| View previous topic :: View next topic |
| Author |
Message |
linuxguy
Joined: 16 Oct 2006
Posts: 120
Location: Beaverton, OR
|
 Posted: Mon Oct 16, 2006 1:25 pm Post subject: Hybrid Octapod |
 |
|
Hi,
I have an application that might be able to use the current sense capability of the Open Servo to good advantage. I have designed two hybrid robots - they have both wheels (4) and legs ( . The wheels are all independently driven, steerable, and fully retractable.
If I am not mistaken, I could use the current sensing to detect when a larger than normal load is put on a leg joint. This might help me tell when the robots need to switch from rolling to walking modes for locomotion.
Is this correct?
I have CAD models and pictures of these two hybrids if anyone is interested in seeing them.
8-Dale
_________________
No, Mr. Jobs, the BiPod is a ROBOT. It does not play music OR interface with iTunes.
The Dynaplex Network - Robotics, Open Source, Linux, and Technology Forums |
|
| Back to top |
|
 |
ginge
Site Admin
Joined: 14 Jan 2006
Posts: 1028
Location: Manchester, UK
|
| Posted: Mon Oct 16, 2006 7:51 pm Post subject: |
|
|
Sounds interesting! I do like the idea of being able to retract all of the legs, and rely on wheels. It's certainly a different approach to the one I was going to take, which was to have a lockable wheel at the end of each leg.
If you are interested in current sensing, you may way to check out Andy's page at http://vizlog.com/robot/ He has done exactly what you describe with current sensing, to the end of feeling for uneven ground.
Barry
_________________
http://www.headfuzz.co.uk/
http://www.robotfuzz.co.uk/ |
|
| Back to top |
|
|
linuxguy
Joined: 16 Oct 2006
Posts: 120
Location: Beaverton, OR
|
| Posted: Mon Oct 16, 2006 8:14 pm Post subject: |
|
|
| "ginge" wrote: | | Sounds interesting! I do like the idea of being able to retract all of the legs, and rely on wheels. It's certainly a different approach to the one I was going to take, which was to have a lockable wheel at the end of each leg. |
I explored that possibility already and dismissed it as not being feasible or practical. Well, that is true at least when working with the Lynxmotion Servo Erector Set. I know it is workable with custom machined parts, since some Japanese have done it.
| "caduser1" wrote: | | f you are interested in current sensing, you may way to check out Andy's page at http://vizlog.com/robot/ He has done exactly what you describe with current sensing, to the end of feeling for uneven ground. |
I am interested in possibly doing this, but within the servo where the data is most current and immediately available. I need a practical way for the robot to tell when the wheels are being stalled and it needs to start walking.
_________________
No, Mr. Jobs, the BiPod is a ROBOT. It does not play music OR interface with iTunes.
The Dynaplex Network - Robotics, Open Source, Linux, and Technology Forums |
|
| Back to top |
|
|
ginge
Site Admin
Joined: 14 Jan 2006
Posts: 1028
Location: Manchester, UK
|
| Posted: Mon Oct 16, 2006 8:19 pm Post subject: |
|
|
I am working on adapting the OpenServo for use with a servo modified for continuous rotation. With this you could detect the current velocity of the servo against it's power usage. As there is a correlation between the consumed current, velocity and voltage, you can fairly accurately detect when the motor is struggling. This way you can hand off control to your legs (or which ever scheme you have devised for movement).
Barry
_________________
http://www.headfuzz.co.uk/
http://www.robotfuzz.co.uk/ |
|
| Back to top |
|
|
linuxguy
Joined: 16 Oct 2006
Posts: 120
Location: Beaverton, OR
|
| Posted: Mon Oct 16, 2006 8:26 pm Post subject: |
|
|
| "ginge" wrote: | | I am working on adapting the OpenServo for use with a servo modified for continuous rotation. With this you could detect the current velocity of the servo against it's power usage. As there is a correlation between the consumed current, velocity and voltage, you can fairly accurately detect when the motor is struggling. This way you can hand off control to your legs (or which ever scheme you have devised for movement). |
This looks like what I would need! I'd be happy to do some testing of this for you once I start building my hybrid. I plan to build the wheel system first, since I can play with that while I am gathering parts for the legs. I have designed several custom legs so far, and will be using a leg with 7 1/2"" to 9"" reach on this hybrid. Being able to sense stall on the wheels and load on the legs will be important, since this this robot or one similar will be a stair crawler.
