About my blog

The aim of the ITD course (ID4220) at the Delft University of Technology is to provide Design For Interaction Master students with in-depth theoretical and practical interaction design knowledge to help develop future products based on user-product social interaction. ITD proceeds through a sequence of iterations focusing on various aspects of the brief and the design, and culminates in an experiential prototype.

This blog is managed by Walter A. Aprile: please write if you have questions.



De meningen ge-uit door medewerkers en studenten van de TU Delft en de commentaren die zijn gegeven reflecteren niet perse de mening(en) van de TU Delft. De TU Delft is dan ook niet verantwoordelijk voor de inhoud van hetgeen op de TU Delft weblogs zichtbaar is. Wel vindt de TU Delft het belangrijk - en ook waarde toevoegend - dat medewerkers en studenten op deze, door de TU Delft gefaciliteerde, omgeving hun mening kunnen geven.

Posts in category Interactive Exercise Machine

Exercise1, our final concept: FLEXCURE

At the end of the project we presented our concept on the exhibition day. Also specialists that were involved earlier in the project, were enthousiastic about our concept ‘fitting the trend of using more interactive and therapy supporting healthcare products. See our latest video showing our final concept and prototype "flexcure":

YouTube Preview Image 

The Final Exhibition of Exercise 2

Yesterday, we showed off our final deliverables by turning a studio at Delft University of Technology’s Industrial Design Engineering Faculty into a contextual environment for the Magic Stick. When visitors first arrived at the studio they could pick up a small tri-fold pocket sized brochure of our product.






 Next, they could enjoy the living room environment, which was built to  illustrate the patient could use the Magic Stick anywhere.  While in the living room visitors could sit on the couch and watch our video about the Magic Stick as well as watch the video of Exercise 1.

YouTube Preview Image

 Final Video of the Magic Stick 

From there, they could move onto the dining room to enjoy candy. We also
created a sports therapy office with a waiting room. Here visitors
could test out the Magic Stick with a "therapist". 

Our contextual Studio

During the exhibit, we received lot of good feedback from visitors and one of the therapists we initially tested our maraca concept with visited us with her child. They both seemed very enthusiastic about the Magic Stick. She said, it would be a nice addition to her hand and wrist therapy tools.  This reaffirmed that the 6 months of work we put into building the Magic Stick could actually turn into a real success for hand and wrist rehabilitation patients. 



Finalizing prototype; sewing and laser cutting

The exciting time where things come together, we are finalizing our prototype! The electronics part with arduino, bending sensor and display with status bar is fully functional. This week we have therefore been working on the hardware of the prototype. This included sewing for embedding the batttery in the self made wristband and lasercutting for a small and precise case for arduino and display. Look at the details of this case; it mentiones "ITD 2011" on top and "power" on a side for the connection of the battery. There is also space to connect to laptop and to reset when necessary. See our pictures:

 Sewing the cover of our wristband:



 Stitching velco on to make the prototype fit on every wrist:



 Embedding the battery in a little pocket of the wristband:

The battery sticking out a bit for the picture, but will be fully hidden in the band: 
Laer cutting in progress:
The seperate parts of the case for arduino and display. Notice the lasered texts on the surface: 









The making of…

Last Friday
the morning started with fine-tuning the prototype, making photographs of the
prototype for the poster and brochure and setting up the paper. After that is
was time to get ready for shooting the movie. The storyboard was discussed and
checked for the last time, a list of needed materials was made and the locations
for shooting were determined. We went to several places to collect all the
materials needed and started shooting in Studio Home. The next location was the
therapy room of Marijke Vaske for the scenes at the therapist. It was very nice
that we may use her therapy room! The environment fitted perfect and the lighting
was great. In the late afternoon we shot the last scenes in studio home and who
thinks that playing a role in a movie is a lazy job, is wrong… it was hard
working and muscular pain from playing hand model, but we have had a lot of




Next week,
the 29th of June, during the exposition the end results will be
shown. For now we are busy editing the movie, collecting all the materials for
the exhibition, writing the paper and finalizing the poster, brochure and



Preparations for the exhibition!

We are already preparing the exhibition day. Exercise 1 and 2 will be together in one studio. We will decorate one half of the studio as a doctor’s office and the other half as a living room. Why? Well, because both our concepts are related to hand therapy but are also designed for home use!
We made a list of all needed things, so we can divide everythings.
The couch is already there!

Exercise 2: In the final throws of Magic Stick, Part 1

The Magic Stick
Time is running out; in about two weeks the ITD exhibition takes place. That means that
we have to speed up a little bit with all the stuff that still needs to be done. Last Friday we therefore
divided the tasks efficient: Barry and Dorine worked on the poster, brochure and set-up for the
report, Fawn made the scenario for the movie and searched for locations to shoot, Ruud worked on
the prototype and Eline made the 3D computer model of the prototype.

