Training New Drivers

FIRST General Forum
1

Hello, so my question to the CD community is this: how does your team go about training and selecting new drivers? In the before last year, there was no official driver selection process, it was just the senior members, or whoever worked on the robot the most. Then last year after going to a seminar hosted by team 701, we switched this up and had driver tryouts. I just set up a small obstacle course in our lab, and everyone gave it a couple tries. Whoever had the best times was the driver. This was a step in the right direction, but Iā€™d really like to know what I can do this year to make it even better, thanks!

1 Like
2

Whatā€™s has worked for our team in the past is trying all the drivers out at a competition. Either a week zero scrimmage or a offseason event. And then which ever drive operator Combination looks the best on the field under pressure should be the choice for ether years competition.

3

I know there are teams that set up different ā€œstringsā€ of drive teams, based on skill level. Often the second and/or third string will be younger, and will be trained by driving at off season competitions. This year, our team has taken a list of people who are interested in positions on drive team, and weā€™ll be training them and then giving them control at an off season event. Itā€™s good to give everyone that wants to try it a little bit of training beforehand because sometimes thatā€™s all they need to be able to show their full potential. The Air Force doesnā€™t just toss people into jets and pick the best ones, they teach them how to fly first.

4

Other than having them run through an obstacle course or a practice match we like to have a rules test to see who knows the rules the best. If the driver doesnā€™t know the rules you are going to have a bad time.

5

We set up ā€˜stringsā€™ or rankings during the off season through watching our members practice and seeing who looks the most promising. Then we put our drivers to the test on the field at an off season event. Itā€™s important to remember that someone who is super good at driving around obstacles in the lab can freak out and lose their cool on the field. Remember the most important thing is that your driver knows ALL the rules. I wouldnā€™t be surprised to learn that only 100 teams in all of FIRST do this. Read them, reread them, rereread them. Drivers should be able to tell you what the most common fouls are and how much each one is worth. In short the driver has to be good at driving, good under pressure, and willing to put a bunch of meticulous work in. Good communication helps too.

6

This year, for the first time, we have a dedicated driving mentor. A former student who is now an engineer for our sponsor (perfect FIRST paradigm) recognized that the driver was almost as important as the robot in a competitionā€“though others might argue that the driver is more important. Anyway, he volunteered to set up a driving selection/training program which began in early fall. The team constructed a ā€œrobustā€ driver training robot complete with a full sized car battery for endurance. He developed a rubric for measuring various driver attributes and set up a training area in which any member of the team who made the initial cut would then have a great deal of practice time to see who had the best overall skillset.
The training was very intensive, even going so far as to create artificial stress in an attempt to see how drivers might respond under game conditions. The results were positive. We were able to empirically measure various attributes of each of the drivers and were very happy with the results. Unfortunately for us, however, the top three drivers are seniors and weā€™ll have to select a new driver next year.
Another positive aspect of this selection method removes the argument, ā€œI worked harder than he did and I deserve it moreā€, or, ā€œIā€™m a better driver and he was selected because ofā€¦ā€. After each round of testing results are posted and everyone knows that the facts donā€™t lie, itā€™s not subjective anymore.

7

Our process looks something like this:

  1. Offseasons:
    With the exception of IRI, we have students who are motivated and have expressed interest in driving, drive coaching, or human playering (is human playering a word?) operate the robot, as long as they attend one of our robot operation and driving training sessions (where we go over controls and SOPs). This lets these students get enough training to feel comfortable with driving prior evaluation. One of our mentors works on a rotation schedule to ensure that everyone has an equal opportunity to be in all of the positions that they want while working with a variety of other students.
  2. Rules Test:
    All prospective drive team members (driver, operator, coach, and human player) must take a multiple-choice rules test developed by our drive team mentor.
  3. Drive Coach Essay:
    Drive coaches also are given a few essay questions that present different scenarios, and ask how the candidate would respond.
  4. Tryouts:
    Drive coaches come up with drills, and coach permutations of operators, drivers, and human players through them. Not only does this help evaluate everyone individually, but this allows our drive coach mentor to see how people work together ā€“ the whole is more than simply the sum of the parts.

