This post is a journal for the module “Embodied Interaction” at Zurich University of Arts in the fifth semester of the Interaction Design course.

17.09. Monday – Kick-Off

Like most modules, our docents began with a short kick-off for the whole module where we were introduced to the overall topic and received all the necessary information for the following five weeks.

Mini Exercise

To begin, our teacher Joëlle introduced a small challenge for the next 24 hours. The assignment was to track data or traces of our body in an analog way. This should help us to get a little into the topic and make some more thoughts about it.

I found my topic really quickly because I was highly fascinated by skin conductance, more precisely Electrodermal activity, short EDA sensors.

Actually, it’s quite a simple thing. If we get stressed or excited (very generally), we tend to sweat on our fingers. This can be easily measured by the change of conductivity on the skin caused by the sweat. The fascinating thing is, that this can change within seconds so correlations can be easily made. On the other hand, this also makes it hard to interpret since our body is so sensitive and it is not an absolute value. At least it’s very hard to control by the body and often times it’s driven by emotions. Also, it is technically relatively easy to measure.

This sensor is – in combination with other sensors – used in misleadingly called lie detectors. But since these sensors only show discomfort or stress, it’s not really possible to completely identify a lie and so it’s controversial if these machines should be called like that.

To measure my sweaty fingers, I decided to draw a spot with a soft pencil in my notebook, put my thumb and my index finger on it and smear my finger down to the page. With the gradient of the track, I was able to see how wet my fingers were. Bike and Notebook

I repeated the procedure after sitting on the atelier terrace, after the bike ride, while cooking, while eating spicy food and after doing dishes. Instead of doing it on a fixed timescale, I wanted to examine some certain events.

The actual thing that interested me was the effect from my body on my fingers but after washing the dishes (and drying my hand) my fingers obviously were very dry. This is more an effect from outside directly on my fingers and not through my body and brain. But I also continue to observe these things. Bike and Notebook

17.09. Tuesday – Introduction in the topic

I continued with my data tracking after I woke up and when I drove to school, back home because I forgot my notebook and back to school. I’ve been really sweaty, not only on my fingers.

Back in school, we presented our work an received a little bit of feedback by our teacher Joëlle.

Methodology, case studies

Also, Joëlle made a very interesting presentation about methodologies and case studies. The methodology included datagraphy, personal data that could be gathered (body data, physiological data, live stats, network-based data, and admin data).


Also, she introduced us to a taxonomy of digitally fabricated data outputs. I think this helps us a lot to not get completely lost in this field which still if fairly new to us.

  • Data Sculptures
  • Form Chasers
  • Fixture as Structures (metaphor for expressions i.e. sofa -> brain waves)
  • Performers (Interactive, music from answers)
  • Mavericks (machines that are different i.e. satellite path on a card)
  • Body as interface


Following ways to translate data into a representation:

  1. XYZ translation (into 3D space)
  2. Metaphor as data
  3. Existing grammar
  4. (unreadable on my notes)
  5. Purpose-driven

18.09. Wednesday – Bodystorming and Inspiration

Single Performance

This morning started with a small assignment that was given to us on the previous day. The idea was to make a short performance of approximately three minutes to display the analog data we collected. It was very funny and still interesting to watch and sometimes we couldn’t hold it back to laugh very hard.

I decided to represent the data with sound. So, I took the OP-1 pocket synthesizer (which helps me in all situations) with me, chose a synthesizer and pitch, and mapped a single note on a built-in step sequencer to make sure the note is always played with the same length. I translated the gradient I had in my notebook from putting my finger on the pencil spot from the previous in the decay of the note I played. Therefore I went back into synthesizer mode and mapped the decay of the sound to the gradient I perceived and named the event I was recording my sweat on the fingers, such as while cooking, after cycling and so on. It maybe wasn’t a “productive” exercise but still, it helped me to gently get into the topic.

I think the fascination about data is that it could give us accurate, unbiased and objective insides when we do not even have to think about it. These advantages are usually not possible in an analog way. But especially because these measurements are biased, it helps (in my opinion) much more to be aware of what we are measuring and the overall situation. It doesn’t tempt us to focus too much on every detail and this is especially helpful in the beginning.

Performance about another person

After our personal performance, we were asked to form groups to observe the traces/data a group partner leaves and make a little performance afterward again.

