Prepared by Daria Klimova and Altynai Nogoibaeva

Project description: 

Have you ever been in the situation where you are apart from a loved one and you really miss them? Be it a friend, a family member or even a partner. Flair takes care of this problem! This question has been extensively researched and resulting data has been mapped out to mimic this human interaction. Social support can broadly be defined as the perception of meaningful relationships that serve as a psychological resource during tough times. Hugs as an expression of compassion and emotional support are crucial aspect of interpersonal interaction. One of the most important findings indicates that hugs have a “stress-buffering effect”.On this basis Flair was developed to bridge the expanse between people. It is an outerwear that allows to send hugs through a smartphone. Flair is made of inflatable actuators that recreates the sensation of touch by inducing emotions perceived during the hug which are sent by a distant loved one. Flair connects via bluetooth and transmits the message to the actuators, thus embracing the wearer.
A Hug a Day Keeps the Doctor Away!


  • Sealing:
    • CNC lasercut – Raylogic 11G
    • Iron
    • Soldering iron
  • Inflating:
    • Arduino 101
    • 12v water pumps
    • 2 relays
  • Polyethylene -> buy here
  • AutoCad
  • Laserworks
  • Rhinoceros 3D
  • Adobe Illustrator
  • PreForm
  • Blynk


A hugging wear? How did we come up with this idea? Simply. Our team decided to explore the inflatable actuators. At the begging of our research we made a decision of exploring the inflatable structures. We started with a design of the stationary device that had a power to amend the physical space through hugging and by allowing people to be comforted or “gathered” together. Once we assembled out first prototype ‘Flair 1.0’, we observed that the material that we chose to prototype with was not effective at generation enough force to change the shape of the whole structure at the human scale. After several iterations of solving this difficulty by using various thickness polyethylene film. The result of the scaled model was still hardly achievable; however, we did not want to give up the idea of embracing and comforting people. We made a decision of moving from stationary block to wearable. That is how, the idea of a hugging shirt was born. Why send hugs? We understand how it feels to miss someone who is thousand miles away. Our motivation is to bridge the expanse between people.


To analyse the “physics” of hugs, we asked volunteer to hug each other. We recorded a total number of 26 hugs where 8 hugs were collected for male-hugging-female, 6 for female-hugging-male, 5 for female-hugging-female and 7 for male-hugging-male cases . The result on contact area summary you can see in the figures below.  


We commenced the project with the exploration of seaming techniques. The first method that we used is ‘a conventional’ heating with iron or soldering iron and baking paper (Figure 1). Such method of sealing was not suitable due to the fact that there were many difficulties with the temperature control, thickness of the seam and efficiency and effectiveness of manual sealing. To add, due to its inefficiency, the manual sealing was used only for broken-seam repairing. After early stage experiments with soldering tools and iron, we sought for more efficient ways of seaming. We assumed that lab’s CNC lasercut could be used for more than cutting and reckon that if we find the right setting we can hack lasercut for polyethylene sealing. We set different configurations – changing speed, power, table position (amending Z-coordinates). The first ‘optimal’ setting we found resulted in thin and quite weak seam. Figure 2 shows different settings seam result. Further search for ‘optimal settings’ end being in better seam – three time thicker and stronger and better air tightness (Figure 3).


Step 1 – Simple patterns
The exploration of ‘pattern behavior’ we started with designing simple shapes. Firstly, we made pattern of horizontal lines with varying intervals for each model. This experiment resulted in different inflation ‘behavior’ which you can see in Figure 1 – 4. Besides, we studied the inflation of the patterned/repetitive ‘prints’ as shown in Figure 5, 6.
Step 2 – Folding shapes
The next iteration of our research was focused on exploring the patterns that enable ‘self-folding’ behavior. We discovered several techniques that allows us to turn 2D flat models ‘cutout’ into three-dimensional shapes just by inflating, such you can see in Figure 10 – an example of a self-folding structure. What realised is that, the model/structure folds better if ‘guiding’ layer is attached (for such layer a stiffer material should be used). As an example, Figure 8 demonstrates the ‘behavior’ of the structure with double layer of polyethylene. In cases with combined material, the inflated layer works as an actuator which pushes the other layer either up or down depending on the direction it faces. The same principle was used in Figure 9, where cardboard was used as a second layer. Functionality –  reversible folding and unfolding of the structures.
Step 3 – Human scale shape
Having studied the ‘behavior’ of folding structures at the small scale, the next iteration is to increase the scale closer to the initial idea. Why closer and not full scale? The available lasercut working area is constrained to 1200х900mm. In addition, the material that we worked with did not have enough pneumatic potential force and was quite fragile when was tested on the ground (tiny, sharp particles that tore or jab the plastic). Figure 16 demonstrates the first close-to-planned scale working prototype. It was automated and reacted to human when approached. It was enabled with a distance proximity sensor, pump and microcontroller and worked as self-folding structure (Figure 9).
Step 4 – Wearable
As we realised the drawbacks of the stationary self-folding structure, we decide to explore a potential of the wearables. Prior to the design of the ‘wear’ we studied the ‘physics’ of hugs and mapped out the areas of the hug that we want to mimic.  Figure 17 shows the first wearable prototype, testing session of which pointed to the lack of significant effect on the pressure points due to its  primitive pattern design. Figure 18, 19 show the exploration of design pattern for other body zones such as neck and arms. With further research and continuous iterations and simplification, we finally arrived to our MVP that mimics the hug accurately.


To enable remote activation of the pumps, we used BLE enabled microcontroller. To be able to use BLE we used Blynk platform that enables to control microcontrollers with smartphone. In the mobile app, you can create your own graphic interface, such a digital dashboard, virtual buttons, gauges and on. In our case, we used the basic Blynk ‘blink’ code that is available in the library examples and uploaded it to the Genuino 101 board. Prior to enable BLE function, microcontroller has to be flashed. It is done via Arduino IDE. In the mobile application, we created project where we placed two buttons for each digital pin. Those buttons we set for D7 and D8 and the output of the buttons switched to High or Low state depending on the state of the button within the mobile app. To each of these pins (D7 and D8) on the board we connected relay’s signal wire, 5v and GND.  


Testing of the ‘Flair’ was done with 5 people. Besides feeling embraced, several participants of the testing session said that they actually felt empowered, as ‘I feel like I turned into hunk and I am ready to smash things around! Awesome!’. Below are the testing session summary.

Trackback from your site.

address: 20 Myasnitskaya ulitsa
(metro stations ‘Lubyanka’ and ‘Kitay-Gorod’)
Moscow 101000 Russia

phone: +7(495)772-95-90 *15026

email: city@hse.ru