New FTC Platform

[gblake]

Originally Posted by Rick TYler

I'm in Atlanta right now -- the information above is correct. I stopped by the demo field and and never saw a robot moving under remote control -- but the demo teams said they only had two weeks to build their 'bots so that's not a big surprise (at least to me). Overall, it's a lot of value for $450, and includes three SDKs in the box (Labview, RobotC, and something else -- the LegoNXT environment?). Got to run to dinner now...

If I remember correctly we ("the teams") had 4 weeks or a bit longer to build our bots. Whoever told you two weeks did not get to use all of the time that was available.

The bots can, do and did move under remote control. However, what you saw was evidence of the non-trivial software problems that we encountered in this version of the kit and field control system. I presume that more mature versions of all the different software tools will be shipped; but I wouldn't be surprised if the first season encounters a few bumps in the proverbial road when hundreds of FTCers truly start field-testing them.

Blake

[thatguy]

gblake how does the new platform compare to the Vex one?
I was only able to catch a few minutes of the webcast match
Anyways, how easy was it to develop a working robot with an arm and it seems as if you can use a bluetooth playstation controller? If so and you did use one, is it fun to use?

[basicxman]

yea i think the new kit contains an entire NXT kit

[gblake]

Originally Posted by wilsonmw04

watching the demo. One thing popped into my head: cool robots, funky game. I really hope this is just a demo and not the direction the games are going to be heading. It's not very game like is it?

I have a hunch that the "FIRST Response" game/challenge was chosen for several reasons; one of which was that in a disaster situation, robots and the people controlling them can be expected to cooperate, not compete.

Cooperating robots don't intentionally collide with one another (the game rules included a penalty for interfering with another bot). Robots that don't collide with each other don't convert each other into piles of LEGO rebble.

Also, with their obvious resemblance to LynxMotion parts, many of the structural metal parts in the kit FIRST gave thier "Showcase" participants seem better suited to building walking/crawling and manipulating machines, instead of the typical FTC, go-karts of recent seasons.

Based on the above, I do expect the nature of the FTC games to change in the direction of becoming more like FLL challenges rather than staying the same as in past seasons, or becoming more like FRC challenges.

Blake

[Rick TYler]

Originally Posted by gblake

Based on the above, I do expect the nature of the FTC games to change in the direction of becoming more like FLL challenges rather than staying the same as in past seasons, or becoming more like FRC challenges.

I got some flashes of this in the last few days, but nothing I can really put my finger on. In my very strong opinion, the more FIRST makes FTC look like FLL, the more teams will avoid it. There is no way a juiced-up FLL is going to serve as "high school competition robotics for less money." The more FTC looks like FRC, the better it will be, and this will have to include permanent team numbers, wider variety of parts than even Vex has right now (and despite some obvious strengths of the the FTC kit, it will be a while before they can come close to matching the richness of parts choices in Vex), and serious competitions that resemble FRC games. For my part, I would like to see a an FTC version of "Aim High" using 2-inch foam balls. That would rock. Doing pretend under-sea exploration or toxic waste removal would cause a mass abandonment of FTC.

I'm a FIRST loyalist, and made a conscious decision to step down from FRC to FTC. If FTC becomes SFLL, I'm out of the program. I'll either go back to FRC, or switch to some other competition.

[basicxman]

Yes, technically I'm underage for FTC this year but I was either starting an FTC team or an FLL team, I looked at past games and I simply just did not like FLL. For one completely autonomous is not my style. And yes I don't think teenagers are going to quite care for imagining an oil spill happened at you have to send your robots to clean it up. Also, permanent team numbers would rock!

P.S. aim high 2009 FTC would be awesome too

[wilsonmw04]

I would tend to agree. I want a mini-FRC game, not a super FLL game.
I've been told that numbers will be permanent startin this year.

[thatguy]

Ugh I'm really starting to fear the future of this program. My school has already decided on FTC next year and since next school-year is my senior year, I want to have a super-fun robot experience.

Hopefully FIRST will give some promises for a FRC-ish like competition. Did the testers complain about the kit to the FTC people?