Daniel and I decided to do that together and spend lunchtime to observe each other. We found a few quirks of each other but we were not really satisfied with this idea. After a while, we met a girl which was having her second day at Cast (audiovisual media) who introduced herself very friendly but also in a weird way. We noticed how hard and awkward these situations can be so we decided to make a performance as if we met each other the first time. Included were awkward questions if it’s the right room, what we are studying and a weird handshake/clap/hug; something in between.

Research Verena Ziegler

Verena Ziegler introduced us to her research which changed from architecture to “interactive” fabrics, all with new digitally fabricated technologies.

Along with that, she has shown us tons of great inspiration in the field of embodied fabrication.


  • “Walking City”, Archigram, 1960
  • “Body Extension”, Lydia Clark, 60s
  • “Cool Himmelblau”, 1971
  • “Finger Glover” / “Einhorn”, Rebecca Horn, 1970
  • “Mobile Office”, Hane Holler, 1969
  • Lucy McRae
  • Iris van Herpen
  • “Shrink”, Lawrence Malstrat
  • James Merry (Björn)
  • Nero Oxman

Data output repetition

After our brains were completely blown away by the projects that have been done our docent Joëlle made some repetition. It included:

  • Metaphor
  • Data mapping
    • XYZ Mapping
  • Existing grammar
  • Machine Properties
  • Purpose driven design
  • Data driven design.

This should prepare us for the following brainstorm session.


Since I’m not such a fan of brainstorms I tried to make something different. I knew that we were making a speed dating brainstorm and from my previous experience, I knew that sometimes it isn’t working with every person. So instead of pressuring me, I accepted that sometimes I had no idea but I asked questions to my brainstorm partner to challenge them for some ideas.

After all, we found some pretty interesting ideas which we all collected in front of the class and briefly discussed.


The goal of the day was to build groups and find a direction of the topic we want to focus on. During the brainstorming sessions, Daniel Holler and I had a great flow so we decided to go further on our ideas. After a (literal) ping pong session, we agreed to focus on sweat and do some general research what sweat can mean in society.

20.09. Friday – Working on the topic

This is a post written by my co-worker Daniel E. Holler on his Blog.

The project is slowly gaining some shape and color. On Wednesday, the big performance day, we had the chance to enjoy some theatrical, didactical and funny performance presentations from our classmates. After having had the first contact with personal data and how we can creatively express it, we had to dig deeper in the afternoon and closely observe some ticks or little habits of strangers, and again perform it in front of the class. This lead to a more concise finding of ideas, in the form of a creative speed dating.

The Idea

After the speed dating my idea slowly crystallized about the same subject as Vinzenz’s: We want to work with sweat. We were intrigued by its simplicity, its raw nature but especially by its very variegated purpose for humans and animals.

This morning after having made a quick research session about the subject, our ideas started flowing and flowing. From highlighting sweat and making it a (fashion) statement to make a portable sweat diary, to have some sort of second skin where sweat could be represented.

The Trouble

Currently, we are having a little trouble to connect the dots: the measurable data, the product, and the purpose. We also dove into some metaphors which are related to sweat. Quickly we found the strigil, a sort of scraping tool used by ancient Greeks, in order to cleanse the skin before and after doing sport.

A typical strigil

The significance of sweating in the ancient cultures we found something really interesting: Sweating was regarded by many cultures spanning the whole globe as cleansing. Sweating cleansed the body psychologically and physically in a sort of ritual. This sacral importance of sweating completely lost its meaning in nowaday’s society, where sweating is regarded as something to be frowned upon.

The Concept

The current state of our very raw concept would connect a simple monitoring of when and how much a body sweats, transcribe the data into a digitally fabricated object (preferably sturdy and durable materials) and for example, wear the solidified sweat (as we all know sweat probably the most ephemeral secretion of the body). The solidified sweat would hide bodyparts but at the same time, the “undesired” sweat is made visible. Furthermore, we would like to connect these aspects of our concept with the metaphor of rituals, but we will start to experiment first with the sweat sensor in order to guide our concept into the right direction.

20.09. Friday – Rhino & Grasshopper Introduction

For 3D editing, we had a day of introduction for the software Rhino with the visual coding extension Grasshopper. This should enable us to process data into parametric design models for this course.

Drawing, Primitives, Boolean Operations, and Loft

Since we already had an introduction in Rhino last year, the overview in the software was a little short. We right away jumped into the curve drawing tool which we used to draw an arrow. A handy tool is the search bar where all commands just can be typed and the command bar when the drawing tool is selected. This way you can select the starting point first (comma separated for x and y coordinates) and after pressing enter the actual length of the stroke you want to draw. The drawn curve we than extruded. To have a better view on it we changed the preview method to a version which was previewing the model in 3D instead of just lines by a right click on the camera perspective name.

