If you have not yet installed Allegro Application Studio (AAS) of your CAN interface and driver, please see [[Allegro Application Studio (Installation) before you read this manual.

Contents

Running AAS

When installing Allegro Application Studio (AAS), you were promted to select an installation path where are files associated with AAS would be stored. This installation path with be referred to as [installPath] from here on out.

Within the installation directory (at the top level), you will find a file titled Allegro.exe file. Also, a shortcut to this executable should have been placed on your desktop during installation. Double-click Allegro.exe to launch AAS for the Allegro Hand.

AllegroIcon.jpg

Which Hand?

When you first run Allegro Application Studio (AAS), you are promted to select the hand (right/left and actual/virtual) you would like to use.

HandSelection.jpg

Actual

Actual System refers to the Allegro Hand hardware. If an actual system is selected, AAS will load the corresponding (left/right) virtual model and open CAN communication with the actual Allegro Hand. The virtual Allegro Hand seen along with the Actual System simply mimics the joint kinematics (no physics or contact dynamics) of the actual system based on the encoder values obtained over CAN. Motion commands from AAS are send directly to the Actual System while the encoder feedback controls the AAS virtual hand.

Virtual

If a virtual system is selected, a dynamic simulation is run using the Allegro Hand virtual model. This simulation includes full physics simulation including contact dynamics. The virtual system can be simulated without the actual system. This ability is useful for testing algorithms in a dynamics environment before trying them out of the actual hardware.

Virtual Hand Simulation

First, we will select the Virtual Left Hand System and click Finish to start the simulation.

Selection either virtual hand system at the Allegro Application Studio (AAS) hand selection prompt will load the AAS dynamics simulation environment and a virtual hand model. The simulation begins running as soon as the AAS simulation window is loaded.

AllegroVirtualHand.jpg

AAS is provided along with the Allegro Hand for two main purposed. AAS allows for ease of testing CAN communication between your PC and the Allegro Hand hardware. Secondly, AAS provides several robust grasping algorithms for use with the Allegro Hand. The algorithms provided with AAS can grasp a variety of object geometries for demonstration or for use at the end of a manipulator.

The grasping algorithms provided with AAS can be explored via the graphical user interface (GUI) buttons at the top-left of the AAS simulation window. Clicking these buttons will send a command to the virtual CAN device on the virtual Allegro Hand model.

Home

Let's start by clicking the button entitled Home. The position assumed by the Allegro Hand is a starting position that ensures that all joints are oriented properly for executing a grasp.

Ready

Click the button entitled Ready.

Grasp 3

Click the button entitled Ready to prepare the hand for another type of grasp.

Now Click the button entitled GRASP3. This grasping algorithm is a torque-controlled, three-fingered grip. This grasp can be used for pick-and-place style object grasping as the object is held between the tips of the thumb, index and middle fingers.

Grasp 4

Click the button entitled Ready to prepare the hand for another type of grasp.

Click the button entitled GRASP4. This grasping algorithm is a torque-controlled, four-fingered grip. This grasp can be used for pick-and-place style object grasping as the object is held between the tips of the thumb and three fingers.

Pinching (I)

Click the button entitled Ready to prepare the hand for another type of grasp.

Click the button entitled Pinching (I). This grasping algorithm is a torque-controlled, two-fingered pinch. This grasp can be used for pick-and-place style object grasping and more dexterous manipulation as the object is held between the tips of the thumb and the index finger.

Pinching (M)

Click the button entitled Ready to prepare the hand for another type of grasp.

Click the button entitled Pinching (M). This grasping algorithm is a torque-controlled, two-fingered pinch. This grasp can be used for pick-and-place style object grasping and more dexterous manipulation as the object is held between the tips of the thumb and the middle finger.

Envelop

Click the button entitled Ready to prepare the hand for another type of grasp.

Click the button entitled Envelop. This grasping algorithm can be used to fully envelop an object within the the hand's four fingers. This algorithm can handle a variety of object geometries.

Sliders

The sliders at the bottom of the AAS simulation window provide a secondary form of visual feedback for the sixteen joints of the hand. Each column represents a finger with link 1 being the yaw joint attached to the palm and link 4 being the last joint at the end of each finger.

AllegroVirtualHandG4.jpg

Actual Hand

A new Allegro Hand, either virtual or actual, left or right, can be loaded by clicking New in the File drop-down menu (File > New).

Again, you will presented with the AAS hand selection window. This time we will select the Actual Left Hand System. Don't click Finish yet!

If you have not yet installed your CAN interface and CAN driver, please see




You are not allowed to post comments.




Copyright & Trademark Notice
Allegro, the Allegro logo, RoboticsLab, the RoboticsLab logo, and all related files and documentation are Copyright ⓒ 2008-2020 Wonik Robotics Co., Ltd. All rights reserved. RoboticsLab and Allegro are trademarks of Wonik Robotics. All other trademarks or registered trademarks mentioned are the properties of their respective owners.

Wonik Robotics's Allegro Hand is based on licensed technology developed by the Humanoid Robot Hand research group at the Korea Institute of Industrial Technology (KITECH).

Any references to the BHand Library or the Allegro Hand Motion and/or Grasping Library refer to a library of humanoid robotic hand grasping algorithms and motions developed and published by KITECH researchers.
J.-H. Bae, S.-W. Park, D. Kim, M.-H. Baeg, and S.-R. Oh, "A Grasp Strategy with the Geometric Centroid of a Groped Object Shape Derived from Contact Spots," Proc. of the 2012 IEEE Int. Conf. on Robotics and Automation (ICRA2012), pp. 3798-3804

Wiki maintained by Sean Yi <seanyi@wonikrobotics.com>

KitechLogo.jpg Wonikrobotics logo.png





Whos here now:   Members 0   Guests 0   Bots & Crawlers 1