Planning the final project.

After all these months of learning all the different techniques and such, there's a lot more concrete information available to think about the final project. My original idea (a toy robot) was incredibly vague.

After the first few programming classes, it quickly became clear to me that it was definitely not my strong point. The design aspects of other classes (inluding molding and casting) suited my interests and skills much more.

After the class on Input Devices, I was intrigued by the light sensor, and started thinking of how it would be fun if you'd make a device that chased after a dot of light, sort of like a cat. This idea stuck with me, and I started searching the internet for examples, hoping to incorporate it into my final project.

Soon I found a page in German (found here, where someone had made the exact thing that I wanted to make. Even better, source code was provided, along with the message that he loves to share it with the world so people can recreate and improve on his design.
Another positive point to this find is the fact that at first glance, most of the electronic components are the same or similar to the ones that we've been using in the lab.

What do I need to make this? The schematic provided on the page provides me with most components to make the board:

• 2 Light Dependent Resistors (photoresistor)
• 2 Small Signal SMD Transistors (labeled BC849 NPN)
• 2 Mini DC electric motors
• 1 Battery (the original design uses a LiPo 3.7V 140mAh)
• 3 Resistors (2 x 10k and 1x 10 Ohm)
• 1 Switch
• 1 ATTiny (45 should be fine)
• 1 LED

Most of these parts are available at the lab, others I've found can easily be ordered online for a few euros per component; the battery could be gotten for around 5 euros, and the photoresistors for about 1 euro (prices quoted from Conrad.nl)

The board will be designed in Eagle and then milled out on the modela, after which it'll be all soldered together.

One of the other pitfalls is that while the code is provided, it is provided in German. If the code doesn't work at once, and needs to be modified (which is likely) the code will need to be translated back to English. It might take some time to find all the correct commands for this, without breaking the code. With just a board and motors the little bug won't move very far though, so we need to make some more things.
Especially wheels. These wheels should be easy to design in a 3D modeling application, and then 3D printed. This will keep the wheels light, but still sturdy. The PLA material for 3D printing is readily available at the FabLab. Alternately, the wheels could be cast from a light resin, or possibly even a type of rubber. The key here is to keep the wheels light.

Time permitting, if all this works there might still be a chance to design an enclosure for the little bug. This would have to be designed and produced from a fairly light material, so the motors don't give out. I think the best would be some sort of acrylic, laser cut into shape. As it's already kind of buglike, I intend to design something like an insect-like shell. The LDR's could even become stalk-like eyes (either wrapped in tape or running through some small tubing).

Given that I'm not starting from scratch with this project, but building with/on somebody elses project, I should be able to finish this in the two weeks left for working on the final project.

In the first week, I want to get the board done, and all the parts assembled. If time permits, I will also start working on the 3D model of the wheels.
The second week will be dedicated to getting the code to work, and to designing the 2D aspect of the (snap-fit) enclosure.

If all goes well, by the end of the two weeks I hope to be able to show a working prototype of the lightbug.
In evaluating the project, I want to at least have the mechanics working. This means I want to be able to show that the board can be turned on and off, that the light sensors are able to get a reading of light, and pass the value on to the motors which will then drive into the direction of the light.
In essence, if I can prove the code works, the project is a success.