I'm working with a small group of students to pilot a new education program called Robot Inc. We're challenging students to start their own robotics startup. We have a thriving First Robotics program but I've always been disappointment by how much money and effort goes into those competitions only to build a robot that really doesn't do real work. Additionally the robots are always very similar year to year and are remote controlled instead of autonomous. If that same effort and funding could be going to building a real robot that does useful work, then I'm sure we can find willing buyers. We're using the AR2 to give the students real robotics experience before they create their own custom robot.
This robot was 3D printed on a Uprint FDM printer. This type of printer creates almost solid parts and is highly reliable. Additionally, we laser cut the covers out of 1/8" black acrylic The cover support for J2 was 3D printed as one piece on a Mark Forged Onyx printer with kevlar reinforcement to give it a bit more rigidity. We still have some adjustments to do as well as clean up the wiring but its looking pretty good. We're having fun getting used to the software.
One cool upgrade would be to use Dynamix servo motors with feedback. If those were used then you could "teach" the arm by physically moving it to the position you wanted then teaching it that position. You could also then incorporate crash detection. It would definitely increase the cost substantially but could be a worthwhile improvement for the pro users.