h a l f b a k e r yLeft for Bread
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Imagining a gyroscopically stabilised (Segway type) vehicle supported on a single captive spherical "wheel" much in the manner of an inverted roll-on anti-perspirant bottle.
The spherical wheel would be approximately 300mm dia and constructed of cast aluminum, with 10mm dia x 5mm deep cups cast in
over the entire surface of the sphere, to facilitate mounting (?epoxy) of several hundred high-flux-density rare earth magnetic discs. The sphere & magnets are then spray coated with a long wearing rubber surface which is baked on. The mass of the vehicle above the sphere is supported by 4 ball casters which ride in spring suspensions fixed to a frame structure above which contains batteries and control electronics. The 4 casters roll directly on the spherical wheel.
Drive to the spherical wheel is accomplished via magnetic induction. Induction drive coils would be placed at various points around the sphere in the structure above it. Gyroscopes via microprocessor then manage the currents to the coils, creating opposing forces which rotate the spherical wheel in the direction necessary to keep the vehicle in balance a'la Segway.
Power consumption requirements would be around double that of a Segway. Segway only has to manage balance in X axis while the ball-wheeled vehicle must manage balance simultaneously in X & Y axes.
Such a device would be well suited to moving around in environments generally suited to foot traffic.
Would be able to move accurately in very small increments in *any* direction as opposed to other wheeled vehicles employing casters or a coaxial pair of wheels e.g. Segway, all of which must make some motion to compensate for caster effect or inability to move sideways.
Has applications in domestic & industrial robotics and mobility aids for disabled people.
As a disabled mobility aid, this vehicle would be most useful for a mobility disabled person in a common kitchen, workshop or grocery store, where precise motion control is important to get the user within comfortable arm's reach of storage areas or work surfaces.
This vehicle could easily incorporate a means of raising or lowering the seat to increase the arm's reach range of the user to that approaching an able bodied person. This would allow disabled people unassisted access to common facilities like grocery shelves.
As a domestic robot, control electronics and sensors would make this vehicle ideal for semi-autonomous operation, perhaps being able to fetch articles for a disabled person via voice command.
Potential also to be a 'follow-me' home companion/health supervisor robot for ill people. Could summon medical help (potential for onboard video link for emergency services personnel) in case of heart attack, epileptic seizure, etc. so patients could more safely live at home, away from all the sick people in hospitals and aged/disabled care facilities.
A machine with such capabilities would be easily able to dock itself to a charging station somewhere out of the way when needed.
Attachments for a domestic robot could include a detachable vacuum cleaner which could be self-stored and self-attached at the charging dock.
In manufacturing, such a robot could be used for autonomous resupply of parts & materials to humans working at assembly workstations as well as conveyance of goods to various points in the process.
Asia Science Letter
www.tokyo.afosr.af.mil
Hope URL is correct for maglev wheels, way down at the bottom but the other stuff makes the hair on the back of my neck,uh. People are seeing things in experiments they have no names for. [mensmaximus, Dec 27 2004]
HB archives: "Orb-It"
Orb-it
Idea circa 2002. Very similar, includes an illustration and several links to prior art. [bristolz, Dec 27 2004]
Waltzing IV Stand
Waltzing_20IV_20Stand
You can tell I was still on meds when I wrote this. [Shz, Dec 28 2004]
[link]
|
| |
Hi, thanks for the description. It sounds as if it has enough equipment to stabilize a camera dolly. Therefore you are back to three wheels if you want to rent it out. There will be much video shot this century, you don't mention it as part of your bot. Forget about vacuuming the floor; I like to give autonomy to bots, it may spread to people. |
|
| |
As far as disabled use, I have stretched a wheelchair to twice its length, but it needs your wheel up front, not the two little wheels, to be empowering. Trouble is, in a house, only a regular wheelchair can manipulate the tight turns. I don't need a wheelchair but when and if I do, I want a stretched one/front power wheel and also one that can pull itself up the side of a ten story building. It should be that strong. |
|
| |
Brau, point well taken. That's the last of the fundamentals required to make the spherical induction motor work. Agreed that it will take some smart electronics to make this go. |
|
| |
Since the ballwheel may/must rotate in any direction,the magnets would all be mounted with either N or S end outward. |
|
| |
The induction drive coils, placed in regular locations around the upper half of the ballwheel, would be pulsed sequentially by the CPU to force the ball to rotate in the direction required to counteract forces acting to topple the unit or to effect motion in the desired direction. |
|
| |
Hall effect or optical sensing (e.g. optical mouse) could be used to sense the direction of rotation of the ballwheel for feedback to the CPU. |
|
| |
Thinking big, acting little, uh, there is a maglev design that has high powered magnets in the wheels to achieve levitation.
