Hybrid Challenge-No Robot Left Behind

[Doug Leppard]

Every year we put in delays controlled by switches. This year we will have 0.0, 0.5, 1, 2 second delays. We do hope and work with our partners in this area.

And I agree it is a crowded area this year and collisions will happen if all take off at same time.

[thefro526]

Tdorthebnator and I were talking about this the other day. This Auto/Hybrid mode is easier than 06's (if you were a dumper or defense). I'm gonna go harass our programmer on monday about this and see if he wants to get started on it. If your using easy c the code can be written in about 30 seconds so i think most teams are gonna have some kind of hybrid action

[Doug Leppard]

Originally Posted by thefro526

Tdorthebnator and I were talking about this the other day. This Auto/Hybrid mode is easier than 06's (if you were a dumper or defense). I'm gonna go harass our programmer on monday about this and see if he wants to get started on it. If your using easy c the code can be written in about 30 seconds so i think most teams are gonna have some kind of hybrid action

In EasyC Pro which we use it is easy to get the four points.

Once Hybrid startrts, wait a bit to let people clear, drive forward for XX seconds calculate how long it takes to cross line then stop. Most robots will not go straight but the distance is so close that it will be ok. You can even aim different to adjust.

For 8 points: Drive straight until clear the lane, again guess on time and adjust, do a 90 left again guess the time it takes and adjust and then drive straight again for xx seconds. You should be able to get 8 points with practice a certain percentage of the time, but should always get 4 points.

To get various amounts of wait, have a jumper on one of your digital input ports and read that. You can have several for several times to wait.

Ours is a bit more ambitious using encoders and other sensors.

[thefro526]

Last year I did a go straight and turn around the rack with dead reckoning. And I could show that to any on who wants to see it. It wasn't a 90 degree turn though it was more like 45 or so to match the angle of the rack. We used timers though and if we forgot to restart the bot it did some funny stuff.

[TheNotoriousKid]

Well, if it hasn't already been shown evident..."Team 25 Graciously Accepts"...the challenge.

[Jim E]

M.A.R.S. Team 1523 Graciously accepts.

I guess will have to nudge our programmer to write some generic short autonomous code to get them across the Finish line if they cannot do it themselves.

[ShotgunNinja]

Well, my team has shown that they wanted to go with the IR sensor controller, maybe using a TV/VCR remote. But I am in charge of the programming, and I was wondering...

Would something like this (Using ROBOTC) be usable?

task Autonomous()  { // Note: Just pseudocode, but whatever
while (bIfiAutonomousMode) /* Just in case... */
{
   switch(GetIRInputState()) /* Get the state of the IR sensor, in a different subroutine */
   {
      case Button1:
         Action1; // Something like "Move forward, stop"...
      case Button2:
         Action2; // Maybe "Turn Left"
      case Button3:
         Action3; // Maybe "Turn Right"
      case Button4:
         Action4; /* Something along the lines of: "Grab ball off overpass, however possible" */
      default:
         TakeNoAction; // Just "Stop Motors" or something
   }
   BetweenActions; // Something like "Stop Motors, Wait XX MS"
}
EndingAction; /* Play a sound or something. Reset motors for Human Control. Etc. */
}

Any comments?

[timville]

The problem with the code above is that the IR sensor does not remain as a "1" state when you hold the button. I believe the IR sensor will pulse the circut open and close every 100ms.

See my reply to your other thread with sample code on how to use the IR sensor with ROBOTC. http://www.chiefdelphi.com/forums/sh...ad.php?t=63004

[GGCO]

Hmm. The 5 second delay wastes a lot of time. My team wants to get the ball down, along with going around the track at least half-way.

[Tottanka]

Originally Posted by timville

The problem with the code above is that the IR sensor does not remain as a "1" state when you hold the button. I believe the IR sensor will pulse the circut open and close every 100ms.