One of the things we tend do differently from other teams is that we select our drivers during the build season. This is a large factor into why offseason events are so critical for us in getting drivers to have experience. Other teams often select the drive team before offseason events, and stick with this static combination for the following game/season.

We also strive to have a student drive coach. There has been a lot of discussion* regarding whether itā€™s best to have a student drive coach (as the rules allow a mentor to fill this role too). While I have no idea what the best choice would be for your team, itā€™s the happy medium our team has come to.

*Relevant threads and blogs:

8

On our team for as long as Iā€™ve been with the team, tryouts happen typically after the practice bot is complete. First people are tested with an obstacle course run where everyone is watching. Typically this is where people crack under the pressure of being watched and judged by everyone on the team. From there, members trying usually practice doing various roles in the game such as how many balls scored in 2 minutes. Everyone who is trying out has to try out for driver and operator so everyone switches up and typically the best combination is who becomes part of the drive team. The real pressure is also judged at the local scrimmage event Duel in the Desert. Here itā€™s not so much about how good people are but how they react to pressure of everyone watching and other robots on the field. Typically tryouts last 1-3 weeks since we want to make sure we have the best drive team possible. For example, weā€™ve had people on the team that are awesome right off the bat but then never really get much better and people that start out below average but develop to the point where they surpass everyone else. Tryouts also happen every year so just because someone was driver one year, doesnā€™t mean they will keep it the next year. Human player tryouts are basically the same thing except they are able to practice more due to not being reliant on the robot being available. This year for example, our human player had probably 2000 in-bounds to a red x on the floor. Even when the human player role isnā€™t as critical like in 2010, we always have them prepared at the same intensity because although they may not always be able to win the alliance the game, they can sometimes cause losses due to not performing their role correctly.

9

Like others, we give a rules test. But this year we extended it to the entire team. We wanted to make sure that the entire team knew the rules inside and out. We then had a secondary Driver Rules test which was more specific toward the drive team. Very very few multiple choice answers were given, this way you could tell who was actually guessing or just mixed up a number. The Driver Test portion also include situational questions so that the mentors can get a better feel of how the student thinks.

There were a couple problems with this. 1) Grading 2 multipage tests that werenā€™t multiple choice was extremely time consuming (esp. during build season) and 2) There was no time limit for the situational questions. A student could come up with a good strategy but you have no way of know if that took 5 seconds or 5 minutes to formulate. I have heard of teams doing a sort of interview where situations are presented and they explain what they would do or draws on a white board. I think we are going to give this a try next year.

As far as try-outs with driving. We allow any student to try driving during off season events. We also hold a few fun driver practices in the shop during the fall where weā€™ve set up obstacle courses and run time trials. I plan to improve these so they are more structured for this summer/fall. In my mind, ideally we would select a drive team following our week 0 event. However depending on the state of the robot thatā€™s sometimes easier said than done.

Some things I look for in a drive team:
Must know rules
Communication
Smooth control (Slow is Smooth, Smooth is Fast)
Knowledgeable in how the robot works (donā€™t expect to know every nut and bolt but should know what does what)
Communication
No Tunnel Vision (should be able to see the field, not just the robot)
Short Memory (needs to learn from mistakes and move on)
Communication

A couple students asked me last year if particular drive sessions were official ā€œtry-outsā€ or just ā€œpractice,ā€ the way I see it anytime you are behind the glass you are being graded. But that doesnā€™t mean you canā€™t mess around and have funā€¦

10

I just want to say most people arenā€™t able to put themselves ā€˜inā€™ the vehicle the first time they drive and i think thatā€™s the most important training they can get.

