You know you have a great Meetup on your hands when it starts at 7:30 and doesn't end until after 11 pm!
(I hope no one was waiting for us at the Hive. We temporarily relocated to Cosi due to the heat and the lack of A/C at the Hive. Keep checking your email, since we may do the same for the August Meetup.)
Here are some highlights of the Meetup. Please feel free to add your own observations:
- We attempted a conference call with Dr Geoffery Barrows from Centeye, but the background noise made it difficult to hear. The upshot: he's offering us a few sample of his company's AVR-based Optical Flow sensors to try out in our applications! I know our Quad project can make use of it, and I'm sure other folks can put it to good use with their robots. Here's some more info: http://diydrones.com/profiles/blog/show?id=705844%3ABlogPost%3A177115&page=1#comments
. Here's Centeye's site: http://www.centeye.com/
. Don't forget to join their mailing list if you're interested in this!
- New member (and mechanical design whiz) Mark showed us pictures of his 65 lb (!) humanoid robot http://www.facebook.com/album.php?aid=8481&id=100000564018560&l=956082e815
. It currently has about 17 degrees of freedom (27 total DOF when finished). There's no controller yet, but Mark's looking into the Axon (which, if he selects it, would keep the group's projects pretty much 100% AVR based).
- Glen demonstrated his "uGPS" ultrasonic location system, including both text and voice distance output (Glen - you still owe us the BASIC code for this!). He and Nelson described the analog amplifier circuit for the ultrasonic sensors. Next step is for Nelson to integrate the mathematics of least squares intersecting spheres to get 3D position estimates.
- Nelson explained the forward and inverse kinematics of the Delta robot, as described here http://forums.trossenrobotics.com/tutorials/introduction-129/delta-robot-kinematics-3276/
(includes source code, illustrations, and a link to a technical paper). He also showed some OpenGL graphics of the linkages, and is planning to simulate the delta robot in the free 3D graphics package Blender. One issue is that, although Blender does have some inverse kinematic modeling capability, it cannot natively handle the kinematic loop constraint of the delta robot (all 3 arms are constrained at the base and the effector). However, it may be possible to add these constraints via the built in Python scripting interface.
- I showed my Lego NXT based delta robot frame (no controls, just the linkages). I find Lego is frequently a good tool for rapid prototyping. However, it does not seem to handle 120 degree symmetry very well. That and the plastic Lego fasteners made for a very "compliant" structure. However, it was good enough to spark discussion.
- I have made some configuration trade-offs for my Quadrotor project and have actually ordered some of the hardware components. I don't know if they'll arrive in time for the next Meetup, but I will be updating my "build log" here if you want to keep up with my progress: http://www.rcgroups.com/forums/showthread.php?t=1212651
. We will be combining the quad with a delta robot to provide a test stand by which we can measure forces and torques on the quad.
- I mentioned Next Fab in the Meetup - a place in the Drexel campus where you can do everything from renting their extensive set of machine and CAD tools to hiring them to design your project for you (its sort of like the Hive on steroids): http://nextfabstudio.com/
(Mark H - my wife indicated that they may have storage areas for projects such as your robot.)
- The August Meetup will nominally be at the Hive, but watch for a short-notice venue change, depending on the weather. We will discuss the above projects (and those of anyone else who attends). We're also looking to expand our web presence and capability. Finally, I'm working on what should be a really cool robotics field trip for mid to late August.
See you next time!
- RoyPhiladelphia Robotics Group