Drive Systems... ...The FIRST Arms Race?

I know that tomorrow is ship date and this message is likely to go unheeded for at least that long, but this is an idea that has been bouncing around in the back of my mind for a while now and it is coming out now…

I have not really been on a team for 3 seasons now – perhaps 4 depending on how you count things. From the outside looking in, I seems to me that FIRST has become an Arms Race of sorts, every year there are more and more teams with more and more motors devoted to their drive system with grippier and grippier treads.

There are a number of reasons for this trend.
[ul]
[li]FIRST game designs have encouraged pushing matches.[/li][li]Too many motors have been introduced to the kit [1][/li][li]The FIRST KOP transmissions have encouraged 2 Motors per side (4 motor drives)[/li][li]It has become easier and easier to implement multimotor drive systems with very grippy tires [2][/ul]From an individual team’s point of view, I can understand why teams make the choices they make. When everyone has 2 motors per side and custom made tires with super grippy tread, you have to do something extra to get an advantage, so, you put 3 or 4 motors per side and you go with even wider, grippier tires. [/li]
BUT… …from FIRST’s point of view, I think this is bad situation.
[ul]
[li]It is harder on the field elements[/li][li]It is harder on the other robots[/li][li]It makes teams do a lot of extra work just to keep up, for limited competitive value and almost no inspirational value.[/li][li]It is no fun to watch (imho)[/ul]So, I think that something should be done about this situation.[/li]
Possible ideas:
[ol]
[li]Re-think the motors in the KOP – 6 motors in the 1996 KOP may have been a bit thin, but clearly 17 in the 2007 KOP is excessive [3].[/li][li]Limit Power to Drive system by[LIST=1][/li][li]limiting the fusing used to supply current to the drive motors. There are several variations on this idea[LIST=1][/li][li]No more than 120Amps total[/li][li]No more than 40Amps to drive any wheel (this would only allow 40Amp per side if you had a single chain drive all wheels per side but if you had a motor for each wheel and you had 4WD, then you could potentially ahve 80Amps per side)[/ol][/li][li]Intruducing a feature the the Victors that reports current back to master CPU. Have master CPU limit drive current. [4][/li][li]Rate each motor for power and just put a limit on the power per wheel or the power for the total drive system (similar to limits on current via rated breaker amperage only limiting via rated motor power)[/li][li]Require that only unmodified (and relatively slippery) wheels of brand X be used for all robot drive systems[/LIST][/li][li]Make a simple test rig from simple pulleys, PVC pipe and dead weights that can test the pulling and pushing force developed by a robot. Limit robots pushing and pulling forces to something like 50lbs.[/LIST]There are pros and cons to all the ideas. I think 2.4 would probably limit creativity too severely. From a limiting damage to the field and other robots point of view I think I prefer #3 but from an ease of implementatoin point of veiw I like 2.1 or 2.3. The more I think about 2.2 the more fun we could have with the idea (read note [4]). But it would require significant work by FIRST & IFI so I think it may be DOA.[/li]
Think about these ideas and add your own. Perhaps this is just a solution for a problem that doesn’t exist.

As always, comments are welcome.

Joe J.

Notes:
[1] This year we have a record 17 motor in the kit with a total peak power of over 3000W
[ul]
li 2.5" CIM @ 330W[/li]li FP @ 380W[/li]li 3" CIM @ 270W[/li]li BB @ 100W[/li]li Globe @ 50W[/li]li Keyang @ 20W[/li]li Denso @ 20W[/li]li Mabuchi @ 15W (not sure if this is in the 07KOP but it is listed in the tips page)[/ul]As a point of reference, the 1996 KOP had 2 Drill motors, 4 seat motors, and a pneumatic pump with a total peak power of close to 500W (I am going from my memory so I may be slightly off).[/li]

[2] IFI sells great traction wheels, AndyMark sell great shifting transmissions and wheels and planetary gearboxes, BB and others sell great planetary gearboxes, the Nothing But Dewalt Whitepaper has made shifting CIM transmission extremely common even for rookie teams, etc.