8-Dale
_________________
No, Mr. Jobs, the BiPod is a ROBOT. It does not play music OR interface with iTunes.
The Dynaplex Network - Robotics, Open Source, Linux, and Technology Forums |
|
| Back to top |
|
|
linuxguy
Joined: 16 Oct 2006
Posts: 120
Location: Beaverton, OR
|
| Posted: Wed Oct 18, 2006 5:02 pm Post subject: |
|
|
| "robotjay" wrote: | Second of all... 40 servos!?! Good God!  I'm very curious to see what you have in mind. You should post your idea in the robotics part of the forum, if you aren't trying to keep it too much of a secret...  |
The above quote was taken from the Hardware forum. Below is what I will start building, hopefully next month.
Hybrid Octapod, Front view:
Side view:
Top view:
Hybrid Hexapod, Front view, with wheels retracted:
I have already completed the design of a custom body for this robot that has two long decks with shorter add-on decks. The robot will be just under 12 inches long.
The wheels are all independently driven, steerable, and fully retractable for walking. The two frontmost legs will have grippers and the length will be adjusted appropriately to accomodate them.
8-Dale
_________________
No, Mr. Jobs, the BiPod is a ROBOT. It does not play music OR interface with iTunes.
The Dynaplex Network - Robotics, Open Source, Linux, and Technology Forums |
|
| Back to top |
|
|
cadcoke3
Joined: 11 Oct 2006
Posts: 3
Location: Lancaster, PA
|
| Posted: Thu Oct 19, 2006 2:07 pm Post subject: |
|
|
| Quote: | 8-Dale said;
I can handle complexity. It's those simple things that confound me. |
Would you like some ideas on simplification?
Of course, we don't know all of the goals yo have for your robot, so won't necessarily see the reasons you have for so many servos. But, even if the reason is ""just to look cool"", that is OK. Don't let anyone discourage you.
If the purpose is to make an agile robot, perhaps one that can do stairs, then I might have some ideas for you.
Walking robots are generally very slow, and tend to consume a lot of battery power. The dual method of locomotion allows you to have the best of both worlds. However, I think you can take better advantage of those two worlds.
The weight on the servos will be an issue... to reduce weight (and the servo count) switch to differential drive and mount the two drive wheels directly to the body. (put a caster on front and back) Also, the way you have the drive wheels on the servos is going to be prone to breaking them unless you drive at a crawl. This is because running over small objects, or bumping into things would put shock loads on the wheels.
8 legs also seems a bit overboard. 6 would probably do.
Joe Dunfee |
|
| Back to top |
|
|
linuxguy
Joined: 16 Oct 2006
Posts: 120
Location: Beaverton, OR
|
| Posted: Thu Oct 19, 2006 7:03 pm Post subject: |
|
|
| "cadcoke3" wrote: | | If the purpose is to make an agile robot, perhaps one that can do stairs, then I might have some ideas for you. |
This one will not be a stair climber. It is a prototype of what I want to enter in a certain competition someday. I just want to see if this sort of configuration can work in the real world, and I believe it can. It will also be constructed to serve in certain real world applications.
| "cadcoke3" wrote: | | The weight on the servos will be an issue... to reduce weight (and the servo count) switch to differential drive and mount the two drive wheels directly to the body. (put a caster on front and back) Also, the way you have the drive wheels on the servos is going to be prone to breaking them unless you drive at a crawl.This is because running over small objects, or bumping into things would put shock loads on the wheels. |
The drive wheel attachment method is open to change. I am looking at ways to strengthen this in the future.
