High Five Robots (12/11/2017) & (4/6/2018)

I’ve built two high five robots. The first was a small one that I could hold in my hand, and the second was much larger, mostly just for laughs.

The first one was designed, printed, and built in one day to challenge myself. I was having a conversation with a friend of mine in the morning before a class, and we jokingly agreed that I should build a “high five robot.” I thought about it during my class, and decided to make it from parts I had in my dorm room. After class I went back to my room and found an arduino uno, a 9g micro servo, an ultrasonic distance sensor, and some wires to hook everything up. I started sketching out the layout of parts and how I wanted it to look, and quickly came to a dome design, vaguely reminiscent of the Staples easy button. I put a design together, with the goal of starting to print the parts before my next classes, and managed to get everything ready to go before I needed to head out. I guess one of the perks of living on campus is that you can get to classes quickly.

When I got back from my classes, I took the parts off the print bed and wired everything together, hurriedly writing code while my hot glue gun heated up. I was going to a club meeting later in the afternoon and wanted to show my friend, figuring he’d get a kick out of it. I uploaded the firmware, screwed everything together, and threw it in my bag so I could get to the meeting without even testing it. Thankfully, it worked and was a big hit. We started bringing it to events for kids to play with and to help with outreach.

The second one was a much less rushed undertaking, with the goal being to print it on my big 3D printer, so I had to be more intentional with the design since it didn’t print very well at the time. I decided to use larger parts that I also had on hand, including another arduino uno, a NEMA23 stepper motor, a TB6600 motor driver, and three ultrasonic distance sensors. The additional sensors were needed because the robot was much larger, and so it had a larger area where you could put you hand and still get a high five. So after some internal debating and some calculations, I settled on three sensors. This one had two iterations, the first one had the hand attached to an arm, which was then mounted onto a shaft that was directly driven by the motor. This had lots of speed and power, but I decided to take things a step further and use a worm gear to drive the arm, since the motor’s rapid acceleration/deceleration was causing lost steps, and the rotor’s inertia caused it to jump off the table sometimes. I tried to tune this out with different motion curves, but didn’t have much luck since I was going for a high speed, very snappy design.

For the second iteration, I ended up ordering a worm and a worm gear, and reprinted the motor mount. This change made a huge difference, since the arm could no longer backdrive the motor. There were no more lost steps, and no more jumping. After this was done, I started using it for club outreach, and it became pretty popular at the campus’ makerlab too.