[3] I say this as one with unclean hands. Over the years, I have played a key role in getting the following motors in the KOP:
[ul]
[li]Delphi Window motor[/li][li]ITT Window motor[/li][li]Delphi Power Sliding Door motor[/li][li]Bosch Power Sliding Door motor[/li][li]Bosch Drill motor[/li][li]Globe Transfer Case motor[/li][li]Keyang Seat motor[/li][li]Johnson motor[/li][li]Mabuchi motor[/li][li]Taigene Power Sliding Door motor[/li][li]Fisher Price Power Wheels motor[/li][li]Chiaphua (aka Atwood) motor[/li][li]2.5" CIM motor[/li][*]Banebots motor[/ul][4] Perhaps this would even allow a new game twist that might make games more dramatic. During a match, each robot would have a “Energy Meter” shown on the big screen. When a robot is out of Energy, it doesn’t move any more or it moves in “limp mode” where it can move but only very slowly. Perhaps robots could do things to get more energy for their robot or perhaps could win more energy for their alliance (e.g. score a tube on the high rung and you get an extra 10,000 Joules) Could be fun.

in this era of FLat-sidded 148 pound robots is getting very bland and un-interesting to me. back in the day the sloped armor bots had a cool effect on game play but with every other innovation it had its drawbacks. not with the mulitmotor transitions and high traction wheels i think first and to implement the bumper rule to even the playing field. but i also only see this as another step in the wrong direction. i shouldn’t have to use up 75% of my weight budget in my drivetrain. i agree that something needs to be done to bring the focus of first back to making cool end-effectors and creative balances of actuation and weight. at one time robots were very graceful and exciting to watch to spectators even if they didn’t quite know what was going on. I feel that we need to return to that. Thanks Dr. Joe

During seven years with Team #64 and the last year with Team #39 we always valued speed and agility over brute power. We never won the National Championship. This year we sold out and bought some traction wheels and still designed a robot for speed rather than push. Go figure? I guess old habits die hard! I believe Dr. Joe has a valid point and I support him!

Ken

remember in 1996 when you could get a top of the line video card with 8Mb of onboard ram . .

I think thats what were seeing now.

Ill explane why.
this year is my teams 2nd year. at the end of last year we looked at our drive system and said this was good but next year we want one thats just as fast but wont let people push us around. so thats what we did this year. as for next year who knows(weve already started talking about using small BB transmitions with fisher prices ).
and Im sure thats what every team does each year. sit down talk about what worked what didn’t and improve.

now on to power. I personally believe that this nation(and the world for that matter) is coming to an energy problem. politics aside, predicted American Chinese and Indian energy needs will not be meet with current resources and current technology. we need to eather A) have a major oil/coal/something find in a neutral territory, B) have a major breakthrough in energy science(solar, fusion, something else), or C) become better users of what we are given.
because plans A and B are a luck of the draw thing I like plan C. and on that note I think that FIRST should implement some sort of energy efficiency in to feature games.

I’d like to lend my cautious approval to many of these suggestions as well. FIRST gets a lot of its appeal from the large and powerful robots that go out to compete and a lot of creativity stems from the wide range of motors available. So limits would need to be carefully thought out. But let me give a brief example from my own experience:

At MIT, sophomore Mechanical Engineering students take a class called 2.007 (formerly Prof. Flowers’ 2.70), during which everyone individually builds a small robot (10 lbs, 18"x18"x24" max). So by weight, approximately a factor of 10 smaller than a FIRST-sized robot. The catch is that the whole thing must be powered by four baby Mabuchi RC-260 motors, capable of putting out no more than about 4 watts each. So by power, a factor of hundreds less than FIRST kits provide. The motors are also somewhat prone to burning up, and so particular care must be taken not to run at or near stall for extended periods of time.

I’m not saying that this necessarily turns out efficient designs (I know from taking the class that more often than not, it simply provokes endless complaints.) But the design limitation adds another dimension to the planning involved in making the robots. This year, I helped with the introduction of current sensors to the control system which can wirelessly feedback motor current data to the “OI” for use in design analysis and motor protection. We are also implementing an “energy meter,” just like Dr. Joe mentioned, for fun now but it may wind up in the competition in the future.

In all, I still much prefer the ability to run motors as speed sources and know that they will supply as much torque as I want, as in many FIRST applications. But I can see how some restrictions could be beneficial in leveling the playing field and encouraging efficient design.