My school won't do FRC, so I really want something close to FRC to satiate my wants and I really felt that way with this year's FTC. Oh and my friends and I found Quad Quandry somewhat similar to Rack 'n Roll.

[fredliu168]

I got so many complaints about the kit from several teams (who will stay hidden). From my general impression of the complaints, the kit wasn't thought through very well, quite inconvinient in building (holes don't align), programming was difficult, parts were awkward to use, and it generally wasn't suited towards a FRC style game. Its closer to FLL, where the challenge of teams is to perform tasks (just look at the showcase).

I have to admit I am very disappointed in what FIRST has come up with.

(Yes I'm in Atlanta right now)

[thatguy]

So I guess all my fears are coming true. And I thought the new stronger metal would make FTC bots closer to FRC bots...

I guess at this point, only mass amounts of emails can do the trick.

(no I'm not in Atlanta right now, when I saw the webcast I believed the game's purpose to showoff the capabilities of the new kit instead of giving the idea for an actual game)

(I blame Team 74's alliance and Murphy's Law for why I'm not in Atlanta, why did they have to be so good in the latter half of the NYC regional D: )

[gblake]

Originally Posted by thatguy

gblake how does the new platform compare to the Vex one?
I was only able to catch a few minutes of the webcast match
Anyways, how easy was it to develop a working robot with an arm and it seems as if you can use a bluetooth playstation controller? If so and you did use one, is it fun to use?

They are different:
The new kit is not a "better" kit. It is a different kit.

I and some of the other participants will need to collect our thoghts (now that we can reflect on the experience, instead of spending our time scrambling to try to puzzle out how to best use the parts FIRST gathered together to make the kits. That will take a little time.

The Vex product line is fairly well integrated into a unified offering. Sometimes when trying to solve a "game" problem, this can lead you to wonder if a different vendor makes a better mouse trap.
In the new kit, in addition to the obvious impications of including snap-together LEGO plastic parts in the kits, the version of the new FTC kit we used seems more like a mixture of several similar, but not the same, different part styles. Diversity can create strength; but it also created frustration when we had to try to blend those philosphies in a single machine.

The Vex system has considerable openness (the VexLabs fora frequently discuss how to integrate non-Vex sensors, motor controllers, etc.) with the Vex Microcontroller.
I am still fairly unfamiliar with interfaces the LEGO microcontroller exposes, but I think that they implent IC2 communication interfaces. Given that, I presume that one can find sensors or other devices that use IC2 and can write or buy communication software/drivers that can run in the LEGO NXT computer.

Attention to mechanical details has resulted in being able to easily mate almost every Vex part with other Vex parts.
Some of the new kit's parts just do not yet connect easily with others (at least if you are asking me to figure out how to connect them). Attaching a rotating part (a gear on an axle) to a stationary part was a partucularly onerous chore.

"How easy was it to develop a working robot?" At this stage of the kit's lifespan, it was not easy. The mechanics were often odd and clunky. In general, the software support still needs quite a bit of work (the vendor reps were quite helpful and generous; but the product(s) simply are not yet mature enough for release to the general-public).

Putting "Arms" on Vex bots is a challenge because the Vex Servos just don't put out much torque compared to the weight of the larger Vex steel parts; and the Vex aluminum parts are only sold in bundles (and I don't want to pay for bundled parts I am unlikely to ever use.
Putting an arm on the new FTC kits was pretty easy. The aluminum is light, the servos are strong and the LynxMotion-style parts are expecitly designed for that sort of thing.

Yes - With the right software drivers you could use a bluetooth Playstation controller. Was using the Logitech handheld remote controller (plugged into a computer that handled the bluetooth protocol and had a bluetooth dongle installed) fun? Not so much. Video game objects respond well to Playstation/xBox, etc. handhelds because software developers spend hours and hours tuning the conversions of user actions into game actions. Doing that for any realworld object is is not going to work any better than using something like a Vex transmitter if the software involved doesn't get tuned to match that realworld object's behavior. An FTC robot is a collection of realworld objects.