Extrued Arrow Curve

Later on, we just played around with primitives such as cubes, cylinders, spheres, text, and pipes which was a curve combined with a circle curve.

Rhino Primitives

Boolean operations were actually very straightforward as we know it from other graphics software. But there is a huge catalog with different operators which are still to explore.

Rhino Boolean

In the end, we combined two simple curves into a simple loft. This loft can be rendered with straight sections so that it could for example cut onto a flat surface which then again can be assembled into a 3D structure.

Rhino Loft


Then finally we started with Grasshopper to generate some cool stuff. First of all, we had an intro to the overall interface but it’s really messy and hard to learn. The nice thing is, that a double-click on the Grasshopper canvas opens up a search bar similar to the one in Rhino where elements to place can be found.

First confusing thing is that there are two different ways to draw primitives. You can either draw in Rhino and connect it to an object by right-clicking on the Grasshopper element and selecting to connect an object. These Grasshopper objects have a hexagon shape around them. Second, is to generate an object in Grasshopper and this is just done by adding the same kind of element but without the hexagon surrounding it.

Grasshopper Primitives

Generate and Divide Loft from Three Points

Later in Rhino we drew three points and connected them with Rhino (select multiple points) draw a line between them which they were generated into a plane which we divided into two numbers (chosen by a number slider) into parts (with “Divide Domain^2”) and then created a box out of them with “Isotrim” and “Surface Box” where the height was controlled by a slider again.

Grasshopper Divided Loft

Going on with the previous work, we draw a so-called “Bounding Box” which contained a sphere which was subtracting it from the box. This “Bounding Box” then was added into every section we previously divided from the loft. I did not really get how we made this but I guess that’s fine.

Grasshopper Divided Loft

Pipe Curves

The next one was much easier to grasp. We drew two curves, divided them into points and connected these points with a curve. A pipe then let these lines look much cooler.

Grasshopper Pipe Curve

These pipes then can be shifted by a certain factor.

Grasshopper Shifted Pipe Curve

Working With a Face

To finish all up, we scanned a classmates face with the software “Skanect” and then continued to modify it with Grasshopper.

The scan was built up with a polygon structure which is widely used by many 3D tools. Each polygon has three points which we separated using list item and an index for each list item. The function “InCircle” finds the inner center of three points where we generated an area inside and extruded it to a certain amount and closed by the function “Cap Holes”. Then we scaled X, Y and Z coordinates and we had a cool kind of parametric work on a scanned human body part.

Grasshopper Face

26.09. Wednesday – 3D Printing Intro

For the whole day, an introduction to 3D printing and serval software tools was planed. Therefore, the whole class met in the 3D printer corner of the low-level workshop in ZHdK with Roman Jurt, Co-Founder of the Fablab Zurich and lecturer at industrial design.

Class in the lab

Material Archive 3D Beacon

First of all, we were introduced to the new printers Roman assembled and put onto nice racks. The first exercise was to print new beacons for the material archive at ZHdK by turning the model that should be printed in the preparation tool “UP Studio” to get a sense how to achieve the best result of printing and to avoid to much support structure on the print.

I tried to very extreme and placed the object beacon on a tip of the edge expecting it to fall down. But actually it was printed without a lot of trouble, only some edges were not perfect and the extruded letters were not very clean. In the end, almost no support structure was needed.

Almost no support structure

Inspirational Projects

Roman continued with a presentation of projects he made and tools he used. This was very inspirational and I think also important to get a little the way he approaches these projects with much more experience than we do.


Also, he introduced us to serval tools with his own touch to approach them. First, he gave us a little overview of what are the differences and what’s available. Major differences are knobs based and polygon based tools. Knobs are like vectors we know from Illustrator. Very simplified, you can draw curves and process the space or surface in between those curves. More comparable to pixels are polygons which are used in other tools. It basically works with three connected points in the three-dimensional space which provide a surface which can be connected with more polygons to a bigger and more variable, big surface. There is also kind of “3D Pixels” which we know from Minecraft and are mainly used in VR tools (which can also be used for modeling) but are rather niche products.


Tinkercad is maybe the simples 3D tool available and super straightforward. Since we played around with it in previous courses it wasn’t a big deal anymore for us. It was a refresher and Roman could introduce very cool new features in our workflow. So, it’s possible to draw an object duplicate it (CMD + D), move, rotate (or edit it in another way) and duplicate it again with the shortcut. This way, all the edits will be applied to the duplication in a further step which is great to make patterns very quickly.