Joe,

I agree completely with what you are saying. FIRST robots have moved towards making a beefy drive train, and the end effector sometimes a 2nd thought. You need to have a drive train with super grippy wheels and 4 motors just to be COMPETITIVE. I remember (it wasnt long ago) when you put 4 motors in a drive train, you would be seriously feared.

However…
I don’t think that we should cease the development of drive trains by putting barriers on power. Year after year I still get wow’ed by teams coming up with new drive trains (118 you definitely pulled that off this year) and I also get wow’ed by amazing end effectors. I agree that watching a pushing match is very boring. I feel that we are at a point where we cannot turn back and go to “the old ways”, which is not what your saying, but its in the same realm of things. FIRST needs to throw us a HUGE CURVEBALL with a game. I think they tried that this year with the robots lifting robots thing, and we’ll see how that plays out.

The reason drive trains are so strongly pushed for is because of how effective they can be. A solid drive train makes you competitive (theres always defense). A weak drive train (when compared to ‘stronger’ ones) with an amazing arm, may do well, but any solid drive train will put up a strong fight against that robot.

This most certainly is an issue, and its most certainly a sticky one. I feel some of your ideas may be a bit extreme, but some of them may be very effective (40 amps per side, 50 lbs of pushing power). Thanks for bringing this problem up Joe.

Maybe it is time to take the NASCAR approach and start limiting what can and cannot be done with drive systems. While multi-motor drive systems are nice, not every team has the financial or technological resources to build them. I am VERY grateful for the kit gearboxes since we are on a tight budget, but the reality it is no match against a 6 motor gearbox.

Come to think of it,FIRST, NASCAR, NHRA, etc. could form a partnerships and star some new robot related competitions! For Example;

  1. NRHRA (national robot hotrod racing association) - Robot drag racing at its best (After all we already have power tool drag racing)!

  2. NRTOWA (national robot tug of war association) - Two robots, one rope and one mean gearbox.

In hockey North American teams have traditionally produced larger, harder hitting players, while European teams have traditionally produced faster, more agile players.

Again, this is “traditionally” and “stereotypically”… one of the reasons is that the Europeans play on a larger ice surface.

Perhaps the solution is to increase the size of the field so that more agile robots can run around the push bots.

Jason

This may be the slight lack of sleep talking in the following post.

I honestly don’t see a problem with the amount of motors in the kit as is. I don’t like the idea of limiting what you can do either.

When first looking at Dr. Joe’s post, what struck me was the question of innovation (the other I word) versus constraints. On the one hand, innovation can be measured by one can accomplish in limited circumstances. However, that can be decieving. For example (as Brandon mentioned earlier), team 118’s drive this year is rather innovative. There were innovations back before we had this amount of motors as well. I don’t know much about robot drive trains before 2003, but the TechnoKat’s ball drive in 2003 sticks out in my mind. Innovation can be achieved in either way, in my opinion.

With that said, the recent drive trains COULD be considered an arms race. However, besides pushing other robots, what else could they be used for? With it so easy for a team to make a powerful drive train, perhaps it is time for FIRST to step up and make it a requirement in the game itself to have a very strong drive train (heavy game pieces, etc), which could then encourage further innovation and “raise the bar” for FIRST teams.

My team has been involved in First for 3 years. The last 2 years we had a robot that could push other bots out of the way and play good defense. We did fairly well (2005 we lost to the eventual nation champions in divisional semis) but we found that a highly maneuverable bot could evade us and a strong defense never won more matches than strong offense. In 2005 there was no pinning rule so every match we would score a few and pin the best opponent so he couldn’t score. The only problem was we played one team (i think it was the cybersonics) and they would just roll off of us every time we pushed them. This year we made a fast agile bot, but used shifting transmissions and custom traction wheels so we should be able to win most pushing matches. I do agree that some teams are ridiculous (one team had 8 motors) but two motor drive trains are acceptable. First should not restrict energy usage because, overtime, the better style of drive train should be more common. It is not First’s job to tell teams how to build their bots and what motors to use.

Well said Dr. Joe. I would think every person who have seen past competitions will agree with you on this matter. It’s was cool and great to see new transmissions, new drive trains, but I feel we are slowly losing the the offesive side of us and focusing more on defense.