Blake

[thatguy]

Blake thank you for that in depth response. I have basically been researching the new FTC platform for hours a day for the last two weeks (it didn't really become an issue to us until after the NYC regional). I've been communicating the information I have found to my friends and have had some discussions with them on this situation. (I must say chiefdelphi and vexforum.com are amazing resources)

I only wish I could somehow get this information to my teacher, but unfortunately we have school off next week due to Passover.

Again thanks,
-thatguy

[gblake]

Originally Posted by Rick TYler

I got some flashes of this in the last few days, but nothing I can really put my finger on.

I see that my hunch, and Rick's, about the style of future FTC games generated some pretty strong reactions.

Let me remind you that my crystal ball is very cloudy.

Do not over-react to hunches about how the tone of the games might change a bit.

Also, when I think about challenges to work into FTC games I often think about problems like building towers, or turning knobs, or punching numbers into a keypad, or about doing other tasks that would not involve a whole lot of bumping or pushing/shoving. Introducing a little bit of non-contact sports into the challenges might be a good thing.

Blake

[artdutra04]

Now that the new FTC system has been announced, I will share some of my opinions on the kit, based upon checking out the kit first hand over the past month while we built one of the FTC robots using the new kit. (Note: while I had some chances to play with it over the past month, most of my time went into working on our NI Prototype Robot with the new FRC control system.)


So here are my thoughts on the new FTC kits based on my experience so far:

The new FTC kit lacks any sort of continuity. Most of the metal parts are stamped with center to center hole spacing of 0.625 inches. The standoffs given to teams are in 0.5 increments. This makes it really difficult to make perpendicular things line up. If you want a relation in real world terms, the new FTC kit is like watching Back to the Future III without ever seeing the first two.

Most of the new parts are overkill [or underkill]. On a robot that weighs at most ten or fifteen pounds, do we really need structural members than can support fifty pounds of weight?

Kind of a sub point to this: the size (I call it overkill factor) of the new parts really makes working with Vex sized (dimensions between 12" and 18" cubed) robots difficult. The new Lynxmotion parts really seem best suited for larger robots, such as between 18" and 24"-ish cubed.

At the same time, the huge amount of NXT Mindstorms parts included in the kit really make anything larger than 12" cubed difficult to work with...

It's like Goldilocks and the Three Bears, the first one is too big and the second it too small. At least letting us use all the Vex mechanical parts in next years game will fill that "just right" size category.

It needs adapters to mate directly with Vex axles/wheels. While control system wise, the new FTC kit is ahead of the old Vex kits, mechanically, it is lacking way behind. And using the logic that it took Vex a few years to come out with the whole range of products they have today is missing the obvious point: Vex has all these neat products on the market right now. Not a year away, not two years away, but right now. And when we are limited to only three different size spur gears in the FTC kit, and a few different size regular wheels, and most FTC teams already have a bunch of Vex omni wheels, advanced gear kits, roller chain, etc. I feel like purposely trying to alienate one brand for another is quite a bad decision on behalf of the teams.


These are the main issues that I have with the new kits; as such right now I feel like this kit is a barely manageable attempt to cobble anything and everything together. The kit will work, it is neither an elegant solution nor an efficient robotics kit.

If FIRST is actively seeking feedback on this issue, and expresses an interest in trying to solve some of these annoying issues (or alter the format of FTC to match these new 'quirks'), then I will reconsider the above points. Until then, if I had to grade the new kit, I'd give it between a C- and a C+. It has potential if there are major changes to streamline and bring continuity to the kit, but until then it's just a lot of public jabs at Vex (which IMHO is quite unprofessional), while it suffers through a lot of swings and misses with its own kit. Our friends from Texas have the perfect phrase for this: all hat and no cattle.

[wilsonmw04]

Thanks for the reminder Blake. i would REALLY be disappointed if they limited the interaction/defense in the FTC game. I hope the demo was just that: a demo. If they were smart, FIRST would take all the feedback from the teams and make some improvements to the kit and make a great game for the first year of the kit.
FIRST got the price of the transition right. now let's work on the kit and the game.