We already heard many things about Onshape from our Industrial Design colleagues but never worked with it. It’s a professional online CAD to work collaboratively, in real time and with a version control just in the browser. It’s an amazing piece of software btw. But of course the learning curve takes much, much effort than Tinkercad but as Roman says still easier than other CAD. And indeed, after getting the basics it’s pretty straightforward. At the end of the afternoon, we had a group challenge to build a space rocket with the whole class working at the same time which miserably failed because no one had an idea how to deal with the software. So Roman took over and used it as an intro to Onshape.


It’s hard to get for what Meshmixer is actually useful for. But Roman showed us some tools to make a beautiful mesh out of a model, how to close holes and to analyze the object. Also, it has a way to build support pillars inside of the model which would be hard with the auto-creating support structure of G-Code software. Sometimes it’s better to have full control over it.

Personal Project

Most of the afternoon, we had time to work on our project in doing some experiments.

Material Feeling

Daniel experimented in Tinkercad to produce a different kind of structures to give the material a feeling for excitement or uncomfortableness to find a balance in the material and examine when it starts to change.

Bottle top Texture

Grasshopper Experiments

I continued to work with Grasshopper to learn more about how it works. Therefore I watched some YouTube videos and tried to apply some things I learned in the previous course. My goal was to find out how to apply data later on in the 3D models.

Screenshot Grasshopper Experiment 1 Experiment 2

28.09. Wednesday – Refining Concept I

In the previous day, we made experiments to scare some of our friends and measured their finger conductivity on their fingers. We found out that after the short shock, they became much more relaxed which was also supported by some research we found. To move on by we decided to finally focus on work context to force people to not be stressed at work. By scaring them we can first force them to stop buying too much into the current rhythm and help them to relax.

Wearable Reference

We first looked up some references to for wearables to get a little of aesthetic view for it and built something like a mood board. Also, we looked up for similar projects in the field and why they were made.

Then we also found an interesting article for a device that was able to recognize if a real (natural) or a synthetic voice was speaking. Very cool and interesting with the emerging of machine learning which enables always better synthetic voices and could lead to serious problems in communication when people are not used to it and trust voices. But the very interesting thing was how the user was warned of the synthetic voice. It was a small Peltier element that was cooling down as soon a synthetic voice was detected. So we wrote that down for the next task to do.

How to Scare

There are many ways to scare people and mostly they are very well interconnected with our senses so we react before we actually can think about it. This actually is essentially getting scared.

But most ideas such as flashing light and loud noises were too distracting or not as unexpected such as something that pressures you. So, after short brainstorming and discussion, we decided to test the cooling with all the properties that are also working in an office environment.


So we went into the thrift shop “Brockiland” to buy super cheap bags with straps to attach something on our body. In the meantime, we put some water in the freezer to use it after. In the thrift shop, we found a cool backpack with an aluminum frame on which we could strap our bag easily and enabled us to test all sorts of attachments on our body with ease.

On the way back to school, we passed by our favorite electronic store Pusterla to ask for a Peltier element. We were not surprised that they had multiple elements but unfortunately they were too big with 45 mm on each side. This would take a lot of flexibility from us in terms of the size, because we need to build an even bigger cooling system around as well as with the power supply but also how we attache it. It was also not necessary since we don’t want to cool down a small fridge. We also assume that smaller elements would cool down faster.

With the ice, we tried to scare each other and see if the reaction was the same as just scaring people before which was successful. Happy Dani, scared me


From then on, Dani began to sketch some ideas of an advanced prototype so we can better communicate with each other but also with mentors to move on quicker.

Support From the Lab

In the meantime, I wrote a mail to Clemens Winkler, one of our lecturers from the IAD Lab. We knew that he would have some experience with it and could help us along.

03.10. Wednesday – Mentoring, and Sickness


In the morning we had a mentoring. Because Daniel had to work during the night, we switched up with some groups and were the last group for the mentoring. We met some hours before and created a short summary of our journey. This is always a good chance to structure your concept and reflect like writing this blog.