Yes, it is great to watch a robot push another one to the other side of the field (SPAM 2002 pushed SigmaC@T 2002 from one side of the field to the other, it was the coolest thing ever). But what happened in 2003? Why didn’t we see many stacking bots? Why did we just see wheels on box with a monstrous drive train under it? What about 2004? There were more hangers than scorers. I love pink team and I would have liked to see them scoring balls rather than playing defense and hang. One of the pure scorer from that year was team 45 and they were on offense at time. There were more to watch in matches. Anywho, you get my point. I will talk about Team 118’s robot too in here. They come up with a very beastly machine every year, but I don’t think I have seen them playing defense as much. They have a beastly drive, but they only use it when they are being stopped from scoring. Oh man, you guys gave us the hardest time in 2005 on Curie.

If we were to use less motors, I believe it would give all of us a chance to look at the offesive side of the game, rather than just defense. It kind of makes me upset when I often hear, “Defense is the way to go.” But at the same time I love it when I hear, “Best offense is your best defense.”

Or I can see the other point of view where teams prefer to use certain motors for their applications. It would be nice to have all the motors we get, but if they were used to make more innovative machines, I would be happy. I really would like to see matches where teams compete to score, not compete to stop each other.

I think if you start limiting the drivetrain you’ll stiffle some good innovation. The last few years of FIRST have seen a number of interesting drivetrains crop up. Drivetrains are evolving and one route has been the stronger, beefier drive trains we see today, but the agility of robots has also increased. The first holonomic drivetrains only came about a couple of years ago and already those have evolved into the now more common mecanum drives(thanks to andymark). They’re stronger and more agile than their ancestor. Swerve drives have never taken off due to their complexity, but those that have been built have had enough succesful stints in the limelight that one could argue that they’re better than the 6 motor monsters out on the field.

As far as brute force drivetrains go, its really only been the third year of plentiful very strong drivetrains. But even these have been evolving. Three years ago there were very few 6 wheel drive robots, last year I’m pretty sure they outnumbered 4 wheel robots. The reasons are due to IFI’s traction wheels, the 4 motor kit gearbox and the “I have to use an Andymark 2 speed to remain competitive” problem. I’m not sure if those were bad influences, but they certainly accelerated the drivetrain arms race. I for one hope we never go back to 4 wheel drive, those robots were inefficient and had severe turning problems that plagued rookies and veterans alike for a decade.

What will come next I’m not sure, perhaps 118’s new drivetrain is a hint as to what the future will bring, but I think its clear that drivetrains haven’t stopped evolving yet. Maybe we’ve seen everything now, and within 2 years most teams will reach a pinnacle of strength or agility. I know that when the evolution finally stops, once teams can’t get any more out of the drivetrains in 6 weeks it will be time to throw a radical curveball, but for now give it another year or two.

A couple of asides. I think an example of what we don’t want FIRST to turn into is a racing league, where stiff rules make teams fight for 1% or 2% gains in their designs. Also I think this thread suffers from the chiefdelphi syndrome, which is that elite teams dominate the chiefdelphi scene. Numerous teams that aren’t making 6 motor drives and beautifully machined manipulators are out there and don’t post on chiefdelphi. At the same time most of the teams that would make a 6 motor drive end up showing it off on here. It doesn’t change the argument, but its good to keep in mind when you consider that this arms race isn’t really amongst all 2000 teams but just about 50 veterans that have the experience and resources to keep the battle raging.

I’m not so concerned about power in the drive. If the team builds it well and can control it on the field, what does it matter?

I’m more worried about people buying these new, off the shelf mecanum wheels and not knowing how to program for them correctly. What a waste.

i don’t agree with this. There are members from most teams on Chief Delphi. And alot more than 50 teams are capable of producing 6 motor drives but, like my team, do not see the point. They are heavy and unnecessary. A well designed 2 motor drive will still be very competitive. Also you cannot win a match by having an amazing drivetrain. The only thing you can win from a good drive train alone is a pushing match, which FIRST doesn’t awards points for. Something more important than the drivetrain is the driver because a smart driver can avoid a pushing match when possible, or knows when to shift down to push thru someone. If a team wants to spent 75% of their weight budget on a drivetrain, let them, but one of the elite teams will know that scoring is more important than pushing.

