Tuesday, 15 September 2015

AI Updates

Robbie's computer is still broken so I was able to catch up on some tasks I never had time for.
The potentiometer were never very accurate I have designed magnetic encoders as a replacement they are more accurate and and just plug in the the existing structure they will be fitted on the next rebuild.

The overall control was a very hard to maintain and expand natural language is not very good with robot control some verbs are shown as nouns thus wont be interpreted as commands. In NLTK you can define what words are verbs or nouns but maintaining the files is troublesome, I tried pattern_en but it suffers from the same limitations. I also tried WIT a online language processor the learning curve is too steep and I wanted a local solution. Robbie's chat engine on the other hand works well.

I never really looked into pyaiml's capabilities but it can run system programs with command line args. For testing I reduced the loaded aimls to two one for commands the other for general response.
Of course that just puts me back to where it was before but with a lot more potential. Pyaiml will throw a error message for a unknown command I made it so it will append the command to a aiml file I only have to add the meaning later I can automate this but for now I want control over it, this sort of gives me a learning robot.
One of the intriguing possibilities is to query Knowrob ontologies.
For now I can add the name of a person from the face recognition node.
Next task is to make a semantic map and name the objects so when asked his location Robbie will answer “near the desk in the garage” not x,y,z.

All of Robbie's physical task are now controlled through behaviour trees program with action servers any task can be pre empted and resumed if there is a fault or error condition. The behaviour tree also monitors and controls Robbie emotions tasks will give pleasure doing nothing will result in boredom, when boredom reaches a certain level Robbie will do a random task that varies from uttering a quip using a Markovian chains, moving his head or arms to driving around in circles.
Using simulators like Rviz and Gazebo has made these tasks much easier.

Thursday, 18 June 2015

Robbie Upgrade



The original base was causing some problems with arms in the relaxed and travel position the arms
would hit the lower body , I don't want the arms to make Robbie to wide he has enough trouble getting through door ways as it is also I didn't have the time for a total rebuild (moving the wheels further inboard) so I opted for some covers and a redesigned waist joint. I ripped the speakers apart

and put them in the space for the kinect. I have put the kinect back in the head so I can visualise the work space  




Monday, 6 April 2015

Autonomous Update 2

For a robot to be autonomous it must be easy for others to operate it logging in and having to run some launch files is never going to work. The answer is upstart we tried this a few time but it never worked as expected. The version that comes with indigo works now to start Robbie you just have to turn him on. Then you can control Robbie with voice commands or Rviz off course with the greater reliance on voice commands we need to expand his capabilities.  One of the options is show time which demonstrates his capabilities in a continuous display, good for showing off  to visitors. Other commands will be included over time. The original NLTK is showing its age now Pattern_en shows a lot more promise but with a steeper learning curve

Tuesday, 31 March 2015

Get Me a beer

This is Robbie's first simulation run of picking up a object of a table moving it to another table, then 
returning to his station. The simulation didn't show the object attached to the hand but the hand opens and closes in the right position


Wednesday, 11 March 2015

ROS MoveIT Update

I have had Moveit working for a while now but I could never get pick and place work I could
move the arm to a location but that was it. After a bit of research I traced the problem to my
URDF, I didn't place to coordinate frame at the end of the joint and the orientation of the frame
was wrong. After I adjusted everything it worked the arms can now reach more places and the resulting  arm position is more natural. the only real problem is the simple URDF I made for
testing works quicker and has a better result with pick and place(less fails). Next up is a test with
the real robot I have to change the joint configuration on Robbie to match the URDF the wrist yaw joint worked better than the wrist pan joint now the arm has 7 joints plus the gripper not the 8 from before


Wednesday, 4 March 2015

Robbie system update

With navigation working as expected now is a good time to upgrade Robbie to the next stable version
ubuntu 14.4 , ROS Indigo and clean up the directory structure. Putting every thing under Robbie_ros.

The good news is most thing works as before the only glitch so far is Arduino 1.6 will mess up my
messenger commands they have to be loaded with Arduino 1.05 and Face recognition is having a problem with opencv

I calibrated battery voltage to read true value, health monitor will move to docking station pose when battery voltage goes below the set level(12.2 for now) and start 
auto dock. I still have to work on the charge state message.

The most important thing to remember running Gmapping before we start localization


Sunday, 8 February 2015

Autonomous Robot Update


The reliability program is on going, I have identified many problems
that need correcting the most surprising is heat the weather has been
very hot of late causing the computer to reset. On the software side
the old repository had to many old programs that were several versions
old I have replaced the structure with a more ROS like structure.
I have refreshed the moveit package and now control both arms
with Moveit. While it works with out tearing it self to pieces the old
L298N motor drivers are not really up to the task, they lack control below
70 PWM making fine adjustment near impossible. I never had the I2C bus
working with the Arduino Due reliably so I reverted to using a Arduino UNO
I still have the occasional glitch especially when doing restarts and poking
around the boards. That brings us to the subject of wiring, crimp pins
(header types )are not the best for a robot, they have been the biggest
problem to reliability I will go with aviation style solder pins and plugs.
The new motor drivers will be MC33887 with magnetic encoders not sure
what to do with the I2C bus