On Friday, May 30th the prototype has been tested for the first time in practice. I did this in the first kindergarten classroom of a local school called ‘Eltenbos’ in Lommel. Many curious faces noticed immediately that something unusual has taken place in their classroom. This first test I did, I started with 1 toddler a time as testperson. I gave them a free hand after started the applications, I showed them some existing applications and gave them the possibility for themselves that they could decide what to do with it. With this I could conclude whether these young infants had any idea what was expected of them. Some even dared nothing to do, fear from doing something wrong to do so. It was not clear for them (as long as nobody else tipped these toddlers, how to manage this table) how to deal with this new kind of interactionmethod. Even if a piano filled the screen, it was not clear that 1 simple touch could ensure a bright note. After making clear that a simple fingertouch was enough to act, they got a full attention to work with various other applications.
In this next stadium I’m analyzing the captured video that was made out of 2 different angles. One in front of the toddler (to see his expressions) and one on top of the table (to see their handlings).

All applications are working properly. The tabletop knows perfectly where touches are made and sends a signal back to the correct place. Tomorrow, the tabletop will be tested for the first time in practice. The pre-schoolkids will get in touch with this new factor in their daily classroom-environment. I’ve made various existing applications ready for use, they can discover themselves. Meanwhile, the tabletop is colored in black and the white frame on top is finished. There’s also a new piece in front of the tabletop that creates an extra firming and ensures protection of the mirror.

Ready for the big day tomorrow …

The box is ready and everything can be tested together at a fixed location. Unfortunately, the drop of my beamer is not large enough to cover the whole acrylic plate to provide a full screen image. No disaster because this is a prototype, I will fix this later. I adjust the camera on the image that has been dropped by the beamer. In Touchlib I make sure he sees only the dropped image. I rule the filters and see my ‘blob’ as very bright. The calibration does its job very well and the first touches are a fact.

It is important to have a solid box that’s toddler-proof. It should be a hard enclosure that sets and keeps all the hardware in a safe way, on the desired location. Inexpensive still firmly pressed wood is an interesting solution. The box consists of 4 separate parts which can quickly put together and take apart. On this way it becomes mobile and compact during transport. On the surface there is a hole where the acrylic plate and ledframe will come. On the other side of the surface, there is room for a laptop to set up for observations that can be done. A handy and useful work surface. Under this surface is a second floor where the mirror with a beamer takes place. The camera is set on the beamer where it captures the screen trough the mirror.

The next step is the same setup but with a beamer which projects the signal through a mirror on the acrylic. Among the acrylic I’ve pasted a piece of chalkpaper. Especially important now is to know whether the light of the beamer is going to be held by the IR-filter in front of the camera and whether the infrared light will still be visible through the chalkpaper. It works perfect and I can start building my box that holds all the elements and protects all the hardware.

The moment where I can put everything together is reached. For now I use 2 trestles that must support my acrylic. The frames with LEDs are mounted on the sides (provisionally with adhesive tape) and the camera is underneath the plate. With Touchlib I play with the options (mainly through the use of filters), making a clear as possible picture of my touches on the plate.

In this first test I notice that direct sunlight or other ambient light of the sun is a bad factor because it’s the largest source of infrared light. From now on there will only be tested in a dark room with minimal ambient light. The camera gives a clear picture of the touches (after the outbreak of the IR filter in front of the lens!).
Below a picture without the filter on my camera.

For the wiring of my leds I use a wire of 0.5 mm with isolation. I place in every hole a led and hang 8 leds at a time to each other by connecting every positive pole with the negative of the next one. After every 8 leds I hang a 1-ohm-resistor to this row and a wire that I connect to a ‘breadbord. This is a small connection plate which connects the incoming wires very simply. This step, I repeat 5 times on each side and in total 10 times to create a fully connected frame with leds. I test the fresh wired circuit with positive results.

Before I start wiring the leds I first finish the frame where the leds will take place in. Because I am using the most powerful IR LED who can handle a relatively large luminous, I decide to put on every long side 40 IR LEDs. Which means a total of 80 leds. So I drilled on every half a centimeter a hole with a diameter of 5mm (the actual diameter of my leds). Because my plastic frame is deeper than the thickness of my acrylic, I clear a few millimeters of this strip. After this, the hole can be drilled right in the middle on the edges of the plate, and gives a perfect position to be able to reflect in the glass. Thereafter, the frame has to be sawn at an angle of 45 ° sawn to be connectable to each other to create a complete framework.

Despite my electronics background I still keep in touch with my local electronic shop and the worldwide web so my leds certainly will not be overloaded with too much power. I used the ledwizard. This is a small online application that perfectly tells you how to wire your leds and how much resistors from the kind you need. As power source, I use an old PC power supply with an AT connector. This normally gives a direct output to the motherboard of the computer and gives a DC voltage of 12volt. I should put my leds in 10 different rows of 8 LEDs in a parallel circuit. Each row should contain a resistor of 1 ohms so my leds cannot be overloaded. After testing the I concluded that an infrared led has a total different characteristic than a normal led. The anode and cathode of an IR have led have reverse characteristics compared with an ordinary led. Therefore, I connect my source in reversed order. Before I connect everything together I first test 1 row with 8 LEDs and 1 resistor . The leds are shining. Now I am sure that the 7 other rows will also work and I can get started with the wiring of 80 leds.