I completley agree with Tom on this one. There are so many of these new off the shelf parts every year, that teams see, and instantly say “Oh, thats different, lets use it, its spose to work great.” When in reality they have no idea what to actually do with it. I’ve seen teams ruin shifters and wheels and so many objects by not knowing how to control them/program/use/install them. I would much rather see teams design their own parts, possibly using the ideas of others(granted they understand them). These companies who do mass produce these parts, should include more drawings, and more information to help teams understand the use of the parts.

I definitely haven’t been involved in FIRST for as long as you have, but within my scope of experience (5 years in FIRST), I disagree when you say that defense has been on the rise. Even back in 2002 (the year before I joined my team), a strong drivetrain and lots of pushing power was very important. In 2003, defense was king. Ever since 2003, I feel that defense has actually been steadily de-emphasized. In 2004, defense and offense were about equal in importance (the ideal balance in my opinion). In 2005 defense was almost completely eliminated from any place of importance. Last year, defense was slightly more emphasized, but you still couldn’t win the game on defense alone. It’s tough to make a call about defense this year yet, but certain rules (i.e. - you aren’t allowed to de-score opposing ringers) lead me to believe that FIRST (the GDC at least) is trying to balance the forces of defense and offense (if not outright giving offensive teams the advantage).

Also, the addition of more motors to the kit does not necessarily mean that FIRST is trying to encourage a drivetrain arms race. More motors could mean more easily facilitated (and more complex) offensive arm designs as well.

I’m not going to disagree with you on your last two points, but I don’t think that those two facts are necessarily a bad thing. How many teams would want to use the KOP transmissions if they felt that they would be inadequate in terms of pushing power? Correct me if I am wrong (you are probably in a better position than I to judge this), but I don’t think multimotor drivetrains are any more prevalent than they were in 2003, my rookie year. I think that in encouraging teams to use 2 motor/side transmissions, FIRST is just trying to level the playing field by providing a competitive solution that is accessible to all teams.

As for your second point, yes, you are probably right. More defense will result in more damage to other robots. However, one could say that this encourages teams to be come up with more robust designs (a good thing, IMO. It’s important to engineer robust solutions in the real world - not just elegant or showy ones).

I definitely disagree with your last two points. I really don’t think it is too much extra work to design for 2 motors/side as opposed to 1. If I’m going to sit down and design a transmission, it would probably take me the same amount of time regardless of how many motors I wanted to use. Plus, the 2 motor solution is easily facilitated by the KOP (or at least it was in 2005 and 2006).

My main disagreement, however, is with the statement “limited competitive and inspirational value.” Innovative drivetrains have *certainly *shown their value in competition. On my old team, 716 (and many other teams, I’m sure), the drivetrain was always placed at the very top of our priority list. We knew that we would not be able to do anything effectively if our drive system could not meet our needs. For example, we decided that we wanted 2 speeds so that we could play defense effectively, but still be fast enough to move around the field and complete our offensive tasks. This meant that we needed to design and manufacture our own custom transmissions.

How is it less inspirational to design a custom transmission than to design an end effector? How is it less inspirational to design and manufacture a mecanum wheel than to design a turret? I think any design project like this is a beneficial experience to the students on the team. It doesn’t matter whether it’s a part of the drivetrain or a part of the arm; as long as the students see the end result of their work, both are inspirational. Just look through the pictures on CD media tagged with “drivetrain”, “wheel”, and “transmission”, and try to tell me that there is no innovation occurring within those disciplines!

Also, I think that defense is a very important and necessary part of the game (strategically and in terms of its “interestingness”). However, I do believe in a balance between the two elements. In the perfect game, an excellent offensive team should be able to beat an average defensive team, and an excellent defensive team should be able to beat an average offensive team. 2004 was the perfect example of a “balanced” game - good defense AND good offense could win a match. In 2003, the scales were tipped too much in favor of the defensive teams; conversely, in 2005 the scales were tipped too much in favor of the offensive teams.

