Modular Robotic Arm

By Dan Chen, March 10, 2015

The goals for this week is trying to model Modular Self Assembling  Robotic Arm.

4 major modules

  1. Base module or moving plateform
  2. Lifting Module
  3. Rotating Module
  4. Grabbing Module

Options for Lifting (Arm) Mechanisms

Direct Drive with Servo
Pros: Takes up less Space
Cons: Less Torque
Timing belt
Pros: High Torque
Cons: Heavier? Require additional encoder

Actuator
Pros: High Torque
Cons: Takes up more space, expensive.
Parallel-mechanism drive by servo
Pros: High Payload
Cons: Takes up more space, expensive.

5th Option? Boston Dynamics’s Spot? 

hallwaykick.0 

Screen Shot 2015-03-09 at 11.54.04 PM

Parallel-mechanism Design & Test

For parallel-mechanism, I tried to model after this Kuka Robot below. For a module system, this might take up too much space.

PR_KUKA_Industrial_Robot_KR40_PA_01-Recovered

Screen Shot 2015-03-09 at 11.06.44 PM

 

Self Assembling / Dissembling

Requirements

  1. Click / Snap on (Magnet)
  2. Data & Power Connection (Pogo Pins)
  3. Stiff Linkage
  4. Detachable (solenoid)

Self Alignment & Snap Test

I set out to create an interlocking system that can be roughly aligned and clicked in place by magnets. The spot where the two module meets would also have a pogo pin to provide power and data, probably serial communication. If the weight requirement allows I might add a solenoid for detaching the modules.

Take aways: To crate a rough meeting surface that gets attracted by magnet it’s generally a good ideas have a large opening then guiding the structure to a fixed point. To make the modules stuff, it’s good to have corners that locks together after the rough alignment.

 

j1

Version 2

Screen Shot 2015-03-09 at 8.16.45 PM Screen Shot 2015-03-09 at 8.15.34 PM

Future versions should include stiffer join, adding a servos for locking each module.

Combine Clickers and Servos

 

How does it work?

dc