These wheels are independently steerable and retractable, so I can not replace any of them with casters. Without the independent steering and drive, the robot would not be able to rotate in place when in wheeled mode, nor would it be able to take advantage of different steering methods.
| "cadcoke3" wrote: | | 8 legs also seems a bit overboard. 6 would probably do. |
This robot will not always be walking on all 8 legs. It will have grippers attached to the 2 front legs. When using the grippers, it will walk as a hexapod or be rolling on its wheels.  I also intend to make this robot smaller as I go through various iterations of the design - I don't expect to go direct from this design right to the final design.
As a side effect of the wheel system design, there is some unintentional but welcome coolness built in. This robot can lower each wheel independently, therefore giving it the abilty to be a ""low rider"" for showing off. It can also lower the wheels to increase headroom for passing under things.
This robot is very flexible.
8-Dale
_________________
No, Mr. Jobs, the BiPod is a ROBOT. It does not play music OR interface with iTunes.
The Dynaplex Network - Robotics, Open Source, Linux, and Technology Forums |
|
| Back to top |
|
|
cadcoke3
Joined: 11 Oct 2006
Posts: 3
Location: Lancaster, PA
|
| Posted: Thu Oct 19, 2006 9:57 pm Post subject: |
|
|
| Quote: | | These wheels are independently steerable and retractable, so I can not replace any of them with casters. |
While a pure differential drive wouldn't be able to drive sideways, it certainly can turn in place. When you truely want to ""crab sideways"", then you have the option of using your legs. That is what I meant by taking advantange of the dual modes of transport.
Regarding wheeled sideways travel, I 've help design a dual-wheel system for use in a theatrical stage platform. The wheels were independently powered, and also steerable by a belt that ran around both of their vertical axes. It could rotate in place, by driving the wheels in opposite directions, and go sideways by turning the wheels 90 degrees from the front. There were only 3 motors needed to do this.
Of course, a 4 wheel drive system has benenfits in its reduncy - just ask any Mars Rover fan. But, for a walker one issue is that it also adds weight, which needs more power, which needs a heavier battery... etc.
Also, since you can raise and lower the legs, it is redundant to also raise and lower the wheels. You might gain some minimal clearance, but you could probably regain that lost clearance by just having legs a bit longer.
But, regardless of anything I've said or if any of its is incorporated... when I see your robot do it's thing, I will think ""That's so cool!"".
Joe Dunfee
_________________
Joe Dunfee |
|
| Back to top |
|
|
linuxguy
Joined: 16 Oct 2006
Posts: 120
Location: Beaverton, OR
|
| Posted: Fri Oct 20, 2006 2:57 am Post subject: |
|
|
| "cadcoke3" wrote: | | While a pure differential drive wouldn't be able to drive sideways, it certainly can turn in place. When you truely want to ""crab sideways"", then you have the option of using your legs. That is what I meant by taking advantange of the dual modes of transport. |
I undersand. I just prefer to have four solid wheels always on the ground when the robot is rolling along, at least for my applications.
While I do have a serious vision for this overal design, I want it to be fun to work with and show off also. I also like the novelty of my drive system, and the tricks it can do. I planned this design to be an overal experimental platform on the way to a final design to hopefully end with the final vision I have for it.
| "cadcoke3" wrote: | | Regarding wheeled sideways travel, I 've help design a dual-wheel system for use in a theatrical stage platform. The wheels were independently powered, and also steerable by a belt that ran around both of their vertical axes. |
I already have my little differential drive robot to play around with. I like variety, so the second robot can't share any major characteristics with the first one. With this design, I tried real hard to cram all the features I like most about other robot designs into one. I think I succeeded, and now I want to see it all work. I am very excited about the software development opportunities I will have with this design once I have it built and working.