Many kids play video games, but even with them, there isnā€™t a third-person point of view to watch from.

So what Iā€™m saying is you need drive time. Start with a radio controlled car because itā€™s much cheaper to crash than the 'bot. But run it for hours. Drive patterns. Race each other or a clock. Drive obstacles. If you want to be a driver, you need to have your brain in the robot. It doesnā€™t happen automatically.

You can move on to R/c planes or Quads and those will sharpen your skills, but before you can successfully drive the robot and win, youā€™ll need many many hours of third person driving.

Only go to the robot once youā€™ve mastered the part about thinking as if youā€™re in the vehicle.

11

I was one of the drivers this past season for my team. We had a drive team mentor who quizzed us on all the rules, and ran time trials and drive tests. Example: one test was see who could score the most cycles by themselves in 2 minutes and 20 seconds. The driver could decide how they wanted to drive, whether they would truss or not, and what not.
We were also given many scenarios and asked what we would do.
Coming from a students perspective, and having been a driver here is some advice I would give:
-Drive time makes a huge difference, so give each perspective driver to do whatever he/she wants with the robot for a set amount of time. Make no rules for them, and let them get accustomed to the robot. (During this time, see how much they improve, because it will show a general relation to how quickly they will improve in a competition)
-Look to see who drives the robot on the edge of the line, going as close as they can to being out of control, but still having control. If a driver drives the robot at top speed and as hard as they can in practice, then they will do it in a competition. If the driver does not drive the robot to its full extent in practice, then they will not drive the robot to its full extent in competition, which may lose a match.
-All drivers should know every mechanism on the robot nad know what they can do when which each individual part. Many times I see drivers not use all the mechanisms on their robot which causes them to not do as well.

12

I personally have been our Driver 1 (Chassis) for 3 seasons and 4 years of off seasons. My first year trying out there were many requirements before you could try out. Such as knowledge of rules and plenty of hours put into the season. Then tryouts were done as pairs (Driver 1 + 2).Driving was based on 50% talent and 50% communication.

Our Adult coach believed that drivers skill can be taught through practice. Its alot harder to teach communcication and how to work together as drivers.

I Eventually won and ever since have had a different driver 2 every year, but continue to be the best driver we have. Driver 2 is based a bit more on team dedication and how well they work with Driver 1.

Coach (student) is decided by communication skills, how well they understand game, and how they work with drive team. A new qualification is how well they keep calm in match. Human player used to be on skill level. It has now become 40% skill, 60% dedication

This year due to district format, we wanted to get more people on the field so we switched driver 2, between districts. And then put our best in at NE Champs.

Off season is another thing! Due to my experience as driver, Im usually coach and then anybody that has driven the robot in the shop or at a demonstration, is dedicated to the team, and knows the rules can drive. We usually have 6-10 people rotate. We believe that getting kids on the field and putting a smile on their face is better than winning an offseason event which is why we let anyone drive. Also its sometimes funny to watch some people!

13

We did things a little differently this year than we did last year since we ā€œfinishedā€ building sooner than we have in the past. Last year we didnā€™t really have anyone that was interested in being on drive team so I was kind of forced onto the drive team because our coach thought I would be able to deal with the pressure and also be able to go out there and do what needs to be done as the operator and the same can be said for our driver. Throughout the entire season, we did not have a set human player but switched back and forth between a few people at each competition that we went to.

This year was a lot different. We actually had a working drive train and semi-working mechanisms by ~week 4 or 5 so we had some time to actually let people drive the robot around. We had everyone that wanted to be a part of the drive team write their name on a list of what position they want to be on the drive team and then we went through all of the different permutations of everyone to see who worked well together and who was able to work through the theoretical pressures of competitions. They would run through 2 minute cycles and the final score would be recorded along with how well they communicated with each other. After a round or two of running through the different permutations, the list was narrowed down to 2 people for both driver and operator.

