I have a student who’s going to be assembling a 4WD-3 and an SES arm. We originally thought we’d mount the arm on the top of the rover, as shown in the pics on the Lynx Motion site. But I wonder if it would make more sense to mount the arm horizontally on the front of the rover, like a crab claw.
An obvious issue is that the weight of the arm would be constantly dragging the shoulder servos down, so to reduce power draw and prevent overheating we would either need some sort of spring or counterweight arrangement, or we’d have to rethink the design of the arm entirely.
Has anyone done something like this? I’d love to hear suggestions for how to approach this.
Pick up small objects (like childrens’ blocks) and put them down somewhere else. I bet no one’s thought of doing that with a robot arm before.
The reason to mount the arm horizontally would be to give the robot greater forward reach, and make it easier for a camera to monitor the arm (by looking down on it.)
What a totally original idea! I never thought you could pick up small objects with a robotic arm
Heres a possible idea. You could either gets a really expensive servo with uber amounts of torque (Probably will have to go for the mega size). Or you could do something else. What about putting a DC motor to rotate the “shoulder” instead of a servo. You could even do some crazy gear ratios to increase torque. Or you can get it prebuilt: servocity.com/html/servo_pow … boxes.html
Some very nice gear ratios. Amazing torque! up to 1,715OZIN!!
I would love to see picture and videos of this rover running and collecting some small objects! Please post some when you get it working!
Thanks, robodude666. I like the idea of using a RobotZone gearbox to increase the torque without going to bigger servers and higher current draw. And by sticking with a servo instead of a motor, we can control everything with the SSC-32.
Another idea I had was to position the arm on the servo so that when the arm was at its lowest position, the servo was at its mechanical limit of rotation. That way, when the robot was just running around, I could let gravity do all the work of keeping the shoulder down, and the arm still wouldn’t drag on the ground.
We will definitely post some pics when we get this thing working. We’re also adding a Gumstix and a USB webcam. Intelligence will be provided by our Tekkotsu software system (see Tekkotsu.org).
If it is a first time project, there is always value with using previously proven setups, including arms. Depending on the over all goal of the project, significant time and energy may be diverted from the overall goal just trying to get one specific part functional. If you don’t see already existing examples of what you want to do, then there is probably a reason.
Good point. We’ve already built one SES arm, which we’re using to develop Lynx Motion arm support for Tekkotsu. But we’ve never built a rover before. (We’ve been using AIBOs for the last 5 years.) And I just realized that the weight of a front mounted arm might cause stability problems for the rover, though I suppose we could counterbalance it if we didn’t mind the added weight penalty.
To do visual servoing of the arm we need an unobstructed view. That’s one thing the AIBO did well: putting the camera on the end of a 3DOF neck instead of a conventional pan/tilt makes it much more versatile. So another reason to go with the crab claw is that we could then put the camera on the end of a separate top-mounted manipulator that could look out and down on the gripper, or up and around for navigation.
I’ve seen setups where the cam is mounted at the elbo end of the forearm so it would be looking at the gripper area. This mounting also allowed the cam to be used for general visual surveillance as it can pan/tilted via the arm itself. One thing to consider with cams is for small bots the cams need some way to adjust the focus. Most fixed focus cams don’t do well at close distances.
what do you mean? That was a temporary camera connected to the arm with foam tape. I now have a nice camera attached with a beautiful aluminum mount 8)
What kind of camera are you using? And can we see a pic of the new mount? More importantly, how about a pic of the view through the camera? I’m not sure that putting the camera on the arm will afford a good enough view for autonomous manipulation. The gripper, and whatever it’s holding, might obstruct too much of the camera’s field of view.
If I was working on this kind of setup, I would keep the arm on top, do without wrist rotate on the gripper, and mount the sensor/camera out on the end of the arm. But for me this would be on a hexapod.
Joe, it holds but it doesn’t look very stable. You can see its a bit bent. Having a geared servo would give you more torque there and the ability to pickup heavier objects and/or hold them for a longer period of time.
??? ur mistaken, its a pinhole camera…IDK how you thought that, its pretty clear in the picture
And Robodude as I mentioned above: That was an older setup which was TEMPORARY while I was waiting for my 2.4GHZ camera to arrive via UPS. This setup is pretty much the best way to pick up small objects on the ground as many of people have stated to me.
than whats that other thing on the robot? maybe a…OH…wait i forgot,didn’t you mention yo me that you put a light on the bot thats what i was looking at.lol
Yep, Ive noticed that. Its kinda cool though sometimes it makes it hard to see when your image is full of white becuase there are a lot of lights around.
i,m gonna do something like this and had the same thought that it would be tough on the shoulder, i,m thinkin0g of using a power window motor or jims super servo ,
