From UAFCS

Cool things that we could bring into the classroom:

Contents 1 Demonstrations/Kits 2 Networking 3 Simulation 4 Data Analysis 5 Competitions

Demonstrations/Kits

MIT SeaPerch type ROV designs:

• Made from cheap PVC parts
• Lets you teach mechanics: center of mass, buoyancy, thrust
• Tons of electronics: good progression from topside switches, bottom-side relays (driven by separate wires from topside), bottom-side microcontrollers (driven by one serial line), all the way to a networked array of autonomous robots.
• Applicable to science classrooms, technology, etc.

Robotics:

• Teaches sensors, microcontrollers, batteries, motor drivers/relays, motors, and mechanical design.
• Build a micro-mouse type mobile robot.
• Build a stationary robot. This can be as simple as two servos (elbow, gripper) with directly-connected mechanicals. Servos are ridiculously cheap (bidproduct.com) and almost plug-and-play. They take a simple PWM command structure that requires a microcontroller too.
• Networked robots, or arrays of robots, would be a clear generalization.

Alternative energy power generation:

• Pick a generator: solar, hydro, wind, etc
• Hook up to a big machine via a USB microcontroller (Custom board? Arduino?)
• Have kids *manually* play with pointing, maximum power point tracking, etc. I think the trick to making this fun is automated plotting.
• Log data out to a network connection for plotting and persistent storage.

Energy efficient building design:

• Start with a cheap white styrofoam (or cardboard?) box--that's the building shell.
• Add "doors" and "windows" with PVC pipe.
• Add a heater: UL-listed heater is probably too much, but a 60W light bulb would probably work better, especially with a dimmer.
• Add some sort of computer control: at least for thermal logging, but ideally also for a programmable thermostat.
• Add competition of some sort--most efficient frame design considering energy for both light and heat.

Road icing:

• Trade-offs: sliding off road vs de-icing chemicals.
• Sensors and data prediction is really cyber-physical.

Networking

Access experiments and such from the web: e.g., on our own server, on facebook. This can scale from:

• Just build web pages.
• Write HTML code (transitions to general programming mindset)
• Provide remote access to robots, experiments, kits (feels cyber-physical too!)

Experiments as listed above.

Simulation

Write code that kids can play with, to simulate real-world stuff.

It's amazing how simple the core code can be, and still simulate ice solidification, solid mechanics, etc.

Something like this:

http://www.cs.uaf.edu/~olawlor/projects/1998/demos/demo8/demo.html

Existing "simulators" tend not to have any actual code inside, just a scripted series of silly yes/no choices.

Data Analysis

Use some sort of programming (Excel, JavaScript?, Matlab?) to do basic data analysis.

Incremental feedback is extremely important to kids, and this is something that simulations can easily provide.

Competitions

A good way to make boring subjects more exciting!