A.S.T.R.I.D. is the Autonomous Swarming Terrestrial Remote Investigation Device.
I’ve decided to start another robot project for when I get stalled on W.A.L.T.E.R. or want to work on something completely different. The SES will figure very prominently in this new project, which will be based on a slightly modified A4WD1 v2.0 rover kit. It will not use the stock motors because I want to have encoders on the back wheels. I’m not sure encoders would be useful on the front wheels.
Right now, I would like to use the GHM-14 (front) and GHM-15 (rear) motors, but these are not in stock and I don’t know if/when they will be. If that doesn’t work out, I will either use the GHM-04 or GHM-16 motors with encoders on the rear motors.
I’m going to get the A4WD1 kit next month, as I’ve wanted one of these for quite awhile now and the v2.0 rover has most of what I need for doing stuff that would not be appropriate for W.A.L.T.E.R. because he only has two wheels and not enough stability to do what I want to do with A.S.T.R.I.D.
As far as modifications to the A4WD1, I will have to a vertical double or triple SES hole setup to the side panels near where the current motor mount holes are. I will not be mounting the motors in the standard locations. This is where I will mount the new locomotion system to the rover. This system will provide for variable ground clearance as well as increased climbing ability hopefully. I’m planning to use some sort of wireless control for testing and manual over rides in case of problems.
I already have most of the other parts to build the modified locomotion system for the A4WD1. I’ve located the short spacers I need for the 2-56 screws at McMaster, which will go between the side panel and an ASB-24 (no 3D model yet) bracket.
Everything I have learned and will learn while working on W.A.L.T.E.R. will come into play with ASTRID also, since he is still my main development platform. I just need to get some 3D models done for the A4WD1 v2.0 rover parts and the ASB-24 so I can check my designs out before I build. I’m just really starting to realize how much I depend on the 3D CAD stuff for robot designs and builds.
For those who have followed my various designs, you know it will be something quite different than what anyone has seen to date… I think this is an exciting new project that will expand the capabilities of the A4WD1 rover base.
I’ve decided to go a different route and design some custom brackets for ASTRID. I found this very cool rover that uses two swinging arms, with two steerable wheels on each arm. I believe this is all workable in an SES compatible fashion. I’ve almost finished the design for the swing arms, which consist of two custom brackets with a 3" x 3" platform at the top. I’m using 6-32 hardware for this one for strength and adding SES hole patterns where they are needed. 3D CAD renders of the brackets will be forthcoming soon! I just hope I can find a place that can/will make the new brackets for a reasonable price.
I’ll be using gear head motors for locomotion instead of continuous rotation servos. ASTRID will be considerably larger than the robot she is based on. ASTRID may indeed become my platform for RoboMagellan. I haven’t quite figured out how to get the swing arms working using SES components yet, but I am sure a BBH will figure prominently in the final solution There is ample space on the top platform of each swing assembly for an Arduino stack, and I will most likely use a BeagleBoard for the main controller. I’ll have my Rev C3 BeagleBoard this week!
That is an interesting bot. I think the mechnical part could be very easily and cheaply made without the need for any special brackets. A 2’x2’ sheet of 5.2mm lauan plywood to cut the frame parts from, plastic jar tops for wheels, a little hardware for the axel, 8 el cheapo servos, and hot glue to put it together. The programming could take a while if the complex moves are needed.
Oh, I know it could be. I want to make ASTRID an SES robot though.
I already have a good idea as to how to program this. The only part I haven’t figured out yet is how to do the pivot without having the main platform get away from level to the ground. In the original robot, I know he doesn’t use any servos other than those for steering the wheels and moving (8 servos total). I would very much like to keep the pivot as simple as possible, in the spirit of the original robot.
Do you have any mechanical ideas for the pivot? Right now, I am thinking of using a Ball Bearing Hub assembly in some way. It’s an obvious way to do an SES pivot. Keeping the main platform (between the swing leg assemblies) horizontal is what is giving me some pause.
Mount the heavy stuff like the batterys underneith the platform below the platform axel centerline. This would allow a pendulum like effect to keep the platform ~level with the ground.
Ahh, yes! This is by far the easiest solution to implement, and I will try it first before anything else. It’s easily accomplished with just a pair of Ball Bearing Hub assemblies attached to the center platform bracket.
This is exactly one idea I have for a modified A4WD1 rover! I am designing four “legs” to which will be attached a steerable gear motor and wheel. The leg design is almost complete now and I have one almost fully assembled (including servos). This is all part of my ASTRID project. So, there are really two different projects that I am calling ASTRID at the moment.