We enthusiastically presented our work to our mentors and the feedback was quite good. They liked our concept but highlighted the following points to be careful with:

  • Is it necessary to compare with other (worse) health gadgets?
    • They would not recommend it.
  • There are many really bad and useless health gadgets but we should not overgeneralize this. There are also gadgets which have a lot of respectable science is behind and really help people without giving wrong promises.
  • Not collecting data is great but we should also highlight why this is important i.e. that insurance companies are really looking for this.
  • We said we don’t want to be our device be taken seriously. That’s a bit an easy cheat to also not work seriously. According to our intentions, we should rather name it playful.
  • We criticized the promise as a magic cure which some devices make. This is a bit harsh, generalized and sometimes even some “magic things” work. We shouldn’t think black and white.

Sick Dani

To not risk too much, Dani went home to get some rest because he really felt sick.


In the meantime, I continued to work with the Peltier element. It was very straightforward. Connected to the power source it cooled down on one side and heated in another. Then I did some research on hardware that is missed and we need to get from the Lab.


Also, I prepared for the presentation we had in the class. Therefore I just continued to nail down the presentation we made for our mentors and made it more beautiful. I changed the things that were criticized and ask Dani to check it out at home.

Then I prepared our prototype for the presentation by attaching our prototype element to the backpack frame vertically and horizontally with straps. In the Trends & Identity (former Style & Design) atelier, I froze some water in forms that were exactly fitting for the prototype element.

I also felt a little dizzy, which was why I decided to go home a bit earlier to prevent to get sick.

27.09 – 04.10 – The big concept issue

This is a post written by my co-worker Daniel E. Holler on his Blog.

In the past week we made a very big progress concerning our concept and our idea. We had to find a common ground and idea so we decided to work on stress. Stress makes perfect sense, since measuring sweat on our hands is very closely connected to emotions. Furthermore we wanted to investigate negative emotions, but more of the chronic kind, so we could leave out fear and anxiety (even though they also play a role)

In our mentoring, we still were undecided in which direction the subject of stress should lead us, therefore we stayed on the simple path and formulated our “hypothesis”: How can we reduce work related stress at the work?

Well, good question — and at the beginning difficult to answer. What Vinzenz and I did, was not to think about it too much and to dive into the data. We had our GSR sensor running and finally we could test it out, get a feeling of the data and hopefully it would lead us into a good direction.

Initial Testing Grounds

We knew for a fact that sweating on handpalms is triggered by strong emotions. We quickly set up the sensor and suddenly the Scary Maze Game came to our mind; this game, an early 2000s online flash game is very easy to solve: your cursor should not touch the black walls and the player has to navigate through a sort of maze/track. On the last level the difficulty is majorly increased, leading to a very concentrated state of the player, shortly before reaching the goal, a very high pitched screaching voice and a scary picture surprises the player promptly sacring her/him. We tested this game on five people, 3 men and 2 women. The men already knew the game and both of the women did not. The results were the ones we expected, the players got very scared. Afterwards we reviewed our data and plotted it and we were very surprised: In all our tests the sweat (or stress) shortly shot up during the scare, but a second later when our testperson realized it was just a small jumpscare, the stresslevel decreased drammatically and stayed below the inital level.

Scare Plot 1 2 plots of our tested women, the scare and the subsequent reduction of the stress level are clearly visible.

Vinzenz and I were very interested and quickly began researching more and more about it. I researched the psychology of jump scares, which is widely used in horror movies. In an interview with a neuroscientist about horror movies this answer striked me:

“Once the ‘fight-or-flight’ signals cease, the brain releases neurotransmitters and hormones that mediate what we call the ‘rest-and-digest’ system, the heart rate is coming down, the breathing is slowing, goose bumps are relaxing. There’s a sense of internal cognitive relief in the body, and that feels good.”

Further research for references

This is the addicting factor of horror movies for example, but we figured that we could exploit it for our stress reduction device. Before we did further testing we did some deeper research concerning health gadgets and existing stress reduction gagdgets. Our findings were pretty extensive but we could extrapolate some common grounds:

— Stress reduction gadgets work mostly by calming the user. Breathing pattern analysis and breathing techniques, similarly to yoga exercises are very widespread. Also some sort of bio feedback systems are used where the user has to concentrate on their own body fuctions/emotions. Moreover some gadgets such as the Moodwings bracelet (very appropriate name) work with simple vibrations. The vibrations sort of calm the user through a soothing touch.

Moodwings The Moodwings vibrating bracelet, in the shape of a butterfly

— A second extrapolation is that medical gadgets are very controversial. There are several studies showing that medical gadgets base their effectiveness on non scientific data and heavy marketing. Concerns over false promises, abuse of personal data, reliability and validity came to our attention. We thought we could very well built our product, criticizing/counteracting these problematic aspects that many (but of course not all) medical gadgets feature.

