3D Printed Robotic Hand

This robotic hand can be controlled using a glove. For example, when the user points their index finger, the robot will do the same. Sensors in the glove communicate with servos in the robot using an Arduino Uno and code.

Engineer

Kriselle T.

Area of Interest

Electrical Engineering & Computer Science

School

Summit Public Schools: Rainier

Grade

Rising Senior

Reflection

Six weeks ago, building a 3D-printed robotic hand seemed like something that was completely out of my reach. I didn’t know how a capacitor worked, or that soldering irons even existed. As the days went by, I learned more at BlueStamp than I could ever have learned in a traditional classroom. I learned a lot of technical lessons, but I also learned how to be independent, find solutions for my problems, and ask for help when I need it. I’ve gone through two batteries, eleven servos, dozens of yards of fishing line, and more wires than I can count. But in the end, I was able to pull it all together.

At first, I was interested in biomedical engineering. As I learned and experienced more, my interests continued to change. One week, I would want to major in mechanical engineering, and the next, I would be set on software engineering. For now, I’m somewhat sure that I want to go into electrical engineering and computer science, or EECS.

Since documenting my last milestone, I made a few adjustments to my project. I lengthened the wires attached to the glove so that I could control the robot without having to be too close. I also removed the duct tape from the glove and sewed the flex sensors on instead. I kept one layer of duct tape to protect the flex sensors and to allow for some material to sew into.

My code, bill of materials, build plan, and schematics can be found on my GitHub.

The STL files used for this project, created by Unlimited Tomorrow, can be found here.

Final Milestone

Servos Installed In Hand

Wiring for Hand

Rotational Joint in Thumb

Fishing Line in Joints

Battery and Makeshift Connector

All Hardware

Assembling The Hand

To finish my base project, I had to assemble the parts for the 3D-printed hand. I separated the parts from their printed structures and used a file and sandpaper to clean the holes and surfaces of the hand. I ran fishing line through holes in the fingers and palm, then connected each joint using bolts and nuts.

Installing Servos

Then, I had to connect the ends of the fishing line to the servos. I first had to permanently attach the servos to the hand in order to make sure that they didn’t move when powered. Because I wanted to use one more servo than the arm can hold, I had to superglue the extra servo into an empty slot. Because this space wasn’t meant to hold a servo, I had to drill a hole in the side of the hand to make space for the wires in the servo. For the other four servos, I used a dremel to create holes that I could use to screw the servos into.

Threading Fishing Line

Finally, I connected the fishing line to the horns of each servo. This was the most difficult part of the project. If one fishing line is too loose or too tight, the fingers won’t bend or straighten all the way, and the fishing line has to be uncrimped, adjusted, and crimped again. At one point, I resorted to cutting the fishing line and restarting everything.

Powering The Servos

Before implementing it into my circuit, I used a multimeter to test the voltage of my battery. It was giving 8.5 to 9 volts instead of the 7.2 that I expected, so I added a voltage regulator to drop the voltage down. When my servos were connected to the battery, they spun for a few seconds before stopping. Each time I connected them to power, they spun for longer and longer periods of time until I couldn’t control them anymore. I wanted to find a way to stop this, since it would cause the fishing line to snap if it continued. I tried adding different potentiometers, resistors, and diodes until I realized that the circuit didn’t have too much current, it had too little. After I removed the regulator, the servos stopped spinning.

Future Plans

The orientation of the thumb’s base joint causes the finger to occasionally rotate instead of bend. If it’s possible, I want to replace this joint with one that does not rotate. The thumb also does not bend to it’s fullest ability because of the servo that it is attached to. If possible, I want to find a way to turn more efficiently. In addition, I plan on using XBee to make the connection between the glove and the hand work wirelessly.

Second Milestone

Wiring on PCB

Close-up of PCB

For my second milestone, I transferred the wiring for the flex sensors from the breadboard to a PCB. This would allow me to sew it onto the glove when my wiring is complete. To do this, I had to solder 30 different items onto the PCB. I created a schematic for this, but I wasn’t able to follow it exactly. In order for the flex sensors to continue working, wires still needed to be connected to power and ground. However, I cut my wires according to how I would need them for my next milestone, so they weren’t long enough. I had to attach them onto a breadboard and attached the breadboard to the Arduino.

I also added clincher connectors to the flex sensors so that I could solder them to the wires on the PCB. After, I covered the flex sensors in duct tape, then taped them to a glove. I learned that the duct tape was useful for protecting the flex sensors, but they should be sewn onto the glove in order to detect as much bending as possible. To compensate, I re-calibrated my code. For my next milestone, I plan on assembling the parts for the 3D-printed hand and connecting the hand to the glove.

PCB wiring with flex sensors attached

Glove with flex senors attached

First Milestone

I reached my first milestone after I connected my flex sensors to my servos using an Arduino board. First, I drew a schematic using a software called Fritzing. As I made mistakes and adjusted my wiring accordingly, I updated my schematic to ensure that it was still accurate. The flex sensors act as variable resistors that change in value as it bends. The Arduino board takes in these changes in resistance by reading the voltage drop between the flex sensors and the 22k resistors. After, Arduino uses the code to translate them into a certain amount of degrees that the servos should rotate, then sends these degree values to the servos. As a result, the servos turn a certain amount when their corresponding flex sensors are bent.

My code and schematic were based off of the projects of two previous BlueStamp students: Annabel Y. and Sanjana K. My code and schematic can be found here.

Servos and Flex Sensors on Breadboards

Wiring Schematic

Starter Project: Minty Boost

minty boost

Completed Starter Project (No Casing)

Open Case

Inside of Minty Boost Case

Altoid Casing

Outside of Minty Boost Case

My Starter Project, Minty Boost, was a battery-powered USB charger. It helps to “boost” the voltage given by two AA batteries in order to charge any device that can be charged using a USB port. It does this by using a power inductor to maintain a steady current while the voltage is changed by the boost converter. The capacitors smooth input and output voltages, and the resistors are used by Apple iPhones to determine what kind of charger is connected. Minty Boost takes power from low voltages, like the ones given by the batteries, to high voltages, like the ones that an iPhone needs to charge. While assembling this project, I learned how to solder. I also learned about a lot of different components that I wasn’t familiar with before. I also installed the “Minty” Boost into an empty Altoids case, which was originally used to store mints. This makes my project much cleaner and easier to transport.