Announcement

Collapse
No announcement yet.

Off-season training

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Re: Off-season training

    Now that I have two light sensors, next up in my bag-o-tools is squaring to a line.

    The new sensor sensitivity is badly mismatched from the original one, which is many years old. That's a good thing, since it provides a good test case for the tribulations of sensor calibration.
    FIRST Tech Challenge Judge: 2010, Referee: 2017
    FIRST LEGO League Mentor, Instructor, and/or Referee/Head Referee since 2011
    FIRST Robotics Competition judge (Chairman's Award): 2014
    Dean says I'm an "Oompa Loompa of Science"

    Comment


    • #17
      Re: Off-season training

      Originally posted by Tom Mosher View Post
      Now that I have two light sensors, next up in my bag-o-tools is squaring to a line.

      The new sensor sensitivity is badly mismatched from the original one, which is many years old. That's a good thing, since it provides a good test case for the tribulations of sensor calibration.
      The mismatch occurs for other reasons. The distance of light sensor off the mat, or light source direction.

      The light intervention can be corrected by shielding the sensor. To fix the mismatch, you may want to investigate creating your own light sensor calibration / reading for each sensor using the raw value.
      Tony A.
      Los Angeles FLL Region OP
      www.la-fll.org
      https://www.facebook.com/LosAngelesFLL/

      Comment


      • #18
        Re: Off-season training

        Originally posted by LOLComets View Post
        The mismatch occurs for other reasons. The distance of light sensor off the mat, or light source direction.

        The light intervention can be corrected by shielding the sensor. To fix the mismatch, you may want to investigate creating your own light sensor calibration / reading for each sensor using the raw value.
        I'm pretty sure Tom has looked at all the obvious sources. I know that none of my light sensors match perfectly and a couple are off by quite a bit (> 10%). Looking at the construction of the sensor I think physical alighment of the photo transistor is the most likely cause. I also have one sensor where the LED is dim.

        Comment


        • #19
          Re: Off-season training

          Thanks for the tips. I'm not stuck for a solution, I was just commenting that I'm happy to finally have the stuff necessary to investigate the phenomenon.

          I'm managing to stay about two weeks ahead of the team's off-season activities, so when they have questions I'm able to guide them toward their own answers, using the tools they have available. Graphical programming languages that look like bricks are not my strong suit.

          Is there a way to get the raw sensor readings?
          FIRST Tech Challenge Judge: 2010, Referee: 2017
          FIRST LEGO League Mentor, Instructor, and/or Referee/Head Referee since 2011
          FIRST Robotics Competition judge (Chairman's Award): 2014
          Dean says I'm an "Oompa Loompa of Science"

          Comment


          • #20
            Re: Off-season training

            Originally posted by Tom Mosher View Post
            Thanks for the tips. I'm not stuck for a solution, I was just commenting that I'm happy to finally have the stuff necessary to investigate the phenomenon.

            I'm managing to stay about two weeks ahead of the team's off-season activities, so when they have questions I'm able to guide them toward their own answers, using the tools they have available. Graphical programming languages that look like bricks are not my strong suit.

            Is there a way to get the raw sensor readings?
            For light sensors? Use the "Raw Value" terminal. For a color sensor you are stuck with the intensity value which comes out of the "Detected Color" terminal when the sensor is in Light Sensor mode. Looking at "Raw Value", the numbers get smaller when the light is brighter.

            Comment


            • #21
              Re: Off-season training

              Originally posted by Dean Hystad View Post
              For light sensors? Use the "Raw Value" terminal. For a color sensor you are stuck with the intensity value which comes out of the "Detected Color" terminal when the sensor is in Light Sensor mode. Looking at "Raw Value", the numbers get smaller when the light is brighter.
              Thanks Dean, I totally missed the Raw Value terminal on the Light sensor. Sadly, my NXT at home is too old to support the software required for the Color sensor.
              FIRST Tech Challenge Judge: 2010, Referee: 2017
              FIRST LEGO League Mentor, Instructor, and/or Referee/Head Referee since 2011
              FIRST Robotics Competition judge (Chairman's Award): 2014
              Dean says I'm an "Oompa Loompa of Science"

              Comment


              • #22
                Re: Off-season training

                Originally posted by Tom Mosher View Post
                Thanks Dean, I totally missed the Raw Value terminal on the Light sensor. Sadly, my NXT at home is too old to support the software required for the Color sensor.
                Not true. I was part of the pre-releast testing and my beta NXT is running the newest version of commercial NXT software with color sensor support.