The Final Concept Plus Further Testing

Therefore by connecting the dots of our findings in the testings and in our research, we came to the conclusion that our device should scare people in order to reduce their stress levels at work. I did some sketches with the horror movie jump scares techniques in mind: scaring people with quick motions and flashes of light work best. My initial idea was to have the backrest of a chair spring open, letting the user nearly fall off his/her chair. A further idea would have been to have a sort of airbag popping up and scaring the user with its motion. Unfortunately, both of these ideas would be very disruptive during work. Vinzenz finally had the idea to scare the person by shocking the body with cold, for example when people flinch when they are touched by an ice cube.

Scare Plot 2 In the 2 red circles you see the stress level going down, after the Ice Cube Test.

In order to quickly validate this direction we made a very quick and dirty test: We froze some water and sat infront of our laptop shirtless and with our GSR sensor active. After some minutes of being distracted we pressed an icecube onto the kidney region (which is vey sensitive). The results were clear, the stress level shot down again (even though the plot does not look as exagerated as with the jump scare).

Ice Cube Test The Ice Cube Test

Therefore we have three major elements which make up our concept, the data, the user, and the embodiment. These three elements are incorporated into the main goal of reducing stress levels at work.


In the aspect of data we want to make sure that our user’s data is not shared, since our device will only reactiveley work with it, without saving the data. This should be an alert to the current trend of, for example insurance companies giving better offers to people who share their medical device data with them.


In our research I mentioned the fact that many health gadgets work with dubious claims assuring people benefits which cannot be scientifically proven. Our device should also call to attention that there should be a little bit more of self responsibility when it come to the promises of medical devices. People should know that not every promise is true made by the marketing team of these devices. Moreover our research showed that the stress level do go down with a good scare, we feel that medical devices could (and should!) be more playful, instead of being sort of esoteric. Not always is the most expected solution the right solution.


We like a lot the idea of working with primal instincts such as fear in order to counteract the modern man made sickness of work related stress. The psychology of fear is very well researched in academic contexts but also in the entertainment/artistic contexts such as horror movies, giving us a very good springboard. And once again hacking the body in this playful way could make our device actually helpful.

04.10. Thursday – Presentation, Sickness, and Tech


Dani still wasn’t getting better. Actually worse. So, we decided to have him just for the presentation and then leave him home. The presentation was very similar to the previous one we had with our mentors and the feedback was mostly good. Unfortunately, there was not a lot to add but we again realized that we need to work more on the embodiment.


As planned, Dani left immediately after the presentation to go home and recover.

I contacted Luke from the Interaction Design Lab for some technical support. He said that he only had some minutes but that was enough for me.

In the afternoon, we met in the lab. He gave me a new Peltier element with the heat sink already glued on one side. Additionally, a battery with 1.5 volt which was close enough to the prescribed 1.7 volts, a transistor, a breadboard, and some jumper cable for the Arduino. The goal was that I was able to control the temperature of the Peltier element by the Arduino. At this point huge thanks to Luke for the time and patience although he had many other things to do and also more important things going on in life.

The Arduino setup was quite simple. Between the ground going out from the Peltier element I put the Transistor and connected the plus pole to the plus pole of the Peltier element. Through the PWM output of an Arduino pin, I connected the Arduino to the element which then could break the PWM-Signal)really quick to change the “resistance”. Actually, the PWD signal and also the transistor turns on and off the connection so quick that it’s almost like a variable resistor. New thing for me.

One problem I had was that I thought the connection to control the “rhythm” of the on and off of the resistor was in the middle. This is called “Base”, the “input” is called “Collector” and the kind of “Output” “Emitter”. So according to the data sheet this was not true. The base was left (number 1 / B), the collector in the middle (number 2 / C) and the emitter on the right (number 3 / E). Big shout out to Michael Schönenberger for the help. Without him, I would have gone crazy and I learned a lot even how to measure electricity with a measuring tool. Transistor Bipolar Transistor

So technically, we are now able to control the temperature by Arduino although we only need an on and off switch. First setup

09.10. Tuesday – Building Fairly Advanced Prototype

With the fresh week, we already had a mentoring coming where we should present a fairly advanced prototype which should be in “real size” and going into the final direction we want to build the final piece.

Rethinking Peltier

Last week we managed to overcome our sickness but also to make a Peltier element run on a simple Arduino setup.

