Last year my team used 1 XBOX 360 controller to control the robot. We had
only 1 person actually operating the robot during the match. What are the pros
and cons of splitting up the driving and control to two people. IE having a
button board for controls and the controller for driving.
Driving by itself is difficult enough task. Splitting the controls allows the driver to concentrate on moving the robot, while the operator can manipulate other controls.
This year, we did not split our controls because we did not do fuel, nor gear pickup. The things we did do were tied so closely to movement (eg releasing the gear or climbing) that we felt it would be better for the driver to have control of them. In the case of climbing, we weren’t driving at the same time, so the driver could easily operate it as well.
A good reason is separation of concerns. Generally, the driver can take responsibility of things that are highly dependent on robot movement, e.g. picking up gears. Giving the second controller (the operator) responsibility to handle other things - like shooter or climber - frees up button slots on the driver’s controller to allow for more driving options. For example, in 2015, we had dedicated buttons to slow movement by 75% when we had a full stack of totes. If the driver was responsible for everything, there wouldn’t have been room on his controller for something like that.
Separating controls does mean you have to communicate between drive team members, but if that’s an issue, then it’s likely not the fault of the controls.
Our team has always had 1 robot driver at a time to simply as much as possible. If you decide to split robot control, both team members must practice weeks in advance in addition to many hours during the build season.
Pros of dual control:
One controls chassis and the other controls only scorer (shooter, arm, etc.). It makes it very easy for the the chassis driver to know where to be and they have a limited amount of buttons to press so with practice they can become very skilled. The same goes for the scorer controller, as long as they stay on the same page.
Cons:
It is a major time constraint that has to be accounted for before making the decision. When two drivers work side by side they will need to communicate and understand exactly where and when the other will preform an action. Sadly however due to the limited 6 week time frame, many robots aren’t completely finished until the 5th or even 6th week; leaving very little time for both to simultaneously practice with elements and the field.
– Greetings from Team 5557 BB-R8ERS
This is something that every team should consider while making the robot.
What I think is that every team should first start with the default option of having one (1) person control the Robot, and only if the robot needs too much control (too many subsystems, too many manual operations, etc.), should the robot control then be split up into two different jobs.
Team 2383 first experimented with the idea of 1 person controlling the robot in offseason 2017 as our offseason robot only had a ground pickup which swiveled up/down with a feeder on it, and a climber. We had good success with this, ending up as finalists in every single offseason attended.
Even with a few more mechanisms, it can be beneficial to have 1 person controlling the robot whenever possible as actions can be streamlined and requires less teamwork and chemistry. Time is short during FRC build seasons and I believe that if the robot design allows it, one person driving/operating a robot will always be better than two.
We find it easier to have 1 driver to get the robot to where it needs to be to score and an operator to do the scoring task with the two communicating during the match. To do this the drive team needs to be able to know each others thoughts in a way. In terms of control, I had joysticks(I was the driver. My operator had an xbox controller. The last two years we had two xbox controllers. This works best for us. You should see what you believe would work best for your team. I you feel only one driver is needed, run with it. if you think you can have two, run with it.
Perhaps learning about that specific robot may fall into this time constraint, but there is no need to train the drivers to work together only during build/competition season. Instead, have your drive team (ideally same year-to-year drive together at offseasons or practice with old bots. This will let them learn to work together without it having to be done during the limited window.
As far as I know, 2410 has used one driver in years past. We have one controlling robot functions, someone watching the camera feed, and our drive coach.
Somewhat unrelated, but another common option is to use another game controller for the controls. This can give a nicer UX with easier to handle and press buttons, as game controllers have a lot of thought put in to them by the designers about making them easy and nice to use.
Game controllers can also let your operator move around the driver station, say like in 2016 when some field elements could block your view, but with a controller you can move to the left or right to see around it.
Having said that, obviously there’s pros and cons for each - (e.g. game controllers are less customizable and could have less buttons that you’d like than a custom button board). It’s really what ever works best for your team.
Actually answering the question - An advantage of splitting up the controls is that it lets your driver focus on the sole task of driving, without having to worry about other things. Driving and controlling the robot well is really hard and requires a lot of attention, and distracting the driver with other things to worry about should be avoided.
Things that are simple and connected to driver actions - say, placing a gear after driving to the peg - aren’t that distracting, but some other actions - like maybe lifting an intake while driving so it doesn’t get damaged or dealing with a faulty mechanism - are quite distracting.
So as others have mentioned, it comes down to what works for your team and your robot’s complexity.
This is actually a really good question. 236 has always had 2 drivers, affectionately referred to as sticks and buttons. Like most of the other people here, the sticks have control over all driving and driving-related operations. Last year, the driver had control over the shifters, and the ball intake - that’s it. This allowed him to focus entirely on driving and working around other robots. On buttons, I controlled the intake deployment, gear manipulator, shooter, feeder, and climber. Giving me control over essentially all the mechanisms was extremely important.
A lot of the benefits are second nature - I didn’t even think of them until now. For example, in one of our early matches in Waterbury, we lost pressure and weren’t able to open the garage door. Because that was my job, and I could look at our dashboard, I was able to make the call and switch our match strategy on the spot. If our driver had that control, it’s likely we would have lost critical seconds.
In general, having multiple drivers decreases the amount of things each one has to think about - giving you quicker reaction times and more efficient performance at all times (assuming they communicate).
If you can automate your controls so that one driver can handle it all, that’s the way to go; your “second driver” can then be an extra set of eyes and ears watching the robot health and/or strategic locations on the field for the coach. However (as an example), if you can aim your shooter in two directions, don’t have an auto-aim function, and need to track a target while driving, split it up!
IIRC, we’ve only had one driver at a time since our third year. For some games the operator might handle the end game or certain functions, but then [s]he would also be driving at that time.
It depends on the game and strategy. This year, if you went gear-based only without any fuel capability, than 1 driver and 2 coaches would be the way to go. That way, there is less reliance on communication between the driver and operator, reducing time during cycles. The driver focus on the game, while both coaches could focus on gear count, score, where other bots are (and the fastest path to gear cycle), etc.
If you had a shooter (like we did), than a driver and operator would be the way to go. For us, I had general driving capabilities (including 2 crawl buttons) and gear drop-off, while our operator had the rest of the functions (including gear pickup, fuel pickup, close shot, auto shot, climbing etc.). This allowed me to purely focus on driving, defense, and gear cycles, while the operator could focus on the other aspects. While I would drive, he would focus on ball count, dropped gears, hoppers, and robots about to shoot fuel. Though this does rely on good communication, in our case, we communicated well due to lots of driver practice.
Our team used Logitech controllers for both driver and operator this year.
We took a hybrid apprach this year, and it worked really well for us. All of our subsystems have “semi-automatic” controls on the gamepad used by our driver, and as long as things are working normally he can drive, get gears, score, and climb all on his own.
Our operator was there as a second set of eyes and ears, but also had full manual controls for each of the subsystems. If the semi-automatic operation failed (ie, if a sensor failed, or a mechanism jammed) she could override and operate manually. It didn’t happen a lot, but when it did it sure was nice to have that backup option!
Obviously each game is different, and I can’t say whether this approach would work well for other games. We used it a little bit in Stronghold, but not at all our rookie year in Recycle Rush.