Crazy Robotics Idea - Assisted Steering Bicycle

K, so generally, I lurk these boards during the season and off to look for ideas on how to handle the games, rule clarification, and the like. But during competition, I had a zany idea for a off season robotics-esque build, and instead of tinkering around and thought experimenting for a week, then giving up on it, I decided to come to you guys, the FIRST community, for your tools and expertise to mooch off of mostly, but if you guys want to help, we could get something together.

So, you know how we built line sensors into our robots this year that worked on the principle of reflection? So, as a real world application to that, sidewalks in general tend to be more reflective then the dirt surrounding them. Fascinating, right? Not really, unless you wanted to make a bike that could tell the difference between sidewalk pavement and dirt, and could ride the edge of the sidewalk from one point to another, not leading the rider into the grass, or otherwise forcefully dismounting them from the vehicle.

So, the plan as it stands is this, equip a bike (or trike, less falling over and stuff) with these reflective sensors to allow it to tell the difference between drivable pavement and not quite as drivable free dirt, using laser pointers if necessary to amplify the effect. Equip it with an ultrasonic radar sensor as well, and you have a relatively safe bike that can navigate itself in the short-term. Now, in theory, this could be a fully robotic courier bike to haul pizzas and packages around from place to place, but I think it would be easier and more innovative to go with an assisted steering unit, in where you have feedback on the helmet or handlebars of the bike, so someone who was legally or entirely blind could use the bike, rather than being entirely self powered. Plus, I’m cheap and lazy, and putting a cRIO on a bike doesn’t scream “good idea” to me anyway.

Regarding input and output, we also want to have something that can take this information, and along with GPS, Compass, and Web access, plot out a good route to take, as well as being able to relay these instructions to a sight-disabled driver. I had a hard time with this one, but I’m thinking that, barring a feasibility check, we could use a PSOC board similar to one found in the Rookie '09 kit for DIO, and hook it up via mini-USB to an Android Phone. Using the Talking Map data, with compass calibration making sure that the rider is kept on track, the reflective sensors, and the ultrasonic in front, we might be able to have a bike that could safely transport a blind cycler from point A to B under their own power and guidance.

I’m kinda at the conceptual stage, so does this sound worth pursuing, or am I barking up a tree with no juicy squirrels in them?

So thanks, but not quite what I had in mind. I was imagining a human driver on a manual powered bike/trike, with sensors and GPS replacing a deprivation of sight, but if you wanna shoot big, we can shoot big, I guess.

UPDATE: With the release of the Android ADK for use with Arduino, this project got a whole lot more feasible for a lot less hacking and costing monies. So I’m probably going to go in that direction.

One more thing, what would any of you guys recommend for a force feedback output on handle bars that could hook into an Arduino board? I’m looking for a similar effect to a rumble pak, and rather not use an audio cue, as users would be listening for that, rather than cars and stuff. Any ideas?

Cell phone vibrator motors might be a good place to start.

Or any motor really, drill a hole in a chunk of matter, away from its center of mass, and press fit it onto the shaft. This will allow you to experiment with different motors and counter weights (and thus different frequencies and magnitudes of oscillation).