But when we tested it, we were not quite satisfied. The element was getting somehow cool, but as soon we touched it, it was getting our body temperature. Obviously, the surface of the Peltier element was conducting heat/cold very well and so it was not very effective on something that is warm like our skin. So there wasn’t really a shocking moment of cold.

On a (or few more) Friday evening beer the week before, our class colleague and tireless competition (joke) Aurelian said that it might be simpler to work with an ice spray. I kept the idea growing during the weekend and proposed it to Dani. After thinking for some minutes we bought a nasty smelling, ordinary sports cooling spray.


The first thing we did with the cooling spray was again testing. We tested it on us and as well on a female testing subject with the GSR sensor we already used a lot. As well here the result was clear. Happy Spray Testers

Prototype Ideas

So what to do with the spray? Simply spraying it on the back was too simple, not embodied and also kind of dangerous because the skin could be burned. So one of the simplest ideas was to get some plastic tubes in the workshop. In the tube, we stamped some holes and sprayed some spray through it. Close to the body, the result was literally pretty cool because all the pipe felt very intensively cold. The problem was, that it needed to be attached tightly to the body which was not easy with the plastic pipe. We stuck a lot of body hair in between which shows our dedication. Hair outch

Since Verena Ziegler showed us her 3D prints on textile, it seemed as everyone wanted to try it as well and there were samples of it lying everywhere in the atelier. So, we also had to try it. It almost felt so natural to go this way because all around was full with it.


So we tried to stretchable textile and mesh textile. The flexible ones were not really able to let the 3D prints stick on the textile but the ones with a rough mesh were holding more strongly. Still, it was not really satisfying and we didn’t want to create a skin-tight piece of clothing. Attaching flexible textile

Kinesio Tape

So we very quickly had the idea of Kinesio tape. It could be attached directly to the skin like a wearable. The pipes could be directly printed on the tape and formed to the body shape. This was our thoughts but now to practice.

So we spent way too much time on creating a 3D model (first with Rhino, then with Cinema4D) and printed it on the Kinesio tape which was attached elaborately on the printing surface with Duct Tape. 3D Model for Kinesio Print

Finding Shape

On the first trial, we recognized that more surface of the object on the textile would not give us necessary more grip on the textile but more surface to be ripped off while having the textile flexible. Kinesio Shape Sketch

Second thing was, that the shape or curve of the object was giving different properties on bending it. Sounds very obvious but is still very important.

Refining model

Now we tried to have a shape that is curved so it should be more flexible to bend.

It was very time consuming and needed serval trials to print. We had a lot of problems with the model, the wall thickness, witch the Kinesio tape and with uncalibrated or damaged 3D printers. One of hundreds fails

That was very exhausting but we made a small piece before we were completely falling apart. Done and done

16.10. Tuesday – Production

Only two days were left until the end of the module of our project. Since we had to make a presentation with a performance we wanted to finish the production as soon as possible to dedicate the following day fully to the presentation.

The goal was to have all models and to print them all on Kinesio tape.

We had the technical constraint that we only had 3D printers that were able to print a fairly small surface although the Kinesio tape would offer us to print long and thin structures on it. But because this wasn’t possible, we chose to make multiple modular parts that could be interconnected with small plastic tubes.

Therefore, we redrew all the so-called splines for the actual path on the tape and the cross-section of the pipe in Adobe Illustrator. As done in the previous days, these splines can then easily imported into Cinema 4D. We agreed to make three shapes with different bending properties to represent the possibilities.

An adjustment from the last experiments was that we decided to make the pipe flatter. This would give the print on less stiffness which then is better for the overall flexibility. This is very crucial because the tape is attached directly on the body and is exposed to all movements. Subsequently, we had to change the end and beginnings to connect the plastic tubes from the elements to each other and connected from the cooling spray to the first pipe element a bit higher end wider.

In theory, it would have been very easy to model the connector and attach it to the ends of the pipe we created before. But in reality, it was not. Modeling it was a big pain because we used the Cinema 4D “Sweep Tool” to connect two different cross sections (from the pipe to the tube height) to an expanding pipe. To connect all the parts correctly took us a lot of time but was fairly simple. We finally managed it thanks to the help of Alessandro Holler, the brother of Daniel and the assistant of Scientific Visualization at ZHdK. But that was not enough of trouble. Connecting the two parts — the pipe and the connector — was not really working. Alessandro was trying to help with his little time he had but obviously also had other things to do. After trying to unsuccessfully connect the polygons of both parts manually, all of a sudden it worked just with the regular boolean tool which connects two objects together but never worked at the first trial in this day.

