Hey people of FIRST,
I’ve been thinking recently about my team’s current drive team configuration (Pilot, Co-Pilot, Coach, Human Player) and was wondering if that matched up to what other teams had? I’m curious to know if anyone has thought of having 2 coaches and only one pilot, rather than pilot/co-pilot?
This year, our “second driver” was much more of a coach than a driver. Most teams used their second drivers to control kick settings, however our robot did not need such adjustment.
The second driver’s job was to keep track of the balls in the zone we were in. It proved very helpful when balls were just out of site behind the tower, bump, or up against the alliance wall. The human players acted as “coaches” sometimes too. It became very helpful for a front zone robot driver to have help from the human players. The human players were a “spotter” and could tell you if the ball went in the goal, or if you were lined up correctly.
It’s dependent on the game really. If the robot only has to do about 1 or 2 tasks then I would suggest one driver and then 2 coaches premise. Although, most likely there are more tasks that need to be done that another person driving is needed. Personally, I would love to have been able to just do all of the driving/manipulation last year but there were too many commands that needed the assistance of a second driver. If you are going to have two make sure that they can communicate efficiently. That was one thing that we had a decent grasp on last year but there were some miscommunications as well.
I think the usual drive-team combination for an FRC team is similar to what your team uses.
In my opinion, there is generally too much for one person to control on a single FRC robot effectively, requiring a second driver/Operator/Co-Pilot.*
On Team RUSH we typically give one driver complete control over the chassis of the robot and nothing else, and everything else is given to the other driver. The exceptions to this being that in 2009 we attempted to give the chassis driver control over the moonrock intake rollers, which was generally unsuccessful and just easier for the “co-pilot” to operate.
Also, as the posters above mentioned, having a human player that can be aware of things going on on the field and communicate them to your coach or drive team can be a great strength.
*The only exceptions to this that comes to mind are from 2008 where several teams built robots to do laps with few, to no, additional mechanisms for other tasks, the most prominent of these teams being world-champs team 148.
We usually have it setup so that the first driver only controls the wheels (that’s more than enough for one person to handle) and the second driver controls anything else in the year’s game (last year it was kicking, kicker power, hanging control, and a switch to invert the drive controls (It was the only way we could go under the tower for some reason)
The Game Rules define how many may be on the drive team, whether a human player(s) is needed or allowed and how many, whether a coach(s) is needed or allowed and how many, etc. We do not know yet what the 2011 drive team may consist of. We find a robot’s control scheme pretty much determines whether or not a co-pilot is needed.
We have always done it where the Pilot controls the base of the robot. The co-pilot controls everything else. The human player depending on its role in the game it can be another set of eyes. The coach is always a set of eyes to watch out for the driver can’t see.
Our configuration - We have ALWAYS arranged our drive team like this:
Driver - Drives the robot, has control of the drivetrain only. This year, that included the actual drive motors, shifting, and movement of the arm for bump-crossing ONLY (all other arm movement was controlled by the operator). He/she generally has either two joysticks or a Logitech gamepad (not xbox - the xbox joysticks are not symetrical and do not work as well for a tank drive). He/She always looks at the robot only, focusing their attention on where they want the robot to go, what opponents are in their way, and how to get there. This year, they could pick which goal they wanted to go for based on which one was more open.
Operator - Handles the rest of the robot. They generally have a custom-made control box built for that robot, and might have another joystick also (for example, our 2006 robot had manual targeting on the joystick, while 2007 had slide/manual overrides on the joystick). He/She takes orders from the driver on mechanism operation, and adjusts parameters (such as kick distance) on their own.
human player - They generally operate “autonomously”, performing their tasks without interacting with the other drive team members too much. This year, the two goal human players would scream that the ball was in if the driver was pushing the ball up to the goal, so the driver knew to go away and get another.
