System Implementation


2014-09-07: Mini-Dora base we're working with!



[Software]

2014-10-16: Kinect disparity view at 2x speed in ROS.


[Software]

2014-10-16: Hokuyo LIDAR scan at 2x speed in ROS.


[Software]

2014-10-22: Hector SLAM Implemented


[Electrical and Control]

2014-10-22: Test Mini-Dora Track Postures


[Electrical and Control]
2014-10-23: Sent wire diagram to Robert, the Mini-Dora designer/engineer, and verified the workings of Mini-Dora.

[Mechanical]
2014-10-24: First version of BUD-E CAD model



[Electrical and Control]

2014-10-25: Controlled Mini-Dora with RC


[Electrical and Control]

2014-11-01: Mini-Dora Climbing Down Stairs via RC


[Electrical and Control]

2014-11-01: Mini-Dora Climbing Up Stairs via RC


[Mechanical]
2014-11-01: Completion of BUD-E CAD model

[Software]

2014-11-04: RGBD SLAM of the lab


[Software]

2014-11-06: CCNY SLAM at 4x speed



2014-11-06: Noticed a gap between the rear-right track and frame body...


2014-11-09: Discovered Mini-Dora's rear-right motor shaft is bent:

2014-11-10: Temporary mobile base replacement for Fall Validation Experiment--Adept P3-DX:

[Mechanical]
2014-11-10: Shaved T-slot bolts so they fit in the T-slots on Mini-Dora.

2014-11-14: Visited Robert, the Mini-Dora designer/engineer. He showed us how to remove the bad motor and also gave us new replacement motors. This design and the parts are from ~15 years ago!
(Bad motor.)

(Bad motor on the right, good motor on the left. Notice the encoder shaft that we need to shave down.)

[Electrical and Control]
2014-11-14: Soldered all parts together on our Power Distribution Board.






[Mechanical]
2014-11-17: Prototype base-plate finished and attached onto the Mini-Dora.
Mini-Dora with prototype base plate

[Mechanical]
2014-11-19: Shaved down the encoder shaft and reinstalled the motor.

[Mechanical]
2014-11-20: Cut and install new base-plate out of acrylic.



[Mechanical]
2014-11-20: Scissor lift at full height while mounted on the base plate.

[Mechanical]
2014-11-20: Scissor lift at minimum height while mounted on the base plate.

[Mechanical]
2014-11-21: Scissor lift and base plate mounted on Mini-Dora.
All the mechanical components for FVE are ready.



[Fall Demo Video]
2014-12-12

[Mechanical]
2015-01-21: Mechanical modifications for fitting Pioneer base.
Group decided to use Pioneer P3-DX mobile base because the Mini-Dora broke down again and we're wary of spending more time fixing/maintaining that system.

[Grasping]
2015-01-22: Implemented Object Recognition Kitchen (ORK) in ROS and obtained initial results:


Point Cloud of the Cup:

[Mechanical]
2015-01-26: Robot arm assembly.
Crustcrawler Arm AX-12 parts


Crustcrawler Arm


[Grasping]
2015-02-02: Recognition Using ORK. (Note: it wrongly identified the mug as the iced tea bottle.

[Mechanical]
2015-02-03: Plate for mounting robot arm and Kinect.
[UI]

BUD-E Speech recognition capability


[Grasping]
2015-02-08: Tried find-object package for detection. It uses 2D image feature detection.


[Grasping]
2015-02-23: Using PCL to identify bottle on the lab bench and display the center coordinate of the bottle.


[Grasping]
2015-03-10: Detecting the bottle versus a box.



[2015-03-05]

BUD-E responding to speech commands.



[Mechanical]
2015-03-20: Scissor lift structure improvement



 
[Mechanical]
2015-03-28: New top-plate

[Mechanical]
2015-03-29: Scissor lift control by relay module

[Grasping]
2015-03-30: Arm control in ROS using moveit! package.

2015-03-30: Planned trajectory in MoveIt!

[Navigation]

2015-03-30: Online global path re-planning

Video:

Online global trajectory replanning


Improved local trajectory planner



[Mechanical]
2015-04-01: Final mechanical design of BUD-E



[2015-04-01] BUD-E responds to speech commands and navigates autonomously


[Grasping]
2015-04-20: We were not able to get moveIt! to work with our Crustcrawler arm. Thus, we went to our backup plan and used a simple inverse kinematic solver to solve for the bottom three joints (shoulder pan, shoulder pitch, elbow):


[Grasping]
2015-04-21: Arm retry testing. Cup was placed farther than the arm can reach so the retry code can be tested. Verified by check how well the arm aligned with the cup on the grasp tries.

[2015-04-21] Arm Grasp Retry Test



[Grasping]
2015-04-25: Testing grasping retry. Cup was purposely placed too far for the arm initially, then it was put on top of a box to check whether the scissor lift will adjust its height.

‎[2015-04-25]‎ Arm + Vision + Scissor Lift Testing


[Full System]
2015-04-29: Full run from Spring Validation Experiment (Encore).

[2015-04-29] Full SVE Encore Run


Comments