                Comment


                • #23
                  Re: Off-season training

                  Originally posted by Dean Hystad View Post
                  Not true. I was part of the pre-releast testing and my beta NXT is running the newest version of commercial NXT software with color sensor support.
                  Sorry, I should have been clearer that the issue is the Mindstorms software running on my computer, not the NXT brick. There's no free upgrade for the PC software I own.
                  FIRST Tech Challenge Judge: 2010, Referee: 2017
                  FIRST LEGO League Mentor, Instructor, and/or Referee/Head Referee since 2011
                  FIRST Robotics Competition judge (Chairman's Award): 2014
                  Dean says I'm an "Oompa Loompa of Science"

                  Comment


                  • #24
                    Re: Off-season training

                    Originally posted by Tom Mosher View Post
                    Sorry, I should have been clearer that the issue is the Mindstorms software running on my computer, not the NXT brick. There's no free upgrade for the PC software I own.
                    True, LEGO does not provide the commercial version of the Mindstorms NXT 2.0 software separately. However, they do make it available for download at their site. Technically, they make the software available for people who have lost their disks. I'm really at a loss to explain why LEGO can't make physical upgrade or replacement disks available for a nominal fee. That's what they did for the 1.0 to 1.1 upgrade.

                    The version of the software sold by LEGO Education has some additional features, including datalogging capabilities and tutorials. I guess that's why they charge $79 for it.
                    Last edited by timdavid; 03-17-2012, 06:37 PM.

                    Comment


                    • #25
                      Current activities

                      The last 6-week period of the school year has begun, and here's what the "lego robots" after-school activity group is up to.

                      First, the group has a new coach (the fourth one since last September). Spring is track season, so the third coach (a PE teacher) has passed the leadership to a math teacher (who was the volleyball coach), since that is a winter activity.

                      The focus is on adding PID controllers (without the I portion) to the team's range of comfortable tools. The math is simple, but it's not obvious how the robot does math until it's been experienced.

                      The first instance is a distance-following robot, with the ultrasonic sensor pointed toward the front. The PD control loop uses the MOVE block's Power input to maintain a fixed distance from a robot ahead of it. The students tested it by setting the robot on the floor, and having it drive toward their foot. Some robots stopped promptly at the predicted distance, some didn't stop at all. So the team learned about the influence of loop gain, and why it's important.

                      When you line the robots up on the floor, nose-to-tail, and start the first one on a straight drive, the others all launch in sequence, automatically. The chain rapidly breaks down though, since the robots aren't being steered.

                      Now, replace the distance sensor with a light sensor, adjust the parameters, and the result is a PD-based line tracker that uses the MOVE block's steering input. The concept of the use of the math blocks is identical.

                      Once each portion is working, they'll extract the distance PD and line tracker PDs into MyBlocks, then drop them inside a single LOOP block, with a single MOVE block controlled in both power and steering.

                      So, by the end of six weeks, I expect to see a train of tribots driving around a loop track, managing their speed to avoid crashing into each other.
                      Last edited by Tom Mosher; 04-13-2012, 04:06 PM. Reason: fixed typos
                      FIRST Tech Challenge Judge: 2010, Referee: 2017
                      FIRST LEGO League Mentor, Instructor, and/or Referee/Head Referee since 2011
                      FIRST Robotics Competition judge (Chairman's Award): 2014
                      Dean says I'm an "Oompa Loompa of Science"

                      Comment


                      • #26
                        Re: Current activities

                        Originally posted by Tom Mosher View Post
                        ...
                        So, by the end of six weeks, I expect to see a train of tribots driving around a loop track, managing their speed to avoid crashing into each other.
                        That would be very fun to watch. Sounds like the kids are on their way to careers designing autonomous cars.

                        Comment


                        • #27
                          Re: Current activities

                          Originally posted by timdavid View Post
                          That would be very fun to watch. Sounds like the kids are on their way to careers designing autonomous cars.
                          When they get it working, it could be video-worthy. Who knows, maybe driving aids will appear in Senior Solutions.
                          FIRST Tech Challenge Judge: 2010, Referee: 2017
                          FIRST LEGO League Mentor, Instructor, and/or Referee/Head Referee since 2011
                          FIRST Robotics Competition judge (Chairman's Award): 2014
                          Dean says I'm an "Oompa Loompa of Science"

                          Comment

                          Working...
                          X