Rosie – Segway Style Concept
Rosie started out as a concept in the summer of 2003. It was loosely based on the idea of creating a more radical and more fun version of a Segway.It was designed as a single wheel portable device that could be ridden like a BMX bike but without the skill.
Rosie was affectionately named after Rosie in the Jetsons cartoon.
Device Goal: The goal of the device was to be light weight, fun, portable and in-conspicuous. The last part is important because people are afraid to buy something that is too futuristic. The device needed to look somewhat familiar, like an elegant bicycle, to gain acceptance. We wanted to specifically avoid a visual association with the Segway and instead create an association with a bicycle. (Hence the spokes)
The conceptual drawing was made from a picture of a unicycle. The idea was to create a small motor that would either be part of the handle-bar or alternatively an in-wheel motor.
Research
We researched the concept and found that there was no existing device that did exactly what we where trying to build but we did find some interesting devices:
We found a motorized unicycle that as far as we could tell is balanced by the driver manually adjusting the throttle. We found one other reference to a motorized unicycle. Both of these have been gas powered and could only go forward. The driver would make small adjustments to the throttle and his weight to keep the device balanced. This is similar to a bicycle or motorcycle doing a wheelie.
We also found a fairly old invention called a mono-cycle or mono-wheel. This is where the driver is on the inside of the wheel and is balanced by gravity pulling him down to the bottom of the wheel. Many monowheels exist and there are enthusiasts today who still build them with large motorcycle engines. Their large size generally makes them somewhat unwieldy and the wheel generally blocks the field of vision.
Around 2003 Bombardier came out with a concept design that came the closest with a single wheel futuristic motorcycle design. This concept however was never actually built to function.
Again our device was supposed to be a bicycle-style device not a motorcycle.
Technical Research
The problem of balancing a device was actually solved long before the Segway. The problem is generally known as an “inverted pendulum” problem. I refer to this as the broom balancing on my finger problem. Explained simply remember that Force equals Mass times Acceleration. You calculate the force of a falling body on a pivot point (in this case your finger), use gravity as your acceleration, and then apply the opposite force to your finger to keep the broom balanced. If you are able to compute how fast the broom is actually falling you can make adjustments to keep perfectly balanced. This technique is used in helicopters, walking robots and even skyscrapers to minimize the effects of wind.
There are two devices that we choose to use to measure our rate of falling, an accelerometer and a gyroscope. A gyroscope will tell you how much your angle of rotation changes over time. An accelerometer tells you which way is down. Together the two can act like a compass that always points straight down.
Building a Model
We needed to build a model to learn how to use our gyroscopes and control boards effectively. Our model used Lego as a simple building block material. Controlling the apparatus was a PIC board that is essentially a small chip that can be programmed from a computer.
Onboard our model had a gyroscope that sent signals to our PIC board. As the device started falling in one direction the gyroscope would send signals to the processor and the motors would reverse to keep the device balancing.
Using the model we where able to get the device to balance for several seconds. Because the model did not have an accelerometer after a few seconds the device would accumulate gyroscope drift. This drift is caused by slight inaccuracies in the gyroscope angel causing the device to fall over.
Building a real prototype
The actual Rosie prototype was built to be low cost. We tried to use as many off the shelf parts as possible and had a goal of creating the initial prototype for under $1000 in total parts. We constructed the bulk of the prototype but never complete the final design.
Current status
We put Rosie on the back-burner in 2004. It turned out that Segway has a patent on a similar idea. We also felt that if Segway couldn’t build a thriving business from a two-wheel product it would be difficult to do the same on a much smaller budget.
Hopefully this post sparks some ideas and helps others as they pursue their own ideas. Any companies interested in more details on the Rosie concept can contact Raizlabs for additional technical documents and related information.
Special thanks to Matt Malchano and Blake who designed and programmed the initial prototype.
Links:
