Friday, 8 April 2016

Michal Ogrodniczak - Sprint 2 reflection.

Team : Mechanical
Name: Michal Ogrodniczak
Section: Head tilt and pan


Technical Reflection

In this sprint I have been tasked with building and coding up pan and tilt head movement. Servo work itself was not really difficult as at that time we were not interfacing with any other module.
Program consisted of two processes - one listening to the keyboard inputs in format -
<servo no.> <dc>, and other process was receiving this input through a pipe and ‘smoothly’ moved servos. Reason that the movement wasn't really smooth could be pinned down to the weight of the head and perhaps some undiscovered bugs in the system as it seemed to work a little bit ‘off’ at times and other times it was spot on.
Summing up - I ve had fun designing and building the prototype with Shaun, at the end of this sprint we have concluded that tilt servo need more umph and we have already found what seems to be a worthy replacement. We have also ran into issue we’ve all met during sprint 1 - twitchy servos - we hope to resolve this through use of the Rpi servo hat.

Project management reflection

The way the overall project was managed seemed ok at the beginning, but soon we’ve started panicking as work didn't seem to tie up together. This was inherent nature of this sprint as we have selected smaller teams to do specific tasks. This approach worked out for most part as every team had successful demo of their own task. In my opinion, it is safe to say that moving forward each team has substantial knowledge of their domain and it should help during system integration phase. Overall process however would be made smoother if we have one big boss kind of a role that would keep an eye on the progress and could report any issues to the whole group so we could tackle those together.

Teamwork reflection

I was teamed up with Shaun for this sprint, we have got out tasks done with time to spare, therefore it's fair to say we did something right. In case of absences no work was paused and other person just continued at whatever was bringing us closer to the finish line.
Bigger picture stuff - mechanical team - as mentioned in project mgmt. Section - we all did things separately and were not quite sure of the progress made by the other teams, but at the same time it was not crucial for us to know as most of the work was completely separate.

Personal reflection

This sprint was very much different to the first one in my experience. Building physical things and using power tools sure is fun but contrary to coding we had one ‘go’ at it. This fact did put an extra pressure but looking around the room I was reassured that all of us were quite stressed, especially at the demo video day as we all wanted to get paid. Despite induced stress I did enjoy this sprint, as in the end we’ve put some life into the LamPy head and got to learn bit more about multiprocessing and IPC.

Next sprint

Myself and Shaun have following tasks for the final sprint:
  • Put bigger servo for the tilt.
  • Implement servo code for Rpi servo hat.
  • Integrate servo code into Kamil’s thread framework.
Some of the work have already started and we are fairly confident that these changes will resolve any issues we’ve encountered during last sprint - assuming that servo hat people did not pull their claims out of a hat.

No comments:

Post a Comment