1023, 0, 100);
// mappedValue = constrain(mappedValue, 0, 20);
Serial.print("Raw: ");
Serial.print(sensorValue);
Serial.print(" Mapped: ");
Serial.println(mappedValue);
// Trigger melody only on mappedValue == 4
if (mappedValue <= 61) {
playMelody();
delay(300); // short delay to prevent retrigger
}
delay(20);
}
// -------------------------
void playMelody() {
int size = sizeof(durations)/sizeof(int);
for(int i=0; i<size; i++){
int noteDuration = 1000 / durations[i];
tone(buzzerPin, melody[i], noteDuration);
delay(noteDuration * 1.2);
noTone(buzzerPin);
}
}
Final Result¶
After that, I placed the construction under the collar of my sweater and recorded the final video.
How I see this moment:
What it looks like from three angles:
HOW TO MAKE CONDUCTIVE DOUGH¶
After all these processes, I became interested in making conductive dough. It seemed really exciting, and I had some free time for experimenting, so I decided to try it. Conductive dough is a soft, clay-like material that can conduct electricity, which makes it a great tool for testing simple circuits, learning electronics, and creating playful interactive objects.
What ingredients we need:
- Flour we need 1 cup
- Salt - 1/4 cup
- Water ~1/2 cup
- Vegetable oil – 1 tbsp
- Cream of tartar – 1 tsp
- Powdered graphite or fine metal powder – 2–3 tbsp (this is for better conductivity)
Importont notes :
- Graphite gives better conductivity than salt.
- Fine metal powders. They can also be used with aluminum or copper, but graphite is a safer and easier variant.
The dough making process step by step:
What can you do with conductive dough?
You can build small experimental circuits, connect LEDs, buzzers, and sensors, and visually see how electricity flows through different shapes and forms.