I'd like to see this wheel as a snap-on attachment to some sort of maglev public transit. |
|
| |
mens, great idea to incorporate the maglev into a wheel and not into the track. |
|
| |
The train still has to pull the wheel axle along with it meaning there is some rolling resistance and axle bearing friction, but since the main mass of the vehicle would be supported by magnetic levitation and the load borne directly by the wheel, those losses are minimised. |
|
| |
Maglev train design keeps magnetic radiation to passengers down to 10 gauss. |
|
| |
Brau, I think this can be accomplished with an industrial computer like one of these: http://tinyurl.com/2xqnx |
|
| |
Lotsa grunt for this app(800mHz - 1.0GHz), x86 architecture. Will run *UX, Winblows. |
|
| |
The motor controller will be pretty spesh. About 30 induction coils will need to be controlled with 5-20A peak currents. |
|
| |
The code to generate the motion will be most interesting. Handed this one off to a cder I know just for fun. |
|
| |
Optical ballwheel direction sensing can probably be stolen straight out of an optical mouse. There's certainly enough precision there as well as driver software & support. |
|
| |
Won't the ball get dirty? Also, you want to monitor the position of the magnets (which I'm assuming are in a regular pattern) at any given point in time, not the general motion of the ball, right? Otherwise, you'll have to have some way of calibrating. |
|
| |
Tiro, if the thing is used outdoors, the ballwheel will get dirty, but only as much as one's shoes. The machine would be mainly for hard surfaces. Possible to include some bristle brushes to clean any pebbles or grit off the ball before the surface has to move under the coils. |
|
| |
The coil gap will necessarily have to be small for efficiency, but there will have to be a reasonable thickness of rubber on the ball for a wear surface and some gap to allow small grit to pass by. Bit o' tradeoff there. |
|
| |
It's not so much the magnet positions which need to be tracked as much as the CPU will need positive feedback that the ball is being rotated and in the direction anticipated. This is necessary for climbing slight inclines or overcoming obstacles like carpet edges. |
|
| |
The CPU has to momentarily increase and decrease current to keep the rotational speed constant or the rider would get pitched on even small obstacles. |
|
| |
Optical mice are very good at recognising tiny or pronounced surface differences; the same technology could be used to recognise the rotational motion of the ball's rubber surface. |
|
| |
In general, the positions of the individual magnets isn't important as they are all aligned with the same pole outward. A grid of induction transformers would be pulsed such that the magnetic force opposition causes the ball tp rotate. Sequential activation of coils in particular patterns will be necessary to balance the machine. |
|
| |
It will take a bit of snazzy code to make it go! |
|
| |
Wouldn't the poles of the magnets be N S N S? and the coils changing polarity? When the coil is turned off there is reverse polarity as the field collapses. How is this energy utilized in this design? |
|
| |
As I think about it, if all of the S poles on the ball face outward, the induction coils will have to be oriented with their magnetic flux axes tangential to the ball surface. Half of the quantity of induction coils will need their magnetic flux axes oriented 90 degrees from the others. X & Y motion are thus controlled individually. |
|
| |
The N-S orientation of an electromagnet reverses when you reverse the current flow. The controller will thus have to create + & - oriented PWM signals for each coil to effect bidirectional motion of the ball wheel. |
|
| |
"Orb-it" is indeed very similar to what I have cooked here. |
|
| |
The drive system in Orb-it is significantly different, though. Orb-it applies drive to the ball wheel via wheels I am simply using for support in the magnetic induction proposition. |
|
| |
Seems to me that either will work but mine's a lot more complex. However, it obviates the potential problem of loss of balance in Orb-it due to drive wheel slippage due to contaminants on the ball wheel surface. |
|
| |
Also, direction of motion is more readily and quickly controllable in the magnetic drive with no need to reverse direction of a drive motor armature, leading to more reliable self-balance performance. |
|
| |
[it obviates the potential problem of loss of balance in Orb-it due to drive wheel slippage due to contaminants on the ball wheel surface.] If you have any loose surface both of these bots are face down because the designs depend on a rigid surface. Anything that is moving under your wheel such as a one inch round ball as you apply a trajectory force to it can't be comprehended by a computer. |
|
| |
One should be careful with absolute proclamations about the abilities of computing. They're a moving target. |
|
| |
But, computers do tend to become moore and moore powerful. So far, anyway. |
|
| |
Things would be a bit hectic if this got to 3000 RPM. It does all seem very complicated though. |
|
| |
3000 rpm for a 300mm dia wheel would be a speed of 169km/h (~105mph). You can bet things would be out of control. |
|
| |
To recap, this device is for indoor use on carpeted or hard floors. It need not move more than a few metres at a time and certainly no faster than 3km/h. |
|
| |
http://tlb.org/eunicycle.html |
|
| |
Too cool. 'Self balancing' electric unicycle. |
|
| |