Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

I really liked the fact that this years autonomous forced teams to use sensors. I think one of the biggest issues with firsts "autonomous" setup is that teams don't have to use sensors. I'm sure many teams in the first two years of the autonomous mode used no sensor input. My team included. But in the true terms of autonomous mode the robots should be using sensors to track their progress, not simply timing.

With that said, I like having an autonomous mode where the goal isn't in the same place every time. But, I still think theres a need for the game to have some autonomous possibilities not using sensors. If your a rookie team, its a feat to have your robot drive straight, let alone drive into a stack of boxes around a corner.

I think something like what was in 2004 where you just had to hit something that interacted with the field was good fun, maybe having it at different lengths or spots along the wall? Or just in different places on the field. If this years autonomous was that all you had to do was get the vision tetra off the ground to get those 2 caps, i think more teams would have gotten the job done.

A big part of the autonomous just the game itself. This years game didn't have many different things for robots to do, in turn there really weren't allot of possibilities for what robots could do. When the game is designed, remember the creativity of the FIRST strategist. You may not think of every way a team can do autonomous, but if theres a variety of field elements or scoring objects you open up a plethora of opportunities for teams in autonomous.

I would love to see a gyro make its way back into the kit. I think its a very useful tool for autonomous. Just look at Wildstang in 2004, their system was amazing for the first year of autonomous, and the roots of autonomous in 2001 when teams would use the gyro to autonomously balance the ramp. I would also like to see Infrared Rangers, like the ones from sharp. These are cheap enough that teams can buy them, if they know how to use them. I think having something built into the IFI controller to help with say using a sharp gp2d12 ranger would be great for some teams.

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

I would like to see Back EMF sensing implemented into motor control as mentioned by Dr. Joe in this thread

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

Alright, this is a rather big theory. I have no idea whether it'd work or not, but I figure that the worst-case scenario is that I use up two minutes of my life.

I believe it's a fair assumption to say that the sweeping majority of FIRST teams hate the 2004+ joysticks. Some hate them with a passion, while others (such as yours truly, who never had a chance to deal with Flightsticks other than feeling them in Manchester at Kickoff this year where they were used on the demo robots) just grin and bear it.

I've had a few experiences with the Vex system (I just got my kit today, and will be working on it over summer), and I've always felt the controller to be nice, solid, and good at what it does (controlling a robot).

What I'm thinking involves taking the Vex controller (or a non-Vex version, or whatever works the best), and tethering it to a revised version of the OI. Then add one or two legacy ports for folks who want to wire up the custom stuff, or who still won't let go of those Flightsticks.

If it happens, I sure won't mind.

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

1. let teams have new technology and software early - sometime in September would be good. Most teams do not have the time nor resources to implement a new sensor during build season. Our resources are limited so some direction in what to look at would be good.
2. Increase the electronics/ sensor budget --- There is some really cool stuff out there to use if we could get it in the budget.
this years electronics was the best offerd yet - keep up the good work

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

I've seen a lot of comments in this thread suggesting that there be multiple tasks in autonomous mode, with varying point scores. We had that (after a fashion) this year, and I think the variety of possible autonomous tasks should be encouraged. That being said, an increase in the reward for autonomous would encourage teams to develop autonomous modes. Preferably, establish it in such a way that a team that had no autonomous mode would be at a large (but not insurmountable) disadvantage in a competition.

I'm a supporter of keeping autonomous mode at the beginning because I think that dead reckoning should be preserved as a viable autonomous mode. Although it ought to be the least desirable of autonomous modes (and the reward allotment should reflect that), it should at least be possible, as many teams do not have time to construct a more complex autonomous mode. Dead reckoning is rather difficult if you're initiating autonomous mode in the middle or at the end, and these outcomes would most likely only serve to reduce the number of teams that had a functional autonomous mode.

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

I agree. I saw several teams having lots of trim/centering problems causing unsafe conditions in the pits. For example, they would turn the robot on and it would just start moving for no reason other than the joystick was not centering well. To me, a (nearly) runaway robot is not safe.

Personally, I have not come across a better joystick for robots than the old CH Flightsticks. I know they are discontinued, but if we could get CH to make a special production run just for FIRST, or if FIRST could some how get/buy the plans for it and have another company make them, that would be totally awesome. Or maybe somewhere in the deepest darkest corner of the CH fabrication facility they still have all the tooling laying around for these joysticks, and even if they don't want to make any more, maybe FIRST could purchase the tooling so then another company could do it for FIRST.

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

Although I am a fan of the short autonomous period at the beginning of the game, there is no reason why one could not introduce a game element that requires autonomous operation of the robot, while human players are at the controls.

As an example, consider a panel in the center of the field with colored game pieces positioned on it. The game pieces for a given alliance would be on the side of the panel opposite them, concealing the location of the game pieces from driver view. The panel might be up off the floor on posts, so that drivers can see the location of their robot, but can't see the location of an arm relative to game pieces to be picked off the panel by the robot. Teams that do not code automatic seeking and grabbing of game pieces could perhaps use their "human player" on the opposite side of the field, giving directions with regard to where to position the robot, and the arm, in order to grab game pieces, a new level of team work... Teams that code a vision system to find and grab a game piece will be able to operate more automatically, but still might use a human player to direct rough positioning of the robot.

The opposing alliance would get to watch your robot collect game pieces, but might not be allowed to interfere. The crowd would also get to watch the robot grab game pieces automatically, making this aspect of the game interesting to watch.