Coach - This is always Jim Zondag. He looks at the entire field, and decides what the best plan of action is. He usually shouts orders (such as “GO GET THOSE BALLS ON YOUR LEFT AND SCORE THEM THEN GO HANG”). He does not care how the driver does it, just that the driver gets the balls and scores them as fast as possible.
After practice and events, they generally get very very good at communicating with one another. They know what the robot can do, how to perform their action fast, etc. I recently talked to our 2007 driver (Woody), who said that by the end of the season, him and Dave (the operator) knew how to drive and move the arm so that the whole robot would lean forward on the front two wheels and score at the same time, in one smooth motion.
Another important and related topic is how to choose the driver and operator. For us, we pick the driver and operator as a pair who work together well, and are dedicated to the team. They stay the latest on late nights, come every day, and generally try to hold leadership positions. IF the human player requires a certain skill (such as 2009), that will be a deciding factor. If the human player requires little skill (2008 robocoach, 2010, etc.) then the human player will usually be the one who shows promise as a future driver.
I am not sure if this was our flaw that led to our mediocre performance, but we switched our drivers almost every round. Our mentor wanted other people to try it. I was the “coach” for most of the rounds but I just stood there looking like an idiot (mostly because it was my first year, I had no idea what to do… I still have no idea) And we switched the human players around too. All the controls were mapped to one joystick, like the moving, kicking, turning on and off components. So we never needed an “operator” I think that defeats the purpose. You never see video games where 2 people control the same character.
Im planning on so that we only need to send a human player and a driver. Yes I am confident that I can make it fully autonomous; the driver is there to manually override the robot if it goes haywire.
I think you’re going to want a coach. (We always use a student as a coach.) One of the roles of the coach is to coordinate strategy with the other teams in your alliance. This will be difficult for the driver and human player to do during the match.
Fully autonomous is pretty high goal - you may want to be sure you have full manual control first before going after full autonomous.
Not good. Not Not Not Not Not good.
The more a driver drives, the more practice they get. Our current driver and operator (Matt and Kitty) have driven in aprox. 150 matches, 130 of them together. They still have one FRC season left before graduating. They are very very good at operating the machine, and know how to communicate with each other. If your drive team changes in every match, they will not have the experience of a drive team of another team who doesn’t change every match, so even with a better machine, the other team will still win.
There is a universal trend that shows that teams get better as they play at more events. While they make changes to the robots, most of this is because of driver practice.
Have you ever been to an off-season with mentor matches? Teams tend to perform worse when the mentors drive. At the MARC, we had an alumni driver (and he drove for two years in FRC), and our coach (Jim) as our driver and operator. They lost early on. With our regular operator and coach (and a different driver, who had about 2 hours of practice time) we won the event.
Second the not good on switching every round.
Places you can get away with it: practice field, practice matches.
Places you could if (and that’s a big if) you had 2-3 very good teams: Friday qual matches.
Places you can’t: Saturday qual matches.
Places you really can’t: Eliminations, if you’re in them.
I think back to my time on 330. We really only had a couple of drive teams, and one coach. There were backups, but the primary drive team got most of the practice time. When we had a good, solid drive team, the human player switched every year, but the drivers never did. Between 2005 and 2008, we had one event where we did not make eliminations. One. The 2006 Championship. We had one drive team.
Regarding only having one person control the robot: It depends on the robot. But for most robots, the driver would need 3 hands with extra fingers on each hand to do it right. There are exceptions to this, like the Brave Little Toasters that drive around, get in the way, and otherwise make themselves annoying to the offense robots, but as a general thing there are too many motions to control. The only thing worse than having only one person control the robot is to have two people who don’t work together very well. The trick is to find a pair of drivers who work well together such that they are almost extensions of each other when it comes to driving the robot.
The reason video games never have two people controlling the same character is because everyone wants to do everything. FIRST robots always have between 2 and 3. One is the programmer, another is the driver, and the third is the operator.