Without questioning too much, we created the two elements and ran to the printers. But again we had trouble. First of all, the slicing software UP Studio rendered a raft and support structure of the print although we removed it in the settings. Trying for a long time and adjusting shapes did not help. But again, magically we figured out that if we turn the model by 90 degrees it worked. There’s no actual reason for it but it works.

But also while printing we encountered another problem: Some of the printers were not really calibrated well and just ripped off the Kinesio tape we attached on the printing surface. We did not even know that it was possible to calibrate the 3D printers but luckily, a guy from Industrial Design showed us how to.

Finally, after many hours, we were able to print almost six exemplars of two shapes each. We agreed to stop at this point because we knew that the next and final day will be exhausting and we should save some energy for the last run.

Nevertheless, I quickly modeled the last shape at home in the bed, hoping that the next day will be a bit more productive.

Failed again

17. – 18.10. Wednesday & Thursday – The Arrival at the Destination

This is a post written by my co-worker Daniel E. Holler on his Blog.


After starting production on Tuesday, we scaled it up on Wednesday. We had some to do some troubleshooting with the 3D printers, but in the end, we could easily finish the production having in total 18 PipeLinks (we decided on the last day to call our amalgamation of Kinesio tapes and the 3D printed tubes “Pipeline”). With 18 pieces we had six each of our three models. The whole production lasted until the evening; we would go down to the workshop every 1,5 hours in order to start a new print. Moreover, our last model, the double curve — which can be bent horizontally, whereas the other two models are flexible on the vertical axis — was finished on Wednesday.

Our last model during production: “PipeLink Double Curve” Performance Ideation In the afternoon we started to ideate a concept for our performance. Quickly we found a good concept and I started sketching a storyboard since our performance consists of a combination of video and us two performing. Afterward, we quickly gathered our equipment and shot the video in a dark room. More about the performance at the end of the text.

A big part of the video consists of an abstract data visualization representing the stress level, the scare, and the subsequent reduction of the stress level. The data we visualized came from an actual test we performed with Rebecca; you can find her in one of the earlier posts. We chose her data because she had a very consistent and well readable result. We imported the CSV file into Processing and Vinzenz wrote a little program, where the data points were translated into a circle. The circle was drawn on top of itself at every data point, resulting in the circle changing in size. Vinzenz rendered a TIF file for every data point/frame and imported it into the video.

“The storyboard is the instrument of the coward.” -Werner Herzog “The storyboard is the instrument of the coward.” -Werner Herzog

While Vinzenz was busy editing the video, I concentrated on a few stylistic details for the performance. Since we wanted to show our PipeLinks sticking on our backs; with all the tubes for the cooling spray etc; we had to come up with a new piece of garment. The reason being that we did not want to perform half naked. Our solution was to tailor two long sleeve shirts, made of transparent plastic. I received help from Lola, a Trends&Identity student, who had some sewing patterns for t-shirts laying around. I struggled a little bit with the details and finishing, but both of our shirts fit us perfectly.

Transparent T-Shirt The freshly tailored, long sleeve, transparent shirt.

In order to add some very last finishes to the performance and subsequent presentation, we quickly animated our models in Cinema 4D. After rendering them and stitching them together, we inserted the video in our Keynote presentation, as a background image during the Q&A after the performance.

The Performance

The performance consisted of a combination of a video and us two performing. The video we projected, showed our two faces in a split screen while “looking into the computer”. The sequence lasted around one minute. The rendered data visualization was overimposed on the video. We timed our facial expressions (becoming more and more stressed, in relation to the data) to the data plot, culminating in the “big scare”. We wrenched our faces, clearly showing the result of the freeze-shock, and afterward leading to the wave of stress-relieve. In order to support the video and to “map” it into reality, we sat in front of a table, looking into the computer, faking a “real-time” recording of our faces. Because we had our transparent shirts on, the PipeLinks were clearly visible. In order to accentuate the climax (the cooling shock), we had some actual cooling spray hidden from the audience; therefore by carefully timing our action, we sprayed the cooling spray, creating a little wave of vapor around our bodies. The performance ended with us standing up and bowing to the audience.

PipeLink attached PipeLink attached PipeLink attached The PipeLink stuck on our backs, they are even visible under some garments

PipeLink Presentation Vinzenz and Daniel during the presentation

The animated models The animated models