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

Force to use sensors? Wait a minute! I can think of several auto modes that most certainly did NOT use sensors, including the most popular ("Mars Rock"). Seriously, it was quite possible to knock down the hanging tetra (or even capture it) with no sensor except the timer in the computer (if it had one).

If FIRST has one autonomous task that you don't need sensors for, but sensors give a team a big advantage, and another that cannot be done without sensors, I would be happy. The rookies can do something in auto mode without a lot of extra work, the second year and older teams have a challenge that they can do easily with or without sensors, and the teams that love a challenge or have a huge programming staff get a challenge.

Sensor-wise, more (complex) is not always better. Beachbots tried to get their camera working through a preprocessor. Six weeks later, the camera was talking to the preprocessor, but the "Rabbit" (preprocessor) was not talking to the RC.

If FIRST gives teams a complex sensor (camera, IR sensor, anything on that order), they should give it to teams as soon as they register and say "Here. This is a ________. Here are some activities that will prepare you for the use of this _________. Have fun."

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

I was thinking about inter-robot communication, but first dismissed it as a good idea, but too hard to pull off. I've been thinking about it again, and thought about some kind of a standard. Each robot would send out a signal from a specified list. Instead of each time sending out different signals, you would have transmissions such as:
RED1: HEAD_TWDS_AUTOLOAD1
RED2: HEAD_TWDS_AUTOLOAD2
*RED3's programming tells it not to go to the auto loaders, as they are full.*
RED3:WAITING
RED1: AT_AUTOLOAD1
RED2: AT_AUTOLOAD2
RED1: DEPART_AUTOLOAD1
*RED3's programming tells it to go to autoload1, now that it's open.*
RED3: HEAD_TWDS_AUTOLOAD1

These are just simple commands, maybe more advanced ones can be made available (such as a robots coordinates on the field).
I think having that would be real cool, beyond real cool. As cool as bagels cool, and bagels are pretty cool 8)

Towards the technical ends of it, I don't think setting it up as an ad hoc network would work. Rather, have the robots each send the messages to a centralized server of some sort, which then sends the appropriate messages to the appropriate robots (EG: Doesn't send RED1's messages to RED1 or BLUE*, only red team)


In conclusion, Dave, if you give us inter-robot communication, I personally will deliver you 3 dozen Krispy Kremes. You know you can't beat that offer ;)

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

Well, since pneumatics is a technology...can we get a break on the prohibition on non-kit hose? I'm not an engineer, but it just seems silly to be restricted to one big piece of hose when using other hoses, unless someone can point out something to the contrary, can do the exact same thing the same way (read: move air from point A to point B at X pressure without exploding into 1,293 pieces).

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

Hello!
Here's a bit of a messed up autonomous idea- force robots to work together to get a task done in such a way that they would have to communicate with each other. I have no idea how it could be implimented, but as FIRST is all about achieveing a common goal through communication, it might make things interesting.

Sparks

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

I like most of sanddrag's ideas but especially the above. Auto mode must have an incentive, either bonus or multiplier scoring or presetting the field for the alliance. If the camera or any other auto mode device is going to be included, put it in the hands of teams in the fall, no later than November 1.
What is really needed is a method of letting teams calculate (or detect) absolute positioning on the field. Auto mode needs to be a short distance goal with several tasks or a whole field exercise with absolute positioning info. Without these types of improvements, auto mode will be attempted by less than half the teams.

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

I have a few thoughts about auto mode, and the use of a challenging playing field to drive innovative drive trains.

Initially I was somewhat skeptical of an auto mode, having been a driver for 2 of my 4 years as a student in FIRST, none of which having an autonomous mode. For me, part of the inspiration has been and will be being able to go out on the field and drive the machine you built. However, watching over the last few years, it is pretty amazing to watch the robots do the tasks on their own. As a result I now enjoy auto mode for the most part. However I think it should stay at the start of the match. I would much rather be in control of my own destiny at the end of a match then letting the robot decide, something they may be good at, but they are bound to a certain number of options unlike a human.
In regards to the scoring of auto mode, I like having a tiered setup. The harder something is, the more you should be rewarded for it. However, I do not think the auto mode should be made worth so much as to severely impact the outcome of a match. Not all teams are equal, especially in the department of autonomous robots. I would much rather see smaller points for smaller actions that all teams can get at.

In regards to the playing field, I think each year should be and has been different. Thinking back the last several years, back to 1999 actually, FIRST has done an excellent job of varying the levels of the playing field. Although, I would say that some of the more interesting years had some sort of obstacle or playing field level change in them. My favorite field and game would still have to be 2000. Lots of offensive and defensive options, an exciting and simple scoring system, a "finish line" i.e. hanging on the bar, and the field was not all flat. But a varied field to challenge drive trains is something I would like to see more of, something to drive change.

PS. I read recently that corn was used in 1992 on the field. Maybe we need to go to an extreme of say, a beech or muddy playing surface? Probably not something so messy but it would be oh so fun :).

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

If you do decide to go down the route of a more interactive field, maybe something that could be sent to the robot via the communications system that would tell the robot what state the field is in.

EG: In the 2004 game, at a certain point in the game balls were dropped from overhead (not too sure of the specifics, I wasn't around then :p). When that happened, data would be sent to the robot saying it happened so that the robot could act accordingly.

Re: [Official 2006 Game Design] Autonomy And Other Technology Discussions
 

i was thinking along the lines of giving robots simple problem solving problems.