Our mentors then sat down and looked over the data and talked about who worked well together and who didnā€™t. We then had to wait until our next meeting to find out what was going to happen next. As it turned out, the mentors decided who our drive team was going to be and made the other two the back-up drive team. They made sure that each of us understood why the drive team that was picked was picked and made sure that everyone was okay with the final decision.

14

The one thing is that the whole drive team (drivers, human player, coach, etc) need to be best friends. You will be spending a good chunk of time around these people in tight situations, so everyone has to work well together. Also, if one driver has comments about what the other driver/human player did in that last match (for example), then they should only talk with the drive coach about it. Let the drive coach deal with any problems amongst the drive team members, it keeps everyone calm.

15

Our driver is normally the driver for all of their years that they participate in FRC. This is because we believe the driver continues to improve and gain experience as the years progress. The last driver tryout was my freshman year, and the previous driver came back from college during winter break/the start of the build season and held tryouts along with the coach/captains. While this has worked out fairly well for our team, we want to start teaching more students to drive so that the people trying out for the driving job are not learning on the spot.

16

When our team goes to decide who will be driving the robot at competition, there are a couple of different factors that we look at. First, we determine who is able to handle the robot with the most ease. Similar to what you did, we set up an obstacle course and have anyone interested spend some time driving the robot around. Once they get the hang of it, we have them practice various maneuvers to see who can complete them the fastest and with the most accuracy. We also see who performs the best under pressure. You donā€™t want to put anyone in a position that they arenā€™t going to be comfortable in.

Usually we have the two best drivers on drive team, with a third lined up in case something goes wrong. These drivers, as well as the coach and human player, meet once or twice a week to discuss strategies, rules, or anything else that we find important at the time. Itā€™s important for the drive team to get along well, because they end up spending a lot of time together. The better the drive team can communicate with one another, the better the robotā€™s performance will be during matches.

In the end, we try to give drivers as much practice time with the robot as the schedule permits. We all believe that the main way anyone is going to improve driving is with practice. Ultimately we try to prepare them for competition as best as possible, whether it is talking strategy or getting practice time in with the robot.

17

We donā€™t really spend any time with obstacle courses or skills tests, I feel this is a backwards process.

Being part of the drive team is a long and complicated process that is far more involved than just driving matches (at least on 973, weā€™re a small team). Drivers are responsible for tuning/learning the robotā€™s functionality, and then obviously driving it. This requires effective teamwork, communication, time involvement, the ability to handle pressure, the ability to thrive on feedback and criticism, and many other things.

Once weā€™ve narrowed down by the above requirements (usually just the ability to handle pressure and criticism), there are few candidates left. We then use the offseason drive practice and events as a testing ground for the remaining students. Since we run multiple robots and go to multiple offseason events, we can test out about 2/3 of the team any given year (which is often more than are interested in even trying it). This also is kids driving full events, not just a match, so itā€™s a much better experience.

The amount that any team hoping to be competitive should practice is enough that a driver who is better at handling pressure but initially fails the obstacle course should prepare that driver with the skills needed. Itā€™s very difficult to teach a kid to not crack under pressure, much easier to teach how to drive.

18

What we did this year (the only year in recent memory weā€™ve had a solid drive team) was simple. The important thing was that it was tailored to our teamā€™s situation.

We donā€™t have an indoor space where we can set up carpet and game elements to practice on. We can ask for practice time at Ames, thanks to 254, 971, and 1868), but obviously itā€™s difficult to visit extremely often. Any practice opportunity we get has to be arranged by students and we have to work extra hard to find those opportunities and make best use of our limited time. We have a lot of students who are interested in driving the robot, but not all of them necessarily have the motivation required to become excellent drive team members. Even fewer want to spend time watching hundreds of matches and observing the best moves and strategies.

