We wanted to control and experiment with a physical activity. We made the arm/hand mechanism with 5 servos. The mechanism commands and imitates the function/ movement of the human forearm and hand. I made a glove filled with sensors, capable to capture the wearer’s movements (human hand) and translate the values in programming in order to direct and control the mechanism of the robotic hand.
I first started by building the simple Animatronics (robotic hand) with tubing, string, tapes, CD-cover Plastic, and a hot glue gun. By cutting away inside the joints, we could get a smooth movement on each finger. Cording is attached at the end of each fingers (thumb and index are attached together). The cords are attached to the servos that will control their flex movements. I made a forearm wearing a shirt, just to add the “real effect” to the robotic arm.
See: http://www.instructables.com/id/Simple-Animatronics-robotic-hand/
For this project I really wanted to try different techniques and materials.
The glove includes one bright red LED for video, 4 hand made/sewn bend sensors for all fingers movements, and a stretch sensor for the wrist twisting movements. All circuits are traced with conductive thread attached to wires at wrist.
After talking to Rory, he directed me to the “How to get what you want” website: www.kobakant.at. I decided to test the Fabric Bend Sensor because I though that would be the most suitable type of sensor for our project. The fabric bend sensors give good results (sensitivity) for measuring the bend of human joints when attached to the body. It is sensitive enough to register even slight bend and has a large enough range to still get information when the limbs are fully bent. The neoprene is great for isolating the conductive thread stitches and keeps the sensor from wrinkling even when repeatedly being bend.
Materials and Tools
Materials: Neoprene, conductive thread, conductive fabric, velostat.
I also used a conductive ink pen to secure the attachments with wires.
Tools: Fabric scissors, fabric pen, stencil, sewing needle, iron, pliers.
The resistance range of this bend sensor depends a lot on the initial pressure. But this can vary, depending on how the sensor is sewn and how big the overlap of the adjacent conductive surfaces is. The sensor still detects a difference, right down to about as hard as you can press with your fingers. The range is non-linear and gets smaller as the resistance decreases. The problem is that whoever wears the glove will give different values, so you have to remap the values in programming every time.
We also made a pocket to hide the Arduino and breadboard with the rests of neoprene.
Good websites for Wearable materials:
http://www.fashioningtech.com/
http://www.imagesco.com/sensors/stretch-sensor.html
http://www.lessemf.com/paint.html?tp=CONTINUE+SHOPPING
http://www.kobakant.at/DIY/?p=20
We had a lot of fun and spent many hours on this project. It was really interesting for me, who knows only the basic of programming, to learn from Chris Langer a really complex system of programming with the Arduino and Processing. Here are couple videos.
No comments:
Post a Comment