Cam the RamBot
This research was developed as a feasibility examination. Currently, toys have their electronic components within inaccessible, sealed units.
Various literature shows a positive correlation between early introduction of computer science and children entering science, technology, engineering, and mathematics (STEM) as adults.
As such, I decided to test the feasibility of embedding electronics that would be accessible for children, so that they can create, program their toy and play with their design.
For this I decided to use Dr. Leah Buechley and SparkFun Electronics‘ LilyPad series. This set focuses around the LilyPad Arduino microcontroller board. This was chosen because:
- Designed for textiles.
- Literature demonstrates effectiveness in outreach to girls using LilyPad Arduinos.
- Washable
- No soldering required (use of conductive thread)
- Commercially available sensors and outputs with the LilyPad series.
As this was an embedded system, the hardware and software must be balanced. There is very little usable Random Access Memory (RAM) which limits the size of programs. In addition, it relied on a AAA battery for power, so efficient inputs and outputs were needed as well as proper threading of components.
Yet, to test an embedded system for use with children, a variety of sensors and outputs are necessary: children expect to have lots of variety in their toys. I decided to use the following in my design:
- light sensor
- multicolor light-emitting diode (LED)
- buzzer (noise generator)
- button switch
- accelerometer
I intended to have a switch between the light and light-sensor mode and a buzzer-accelerometer mode. This would conserve power. The toy needed to be designed for having the hardware inside it: a soft, concave form rather than a flat plane.
I tested my code outside of Cam, prior to inserting it. That was quite simple and successful, so I decided to begin work with the actual toy. Unlike testing, working within the toy resulted in several difficulties:
- LilyPad buzzer short circuiting.
- With such a diverse and heavy load, the battery heats quickly. This becomes a potential fire hazard with the plush materials.
- Limited space within the toy, allowing the conductive thread to cross and signals to mix.
- Sensor data returned in voltage, limiting its readability.
- Sewing in the head cavity involved limited visibility.
- Thread becoming lodged in the boards’ holes, preventing additional wiring.
- Thread shredding and pieces becoming loose.
So, with a somewhat unsuccessful endeavor, I have been considering future work:
- Testing alternative microcontrollers with sensors, including others in the Arduino line.
- Using alternatives to conductive thread to connect the system together.
- Integration of speech generation software to allow more variety in sound.
- Use of a rangefinder to allow visually impaired children more meaningful inputs.
- Extending the toy’s capabilities using a personal area network (PAN) connection between either additional hardware or the Internet and Web 2.0 services.
- Finding a less heat generating power.