I pursued a slightly different possibility for ASTRID - one I don’t need any custom brackets to build. I built up (complete with servos) a leg with a motor at the end of it. This is a 5DOF leg that provides for swinging out away from a robot’s body, being close in next to the body, and possibly even allowing for the ability to climb. The end effector steers the wheel, which allows for much more flexibility in motion, and allows the wheel to be oriented properly regardless of the way the leg is set.
My next goal is to assemble brackets for three more of these legs. With the design I have in mind each leg will be slightly different, based on its location on the robot. To allow for the symetry I want, this is the way it has to be for proper locomotion. I’m sure I have more than enough SES brackets to build thee more legs, but will have to wait to add servos. Right now, I am using mostly HS-475HB for all but the two heavy lifting joints which are HS-645MG. From looking at the current servo line up, it looks like I will need to switch to HS-485HB servos for my final build. At least I can test the lifting capabilities of this leg design with the leg I have assembled.
This design is based on one innerbreed created for one of his robots. This is the design that would be very appropriate for an A4WD1 modification. All it would need is some SES holes in the side panels next to where the motors would normally be mounted. I’m going to work on getting 3D models made for the legs and a substitute chassis to connect them to.
P.S. I’m getting a new web camera next month so will be able to start taking pictures and videos of my creations again soon!
I completed assembly of the first leg, including motor and wheel. I must say it looks rather awesome! I did a preliminary mechanical calibration of the servos so I could determin exactly how I want the legs to be able to move and the wheel to be steered. As it is now, the wheel can always be moving in the forward direction, regardless of whether the robot is moving forward or sideways to the left or right. I also found a use for the new HUB-16 1 inch SES spacers, since the wheel won’t quite clear the other side of the steering servo when moving toward the left. I think that extra 1 inch of space between the servo horn of the steering servo and the MMT-03 SES motor mount will get me the clearance I need.
I just couldn’t resist, seeing that leg all assembled and not attached to anything… I had to attach it to something! Well, the only robot body I have that can handle such a wheeled leg is WALTER’s body, so I attached the leg to one of the flat sides. I have to say it looks rather amazing! WALTER’s body is very light, and even with two batteries and electronics, the legs should be able to support everything. I have HS-645MG servos in the two heavy lifting joints.
I can’t wait to show off pictures of this extremely unique setup.
Watching that video, it appears only one leg side pivots, the other leg set is permanently fixed to center platform assembly. Sure, your sensors won’t be level, but it’s probably not that big of a deal.
Well, with ASTRID, I will have both sides on a pivot. I almost have the design of the custom side brackets finished and am going to start putting out feelers to find a place to make them. These brackets are pretty long, so it is probably going to cost quite a bit to get four of them made. I probably won’t be able to get them made for awhile. I want to get the same flexibility as what is in the robot that has inspired ASTRID.
Most of the sensors that will be on ASTRID will be in the side assemblies, so they will point where the leg is going. I want to learn more about accelerometers and put a 3-axis unit on each side as well as on center the electronics platform.
You are apparently doing something different in your “legs”. The ones in the video had a pivot in the center, such that the all four “leg” portions (fore and aft) are free to independently move.
The assemblies for ASTRID are not legs. They are like the robot in the video I posted that use the custom brackets I am designing, with each side assembly having a central pivot point and attached to a central platform. Each side will actually have a 2.5" x 3" top area where I will mount a motor controller and possibly an Arduino compatible board to handle sensor processing for that side. I hope to use my BeagleBoard as the main controller for ASTRID.
The “legs” are for a completely different project, which hopefully will be tested out on WALTER, but they will need much beefier servos than I am using now. More on this in my WALTER thread.
OK, “arms” then, not “legs”. They are still simple cantilevers with a single pivot in the center. I believe the video showed a pair of “arms/legs”, two on each side, such that either (or both) could be lifted. All four motors and their respective wheels thus appear to be independent.
I see two Scarab Lunar Robot videos. Oh wait, I see a yellow one before all that. Yes, I was talking about the Scarab, the latest one. It’s much more interesting!
The below shot shows pretty much that one side is attached to the center part and the other side is free to rotate. I read the info on it and the body is made from plywood. Cost was stated at $100. Interesting design where detailed manuvering may be needed.
I’ve decided to take A.S.T.R.I.D. in a different direction and make her a full fledged quadraped. I’ve designed a new 3DOF leg that also has ankle rotation and tilt that I really want to experiment with. I’ve already verified that this leg (using three HS-645 and two HS-475/HS-485 servos) will have the lifting power I need for. Since am changing W.A.L.T.E.R. into a 4WD robot that has independent steering on all wheels, he will be able to do the same moves as the robot that originally inspired ASTRID. I don’t need two robots that can do the same moves, so I will turn ASTRID into a quadrapod walker. I already have most of the brackets I need, less a few ASB-04, to assemble all four legs.