To give you all a quick update of the progress we made on the different tasks heres a short piece
about every task:

The poster and the brochure are almost finished. After a very useful brainstorm we concluded that “the magic stick” fits our product the best. The first impression of our product provides is a rather sober impression. But once in use you hear and feel the magic. And eventually it preforms ‘miracles’.

On the poster we tried to explain how our product works, by whom it will be operated and how it will be used. The brochure gives a bit more explanation. We are also working on our final report. The structure has been created, now it’s just a matter of adding it all together.


We created a rough storyboard of the video, taking inspiration from various movies of a U.S. site called Kickstarter. There were several elements that we were not exactly sure how to make, so we went to a video clinic where we were able to get a quick tutorial in After Effects. The effect we are going for is going to be challenging and time consuming, but we will try our best to make it happen. The storyboard has all the necessary commentary and rough illustrations of the scenes that are accompanied by explanations of the shoot. We also created a shot list to make sure we get all shots necessary. We will begin shooting on Friday and hopefully we will get confirmation on our preferred locations. 
Partial view of Storyboard  
We made good progress with the final prototype: after getting the supplies to the faculty we sanded
the parts to prepare it for the paint job. After a few layers of primer and a couple more series of
sanding, the final colors were applied onto the stick & caps: black for the body and lime green & red
for the weights. To secure the weights onto the body of the device, we chose to add threaded ends
for a screw-on mechanism. This will only be worked out as concept, just as the USB connectivity for
music (a dedicated USB connector was installed on the inside of the device body). Over the weekend
we worked on a custom neoprene grip, which makes the device comfortable to hold during the

 Painted End Caps with Final Color 
 Sanded and Painted Weight Tube 

3D Computer model
For the presentation we need to present the prototoype, in a model, as though it is going to market. We modeled all the parts in Solidworks, but
the USB connector was quite hard to model, thus we found this component model on the internet.
After all the parts were finished an assembly was made. In the picture below you can see all the different
parts in an exploded view. We now have a model to illustrate the components of our product. 

Magic Stick Rendering, exploded view 

Finalizing prototype

The product idea is getting serious developed shapes now. We bought a sweat band, used in sports, to embed our small Arduino and battery. The display is the only that a patient will see, on the surface of the band. This is an arm with the sweatband:

Arm with sweatband


Furthermore we use a bending sensor taped at the top (healthy) part of the vinger. Not directly taped, since the finger has to bend, but the idea is to stick little tubes at the phalanges of the finget. The bending sensor runs through these tubes, free to bend. Also, an additional idea is to place touch sensors on the vinger tips, to measure, in the final stage of the treatment, whether patients are able to bend the vingers totally till vingertips touch handpalm. Since patients are not able to do this ‘total flexion’, it has no use to make use of the bendingsensor to determine whether a patient bends totally. but above all, this would imply some form of callibration of the bending sensor, with a more compicated interface that does not match with our concept of easy and everywhere to use.

First, we will focuss on the bending sensor, but we’re almost there. Ex(er)citing isn’t? See the picture of the hardware, with the working display :). The second picture shows the sweat band partly cut open, to be able to squeeze the Arduino and battery in.  

Sweatband cut open for assembly

Developing our new product idea


After ourmeeting with a field expert, a therapist, where we showed and explained our prototype we had some though time, changing our product idea. The concept was ok, but the physical object we had in mind, a glove, turned out to be not suitable for use in therapy.

Thu; discussions and talks and ordering stuff! The idea has changed to taping a bendingsensor on the top part of the finger connected to an Arduino, instead of the original glove. This does not obstruct the movement of the exercise that patients have to do and it still fits our concept of easy and everywhere to use.

See our picture of our programming hardware. Yes, there is a display! What it should do? See the sketch. It gives feedback on the amount of exercises done and the amount of series done that day.


About our prototypes

Since the last few posts have been descriptions on a more conceptual level, we would like to enlighten you with a more technical story. Read about the ins and outs of our prototype!

First iteration

The first Iteration was done in a single afternoon. We were introduced to the Phi get I/O board and suitable software applications like MAX/MSP. After some struggling installing, we managed to get our first iteration working. However, it wasn’t stand-alone; we used the PC to control the lights ourselves. In the upcoming iterations, this would change.

Second iteration

The second iteration we used a touch-sensor and the Phidget I/O board connected to a laptop. On the laptop, we ran a Flash-application, which was capable to read the data from the touch-sensor. The Flash-application counted the amount of touches and this was translated in the size of the balloon on the computer-screen. In addition, we added a reset button when the balloon exploded, so the exercise could start over again. This iteration was semi-stand-alone. Since it was still connected with the PC, but no user intervention was required anymore.