See my reply to your other thread with sample code on how to use the IR sensor with ROBOTC. http://www.chiefdelphi.com/forums/sh...ad.php?t=63004

yes.
what you need to do looks more like this:

If (RC_dig_in04)
     c=0;
If (RC_dig_in01)
     c=1;
If (RC_dig_in02)
     c=2;
If (RC_dig_in03)
     c=3;
if (c==1)
   pwm01=254;
if (c==2)
   Pwm02=0;
if (c==3)
   Pwm05=160;

[meaubry]

So, now that we have everyone capable of moving, and assuming most teams will still be potentially subjected/limited by signal overload, interference, blocking or stepping on each other - what can we do within the community, that will help reduce or eliminate this problem?

Signal guides (tubes) are helpful - if the signal gets there, but is there something that can be included in the code to help everyone?

Or, is this something that we must reinforce with the Robocoaches that are in control of the IR controls?

Is there a certain IR controller that works best for what we are trying to do? Signals are supposed to be sent as one time event - ideally, not continous (if the button is accidentally held) in the excitement of being a robo-coach?

Is there something that teams can do to there IR that would reduce the "holding down the button" problem?

No robot left behind would be alot more meaningful, in the context of FIRST coopertition, if it could also help eradicate the problem of multiple signals at the same time - that many are concerned about. I just don't know if it's possible.

Obviously, in the spirit of cooperation among alliance members, agreements as to who goes first, second, third, or not at all, can be done prior to the match. And, if that is all that is needed - so be it.

Just wondered if there was something that might be able to be done with the IR itself, or if certain IR's eliminate that problem and we could suggest using that one instead of the kind that cause problems. Maybe, a strongly suggested/recommended IR list, and a strongly NOT recommended IR list.

Mike Aubry

[Doug Leppard]

Originally Posted by meaubry

Is there a certain IR controller that works best for what we are trying to do? Signals are supposed to be sent as one time event - ideally, not continous (if the button is accidentally held) in the excitement of being a robo-coach?
Mike Aubry

We want a continuous push of button to make sure the robot got the signal and that is legal. We will only be sending the signal the 1st second and then it will be turned off by that time the robot should know what ball to go after.

[meaubry]

Doug,
I understand the concern of assuring that the signal was recieved. Pushing and holding for a second or 2 isn't what I was referring to.

I was thinking that the robocoach may hold the button for 10 or all 15 seconds - simply, because they got excited and forgot to let up - or - they panic when the robot doesn't do what it was suppose to do, and continue to push the same button rapidly, over and over.

Just trying to better understand the dynamics of what might happen at the start of a match, and wondering if there might be something that could be more or less done to eliminate problems.

Like I said before, it's awesome that everyone is going to try to make sure that any team that wants to be able to try to move in the 15 second Hybrid/Autonomous time period has help available. And I'm thinking that with a little more thought, a few other issues could be addressed.

Such as -
If the code is simple to drive forward or go forward -turn left, it doesn't need to be IR controlled. That alone would eliminate alot or unnecessary IR signal traffic.
Also, simple delays in starting or ramping up to full speed could help avoid traffic jams - if each team also knows what speed they run at.

Mike

[Doug Leppard]

Thanks Mike. I appreciate your input.

What we are doing is telling our bot what balls to hit and then it does the rest. We figure we have 1-2 seconds to do this and then it is out of there and then there is no more reason to push buttons.

Interesting thing about the IR receiver according to my software guy over this is that once it receives a signal it transmits for 100ms waits 50ms then you can do another signal. Also holding down the IR same button it will acknowledge that button but only once, that button must be released to get another signal on it.

Also for our use, four buttons is barely adequate, six would be better.

[tseres]

if anyone is at the Waterloo and Greater Toronto Regional, come ask for me at the 1565 pit. i have code available that can help ANY team have a near-perfect autonomous. and no, i'm not bluffing

any questions please PM me and/or i'll see you in the pits!!!