| "cadcoke3" wrote: | | Of course, a 4 wheel drive system has benenfits in its reduncy - just ask any Mars Rover fan. But, for a walker one issue is that it also adds weight, which needs more power, which needs a heavier battery... etc. |
Power is my only real concern with this design. One possibility is that I can always drop back to a Hexapod design to take a bit of weight off and reduce power requirements if necessary.
| "cadcoke3" wrote: | | Also, since you can raise and lower the legs, it is redundant to also raise and lower the wheels. You might gain some minimal clearance, but you could probably regain that lost clearance by just having legs a bit longer. |
I probably have 10 different leg designs I can try out on this robot to see what works best. It's not redundant to be able to raise and lower both legs and wheels when the robot is in a given mode of locomotion. When rolling, I don't want the robot to have to use any of its legs, except the front pair for picking up and gripping objects. Likewise, when walking, the wheels will be retracted up under the body and not used unless its necessary to switch modes of locomotion.
| "cadcoke3" wrote: | | But, regardless of anything I've said or if any of its is incorporated... when I see your robot do it's thing, I will think ""That's so cool!"". |
Oh, I believe in this design, at least in concept, even if the first implementation doesn't work out exactly the way I want. That's why I am using the Lynxmotion Servo Erector Set to prototype with - it's easy to change the design in whatever way might be required and I can have custom parts made if necessary.
I am really hoping that the current sensing function of the Open Servos will allow me to do the things I want to do with detecting when to switch modes of locomotion, as well as other things. I think this robot will be a very good test for the Open Servos, and I am very anxious to start working on it.
I have another robot, a four legged walker, that I also want to build at some point - it's a 6DOF per leg robot based on a biped leg design.
When I dream, I dream REALLY BIG.
8-Dale
_________________
No, Mr. Jobs, the BiPod is a ROBOT. It does not play music OR interface with iTunes.
The Dynaplex Network - Robotics, Open Source, Linux, and Technology Forums |
|
| Back to top |
|
|
linuxguy
Joined: 16 Oct 2006
Posts: 120
Location: Beaverton, OR
|
| Posted: Sat Nov 04, 2006 2:02 am Post subject: |
|
|
I've decided to build this Hybrid Walk 'N Roll bot as my next project. This will be an excellent application for Open Servos, as long as I can purchase already assembled electronics boards. So far, I am right on schedule to start construction of Walk 'N Roll in January of next year.
I will start out by building the wheeled assembly, since I can run that as a rover while I am building legs and the wheel assemblies are pretty inexpensive to build up.
While I am working on this, I can be researching the tri-wheel design and ways to make that work.
8-Dale
_________________
No, Mr. Jobs, the BiPod is a ROBOT. It does not play music OR interface with iTunes.
The Dynaplex Network - Robotics, Open Source, Linux, and Technology Forums |
|
| Back to top |
|
|
Delta445
Guest
|
| Posted: Sat Dec 09, 2006 10:10 am Post subject: |
|
|
| When was the last time you guys searched this forum for the answers? |
|
| Back to top |
|
|
trashmanf
Joined: 29 Jan 2007
Posts: 2
Location: Kent, WA
|
| Posted: Tue Jan 30, 2007 7:28 pm Post subject: |
|
|
i just joined the forum and wanted to say that while complex, your idea is really cooooool!!! please keep us posted on your future iterations of the design?
_________________
I am Open Source |
|
| Back to top |
|
|
linuxguy
Joined: 16 Oct 2006
Posts: 120
Location: Beaverton, OR
|
| Posted: Thu Feb 01, 2007 6:15 pm Post subject: |
|
|
| "trashmanf" wrote: | | i just joined the forum and wanted to say that while complex, your idea is really cooooool!!! please keep us posted on your future iterations of the design? |
I have pretty much frozen the design now. The only thing keeping me from moving forward is lack of funds to purchase parts.
8-Dale
_________________
No, Mr. Jobs, the BiPod is a ROBOT. It does not play music OR interface with iTunes.
The Dynaplex Network - Robotics, Open Source, Linux, and Technology Forums |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