The touch sensor of the second iteration  

Third iteration

For the third iteration we decided to go stand-alone, so no user intervention is required anymore. We used first an Arduino board that is programmable and can operate stand-alone, using a 9V battery. We connected eight LEDs to the Arduino so the progress of the exercise could be shown. However, we had some difficulties with sensing the bending-sensor, in order to sense the exercise. Therefore, we had to simulate the exercise using a linear potential meter. In addition, we also though ahead in order to make the system more mobile and compact we played around using the Seeedruino Film; A small flexible Arduino-compatible interface which is very small. However, we had some difficulties programming the Seedruino and we already noticed this solution was not solid enough to complete the end of the course. 

We sewed the bending sensor ourselves 

Fourth iteration

For the fourth iteration, we decided to change the LEDs into a small-display, capable of notifying the user more concrete and precise. In addition, this solution enables the user to get more insight in the exercise and can give the user direct feedback while performing the exercise. Since it was still a prototype, we used the Arduino Uno, a 2×16 LCD display and a bending sensor to show the status of the exercise: “too fast”, “OK” or “too slow”. In addition using a counter, we could notify the user when the required 10 excersies were performed. Moreover, we investigated the use of the ordered bending-sensor in textile. It was working though, but since there was too much bias while sensing we decided to go back to a more reliable method; a bending sensor. We made a list of stuff we had to order for the first iteration, in order to make it more compact and professional looking. This included a small OLED-display, a bending sensor and the Arduino min Pro.

Fifth iteration

For the fifth iteration we went to miniature scale; using a small 2 cm wide OLED-display, Arduino Mini Pro, a 3,7V battery and a bending sensor this design becomes small enough to be placed in a watch. Libraries for the display were created so the Arduino could communicate with the display. At time of writing this blog post we were porting the previous code to make it work for the new sensor and display. We’re looking forward to test the final design in practice when it’s finished.

The little OLED display

Exercise 2: Manipulating the Speed of Sound

 If you keep drawing out ideas and talking about how things will look like, it is hard to find out how it the product will turn out. By building, you are able to see which parts of your design need to be improved. And this was the case for our team. We found the purchase of last weeks materials such as the weights, did not work as planned. Thus, we had to change the building plan slightly. During the building it was harder than expected to integrate the weights. As we had previously shown, we wanted to attach the weights on the stick through magnets. But since the weights are 1 kilo each, we needed very strong magnets. Therefore we decided not to use the magnets in our prototype, but make use of PVC pieces with weights in it, which can be slide over the main PCV part quite easily. We have yet to determine the final look of the slide on weights. 

We made a trip to GAMMA, an all-purpose store, to find additional equipment we needed for our prototype. 



 New weight and pvc cap

 Meanwhile, we also made changes to the hardware inside. Aside from last week’s discovery of substituting the Phidgets for the Wii-mote to gain wireless capability, we discovered our prototype is now more sensitive to movements and is able to measure more precisely, the patient’s rhythm.  This was a bonus, as it will allow the patient to get proper feedback when performing their exercise at the right or wrong pace. 


Interior of pvc pipe with Wii-mote components
Another plus of using the Wii-remote are the existing buttons and perhaps there is something we can do with light. This is something we will discuss next week. 
Next to the hardware, we also worked on programming our prototype. In order to manipulate the speed of the sound while moving the wrist-rehab-stick, we use Max Msp and send the speed variable to the sfplay object. In our earlier prototypes we used a direct manipulation where movement measurements were directly output in the speed of the song. This resulted in fast changes of audio speed that did not exactly match the users movement. To solve this problem we used the mean and the bucket object in Max Msp to create an average speed sound using the previous speed variables. In our case we are using 5 previous samples. The visualization will illustrate this. The red line is a direct manipulation and the green line is the new averaged out way using the mean object to process the speed of the song.
Illustration of the average speed manipulation using the mean object in max msp
We also implemented new feedback in the form of spoken sentences. Until last week the patient experienced vibration and fast or slow music when he/she did the movement too fast or too slow. Now the patient will hear the voice of a therapist telling him/her if the movement was too fast or too slow, when he/she finishes a song correctly they may go to the next song and attach another weight if the patient finishes the song with too many mistakes they have to start over again. Below you will find the feedback for the various movements during the use of the prototype. The below movie also illustrates some of the audio feedback.   
Interactions and associated feedback
YouTube Preview Image 
Video illustrating the movement and sound interaction  
© 2011 TU Delft