Anyway, hereā€™s what we did:

  1. Iā€™m the guy who reads the rules, watches matches, feels strongly about the drive team. So during the fall semester, I started making tests that evaluated knowledge of the rules and our robotā€™s technical attributes (how to maintain it, what to check pre-match, how to adjust things, etc.) and administering them to anybody interested. We will never field drivers that donā€™t understand the rules or the robot forwards and backwards. In the process, I became defacto coach of the drive team.

  2. Most failed the tests. I allowed candidates to retake (slightly different tests) as many times as they needed to pass; if they studied and learned how to take care of the robot, they would have the skills they needed to become good drivers. Those who didnā€™t have enough time or motivation to become good drivers generally dropped at this point. By the end, a small group of students (10ish) moved on to practice at NASA Ames before CalGames.

  3. Then students were selected based on ability to handle pressure during practice. 8 students were left; they comprised two drive teams for our offseason competitions.

  4. Fielded those two drive teams at CalGames (fall offseason) on one robot. Practiced at NASA a few more times. Fielded 2014 competition drive team at Madtown Throwdown (fall offseason). Now, the team has two functional robots and will be using them to train even more students at these competitions.

  5. After we got the 2014 game and as we built our robot, I created more tests. I once again opened up the opportunity to take the test to everyone, and did the same weed out process. We selected two drive teams, one to drive our competition robot and one as backup and also to play defense on the main drive team.

Important note is that all drive teams selected committed to practicing during all practice opportunities, and we committed to be available during several time slots a week in case we were able to go to NASA or something along those lines. We knew we would need more time to practice for this game, so we acquired carpet, laid it down on the blacktop behind our room, and practiced in the open air. (see: http://youtu.be/92IbHU0Z76I?t=46s)

We have to work really hard to create practice opportunities for ourselves, and the driver candidates have to lead this effort. Therefore, motivation is the most important quality we need in our drivers. We find those committed students through the difficult tests and by mutually agreeing on an excessive practice schedule. If a student canā€™t commit to studying the rules, knowing the robot, and/or practicing at the level we agree upon, then they canā€™t drive. Simple as that.

19

Iā€™ll second what M. Lillieā€™s said about the whole drive team needing to be best friends. If everyone on the drive team is just awesome friends, then theyā€™ll have a great time, be more relaxed under pressure, forgive themselves for mistakes more readily, and generally have a sense of camaraderie. Iā€™ve also been on drive teams where some of the members really didnā€™t get along that well (personal rivalry, jealousy, incompatible personalities), and it can get toxic pretty fast. Instead of moving forward, a lot of the post match stuff will be playing the blame game and thereā€™ll be more negative pressure pre match.

Another thing a lot of people forget about is that the drive team needs to know basically everything about the robot and how it breaks (including software). Theyā€™re the last people to interact with the bot before showtime, and sometimes the only people to react to stuff that comes up in queuing.

A lot of people believe that ā€œanyoneā€ can be made a good driver as long as theyā€™re dedicated and responsible enough. While that may be theoretically true, I think innate skill and the abilities to make decisions and ā€œgo for itā€ in a match usually trump dedication (at least for us). If someone already has a great field sense and can make the robot do what they want, Iā€™m willing to overlook a bit of immaturity or irresponsibility. Anyway, dedication does not necessarily make a good driver. Iā€™m one of the most dedicated people on the team, and still canā€™t drive the robot without crashing it into walls.

Despite having a large team by some standards, itā€™s always been pretty clear who should drive, given thereā€™s a limited number of people who show up enough, have the ability, want to, etc. Weā€™ve never really had to do a formal selection process.

20

Thanks for the great advice. So on our team there is a driver and a manipulator. The driver just drives, and manipulator does all the rest. It seems somewhat obvious who our A-team driver will be next year, but since the manipulator that worked with him is leaving, that opens a spot up. Do you guys prioritize individual skill over people getting along? I.E, if the most talented manipulator didnā€™t get along with the best driver, would you not make those two the drivers?