Embedded Motion Control 2017 Group 5 Meetings Overview

From Control Systems Technology Group

(Difference between revisions)
Jump to: navigation, search
(Meeting 1 (12-5-2017))
 
(3 intermediate revisions not shown)
Line 2: Line 2:
Reviewed initial concept design with tutor.
Reviewed initial concept design with tutor.
<li>Team needs to think more on software architecture (components, functions, flows).
<li>Team needs to think more on software architecture (components, functions, flows).
-
<li>Diagrams are not explicit enough.
+
<li>Current architecture diagrams are not explicit enough.
<li>Agreed to meet next week to discuss findings and prepare first presentation.
<li>Agreed to meet next week to discuss findings and prepare first presentation.
 +
 +
 +
== Meeting 2 (16-05-2017) ==
 +
Discussed initial design and prepared presentation.
 +
<li>Presentation will be made.
 +
<li>Information flows need to be specified.
 +
<li>Code the corridor challenge in pairs (Sjoerd & Bart, Rodrigo & Kagan, Angel & Torben).
 +
 +
 +
== Meeting 3 (19-05-2017) ==
 +
Discussed corridor challenge progress.
 +
 +
<li>Focus on framework, easier to discuss if we are more specific.
 +
<li>Think carefully on what we need/want to test (some parameters) Simulation vs Real.
 +
<li>Code the corridor challenge separately, in pairs Sjoerd & Bart (world model) , Rodrigo & Kagan (Motion Controller), Angel & Torben (Action Manager).
 +
 +
 +
== Meeting 4 (29-05-2017) ==
 +
Define testing plan.
 +
 +
<li>Tuesday, Turning 90° degrees with odometry, turn based on the laserdata, collect realtime data for worldmodel tests and tuning potential field (Angel, Sjoerd, Kagan, Bart) .
 +
<li>Wednesday, Everything not done on Tuesday, smoothing turnings, potential field with set point (Kagan, Rodrigo, Bart)
 +
<li>Use pythagoras, check for corner detection.
 +
 +
 +
== Meeting 5 (02-06-2017) ==
 +
Discussed corridor challenge result and progress.
 +
 +
<li>Things that go well (Potential field, detecting the gap and stopping in the right spot).
 +
<li>Things that didn't go well (Didn't stop after turning, due to time delays).
 +
<li>We are not using time delay anymore, we go back to odometry data.
 +
<li>First advance for Pledge algorithm (PICO got into a loop).
 +
<li>There's a problem while sticking to a wall when moving.
 +
 +
 +
== Meeting 6 (12-06-2017) ==
 +
Discussed final presentation and maze challenge.
 +
 +
<li>Tutor had no complaints about final presentation.
 +
<li>New function for align PICO and stay parallel to walls.
 +
<li>Integration of all functionalities tested, except for left and right doors (To test on tuesday).
 +
 +
 +
== Meeting 7 (20-06-2017) ==
 +
Discuss the Wiki page.
 +
 +
<li>Still mising some points.
 +
<li>Update functionality and system hierarchy.
 +
<li>Update IBD and add description.
 +
<li>Add figures and unused functions.
 +
<li>Merge corridor challenges explanations.

Current revision as of 15:28, 20 June 2017

Contents

Meeting 1 (12-5-2017)

Reviewed initial concept design with tutor.

  • Team needs to think more on software architecture (components, functions, flows).
  • Current architecture diagrams are not explicit enough.
  • Agreed to meet next week to discuss findings and prepare first presentation.

    Meeting 2 (16-05-2017)

    Discussed initial design and prepared presentation.

  • Presentation will be made.
  • Information flows need to be specified.
  • Code the corridor challenge in pairs (Sjoerd & Bart, Rodrigo & Kagan, Angel & Torben).

    Meeting 3 (19-05-2017)

    Discussed corridor challenge progress.

  • Focus on framework, easier to discuss if we are more specific.
  • Think carefully on what we need/want to test (some parameters) Simulation vs Real.
  • Code the corridor challenge separately, in pairs Sjoerd & Bart (world model) , Rodrigo & Kagan (Motion Controller), Angel & Torben (Action Manager).

    Meeting 4 (29-05-2017)

    Define testing plan.

  • Tuesday, Turning 90° degrees with odometry, turn based on the laserdata, collect realtime data for worldmodel tests and tuning potential field (Angel, Sjoerd, Kagan, Bart) .
  • Wednesday, Everything not done on Tuesday, smoothing turnings, potential field with set point (Kagan, Rodrigo, Bart)
  • Use pythagoras, check for corner detection.

    Meeting 5 (02-06-2017)

    Discussed corridor challenge result and progress.

  • Things that go well (Potential field, detecting the gap and stopping in the right spot).
  • Things that didn't go well (Didn't stop after turning, due to time delays).
  • We are not using time delay anymore, we go back to odometry data.
  • First advance for Pledge algorithm (PICO got into a loop).
  • There's a problem while sticking to a wall when moving.

    Meeting 6 (12-06-2017)

    Discussed final presentation and maze challenge.

  • Tutor had no complaints about final presentation.
  • New function for align PICO and stay parallel to walls.
  • Integration of all functionalities tested, except for left and right doors (To test on tuesday).

    Meeting 7 (20-06-2017)

    Discuss the Wiki page.

  • Still mising some points.
  • Update functionality and system hierarchy.
  • Update IBD and add description.
  • Add figures and unused functions.
  • Merge corridor challenges explanations.
  • Personal tools