On a side note, how can anyone say that defensive play is boring to watch? I will never forget the 2004 finals and semi-finals on Archimedes when my team, 716, was faced off against some of the strongest defensive (as well as offensive) bots in the world! That defense was some of the most intense (and visually and emotionally awesome!) that I have ever seen (494, you are amazing). It was heart pounding and exhilarating - I wouldn’t ever want to trade that for a technically interesting, but rather boring exhibition of robots that are just scoring (and no one is trying to prevent them from doing so). The presence of defense makes it even more amazing to see high-scoring robots perform - knowing that they are experiencing difficulty, yet are still able to overcome!

Definitely do not impose limits on the amount of power in the drive system. As it is right now, the amount of power a team can have is already somewhat limited by their coefficient of friction against the ground, so potential power is not infinite (after a certain point, the wheels will just begin to spin against the carpet in a pushing match). Yes, the extra power will help in acceleration, but there comes a time when too many motors becomes more costly in terms of its draw on the battery (and less motors available for other manipulators) than it is worth. I don’t fear that defense will intensify in the likes of an “arms race” for these reasons (as well as the fact that it’s obviously not being encouraged by the GDC). Also, it would be costly in terms of battery power to use ALL of the motors in the KOP. Thus, teams will have to learn about restraint in their designs and make responsible choices about motor selection.

Thanks Dr. Joe for bringing up this discussion. It’s very interesting, and I hope I don’t sound as if I am being argumentative. Just trying to share my two cents! You definitely have a much greater perspective on the evolution of FIRST than I have.

– Jaine

This is a real issue. This year, our drive system was biased towards going off-the-shelf when possible. The result was a 6WD with Gen2 AndyMarks. Works beautifully.

Overall, I agree with most of the points raised in the thread. However, I think the solution isn’t so much in meters and lines on the inspection checklist. The solution seems two-pronged to me.

  1. Cut down on some of the drive-grade motors, and perhaps work in some more arm-friendly motors. Drop a couple of the Big Powerful Motors (herein defined as the CIMs, big and small, and the F-P), either individually or in pairs, and throw in a couple more window motors. I mean, sure you can do window motor drive, but most teams will just work with less power there.

  2. Create game objectives that value finesse and agility over pure brute force. Rack 'N Roll, Triple Play, and (to an extent) FIRST Frenzy can arguably meet this criteria. Aim High, Stack Attack, and Zone Zeal couldn’t to the same degree. Make it so we’ve got to move fast to score well, and we’ll do it.

I disagree with you about Rank 'N Roll encouraging a maneuverable drive train. If you don’t have the stickiness and the power to stay put when someone is trying to push you this year, you will have no chance of being able to score a tube. precision manipulation with an arm requires a stable base, and if a team is trying to nudge you, they probably will be able to.

I want to see games with more variety and strategic possibilities. I think adding the three different weight classes of robots this year was a really interesting idea. What if they took the idea even further and limited classes even more? We might see even more innovation and teams working to fill specialized niches, as in the 2004 game. The past three years there has been a single viable objective for teams. This year starts to change that a bit with the idea of a “carrier robot”, but there really isn’t enough differentiation.

Force teams to do something other than push hard and do one thing well.

Greg, I gotta disagree with you on the weight class idea. They are going in the right direction, but I think they need a different type of variance. I really dont see alot of designs changing drastically do to this rule.
Raising the Bar, in my opinion, is the best game FIRST has had. It had alot of variance. You could be a ball manipulator or you could hang, or you could do both. In recent years, you basically had to make an arm or make a shooter. There really weren’t multiple options. FIRST needs to concentrate more on making those options available. They don’t need to have different shaped game pieces (that might make it too difficult), but they need many more options. This year, you dont have much to choose from.

I really appreciate everyone’s very thoughtful comments. It was much much more response than I ever expected on the day before shipping.

I just want to make 2 points:

#1 I never even used the word “defense” in my posting. I am not against defense and that was not the point of my post.

#2 I misstated my point when I said “It makes teams do a lot of extra work just to keep up, for limited competitive value and almost no inspirational value.” I SHOULD have put in there “almost no additional inspirational value”

Trust me, I am a HUGE believer in the inspirational value of drive trains. Building & competing with remote control machines that can DRIVE is a fundamental ingredient that makes FIRST work. My point is just that I don’t believe that having the average number of drive train motors climb from 2 to 4 to 6 to 16 is really a good investment in terms of the added inspiration for the extra effort expended.

Joe J.