A better parallel might be a military or commercial aircraft. The only planes in those categories to have one pilot are some fighters. Most planes (including alternate models of most of those fighters) have a flight crew of at least 2 plus an autopilot. One is the pilot, one does stuff other than actually fly the plane, like dial in a target to the weapons system or check the weather ahead. That’s what you need to be thinking like for the drive team.
Extremely well stated.
By far, the best thing your team can do to improve your robots performance is get the drivers as much practice as possible! Whether that means both of them working together, or if you only have one person driving all of the robot.
In Breakaway, our mentor decided that it would be a good idea to have one person drive every part of the robot at once, and have the operator be more like a spotter, telling the driver where the balls were if the driver could not see them.
We did not finish our robot early enough to have a lot of practice for the operator and I. We did not do extremely well at our first competition, and this is mainly because the drivers had extremely limited practice.
Our best performance was in Ramp Riot, the post season event, because I had plenty of time behind the controls and knew how the robot drove and worked extremely well.
So words of wisdom:
Hope this helps!
Our drives team consist of a driver an operator (who handles whatever scoring apparatus the robot uses) a human player and a coach (which has always been and most like always will be a student with us). We have a backup driver and a backup human player.
Defeats what purpose? What purpose are you referring to?
Anyway, every competitive team has a consistent drive team, and a large majority of them have two drivers.
And to be blunt, if you haven’t figured out how to make a drive team successful, what makes you think your team will be able to pull a fully autonomous robot? Where do you get all this confidence from when you haven’t been able to pull anything resembling AI together in the preseason? Before you get artificial intelligence down, you should be able to figure out how normal “human intelligent” drivers work.
Beside being able to operate the controls, the drivers need to be able to work well together. If the driver goes “left” as the arm operator tries to pick up “right” - not a recipe for success.
If you should decide to switch people around, be sure everyone who is partcipating knows the rules. I don’t know how may stupid penalties have been called because there’s someone new out there that doesn’t know what is going on and does something illegal - blatantly illegal, something that drivers and HP’s who were around from Match 1 would know enough not to try.
i’ll be the brutally honest one who says:
Every match your partners are relying on you to do your very best, doing anything less is a disservice to them. Keeping a consistent driver is absolutely essential to good performance – each match that goes by the driver will improve exponentially.
I love how if someone talks about intentionally “throwing a match” by not doing their best, the entire community flips out… but if someone talks about changing out their drivers…
-John
Most years, we have a driver and operator. In “arm” games like 2007, these roles have similar control demands and require excellent coordination between the two. This changes from game to game for us though, for instance 2009 was less operator intensive. In 2010, we really didn’t use our operator much during the season at all. (In retrospect, we should have given them another official job.) The extra functions were relatively easy for our driver to control, but part of it was just lack of driver-operator coordination. Once we found an operator that fit in well, it got easier.
I’ll vouch for the practice-practice-practice stances, though. Having been a
mechanical and pit captain, captain-manager, operator and coach, drive team practice is the most efficient way to improve your performance. Don’t change your drive team!
Also, Karthik says, ‘the role of the field coach cannot be overstated’. If you don’t know what you’re doing, find out! Ask fellow coaches (especially if they’re from 71, 1114, 148*, 217, 45, 33…) and ask your drive team. Read and practice.
*EDIT: yes, listen to that guy. ^
Oh yea I remember, I won’t be using ANY of the code I am writing before kick off on the robot. Im scrapping everything I do pre kickoff and doing it all over during the build. Never knew people took this so seriously.
I’m saying if in the past 6 months, with all your practice you didn’t get anything close to what you wanted done, why do you thin with only 6 weeks you will be able to accomplish this?
Do you really want to jeopardize your own success? That’s your right to do so. But what about the rest of the team, too? Are you going to put your own demands and delusions above them?
What about your alliance partners? Do you really think they want to be your guinea pigs for autonomy tests? There’s a match going on and they probably want to win it.