QOTW 08-24-03: Old School

[JVN]

Maddie has asked me to step in and post the QOTW while she is on vacation. Unfortunately I couldn't think of anything good, so I asked for help. Because of this, this week's question is brought to you by our good friend Andy Baker.

Andy wants to know:

Question of the Week 08-24-03: Over the years, FIRST has loosened up robot build restrictions. They have changed the robot design challenge. 4 years ago and before building a robot was somewhat of an "Apollo 13" challenge, where we had this pile of "stuff" and we were to build a robot out of the "stuff". During the last 4 years, we have been able to use many more materials and parts. Now, we can build a robot out of whatever we want (within reason and limitations of some common components).

What do you think of this change in robot building philosophy?


Well folks, what do you think?

[piotrm]

I think its great. Opening it up a bit makes designing and building a robot less like solving a puzzle and more like .. well "designing and building a robot".

[Jeff Waegelin]

I think it was a great idea to relax the restrictions. Many of the additional parts restrictions were designed to make things more fair for teams with less resources, but as Dean said at Kickoff, life is not fair. There will always be teams with more resources, and trying to limit them to make things more fair never did much to even things out, anyways. Now, with less restrictions, teams are more free to use what they want, but the $3500 cap puts a reasonable spending limit on deeper-pocketed teams. All in all, I think it is a good decision that will allow for more creative robot designs, generate more student involvement, and perhaps do a better job of making things easier for all involved.

[kmcclary]

Originally posted by JVN
Andy [Baker] wants to know:

Question of the Week 08-24-03: Over the years, FIRST has loosened up robot build restrictions. [...] Now, we can build a robot out of whatever we want (within reason and limitations of some common components).

What do you think of this change in robot building philosophy?

This is much closer to industry, but we're not quite there yet.

I think it is very fair to allow hardware store stuff while still limiting "exotic" materials that give deep pocket teams an unreasonable "strength to weight" advantage.

However, that said, I still wish FIRST would open up the TECHNOLOGIES, especially electronics, and pneumatic controls.

For the new people: The current build rules have separate restrictions for dollar spent on electronics vs mechanics ($400 vs $3500 overall), and NO pneumatic plumbing nor control components may be added over the base kit. Electronic components and CPUs must come from one of a couple of vendors, which don't even give us a discount.

IOW, currently we can spend $2000 on a complicated drivetrain, but we can't even spend 50 cents for another pneumatic elbow... We can buy a tire from anywhere, but we must purchase electronics from only one of couple vendors. We can't assist a relatively weak Robot Controller in its Autonomous Mode with outboard CPU assist designs, unless we're both VERY frugal financially, AND all of the parts used happen to be stock items with only a few specific vendors.

This doesn't make sense to me at all.

Is a goal of this program to teach students how it really works "out there"?

In the "real world", you find things wherever you can, and trade off technologies all of the time for cost/weight/time savings. Many industries routinely replace complicated mechanical designs with simple micros, sensors, and actuator controls to create far superior products. Look at how much simpler automotive fuel injection is to complicated carburetors. All of the mechanical "calculations" previously done with myriad parts of complicated shape are now done totally in software, which allows for rapid modifications as the design evolves.

IMHO, I strongly feel it would be a much better preparation for the real world to simply place a UNIFIED weight, dollar cap, and given actuators (e.g. keep the "$3500 over kit" and current weight and actuator/motor limits), but allow that $3500 to be ANY mix of electronics, mechanics, pneumatic controls, optics, (etc.) that the team wishes, from wherever they can find it, AS LONG AS the documentation of that cost is auditable.

IOW, If a team can somehow figure out how do this contest with simply the kit materials and a light weight $1500 minicomputer with a suite of software, or run those limited pistons with complicated pneumatic plumbing controls without going over budget, GREAT. I feel we should let them do it instead of penalizing them by artificially favoring a close to pure mechanical design. Favoring one technology over others restricts innovation, and prevents our students from seeing how REAL industry does control design trade-offs in the information age.

Let's make next year's contest a REAL battle of technological innovation, by simply limiting actuators, weight, and power, making a unified dollar cap, and throwing the technologies open! Instead of segmenting the dollar and vendor limits by technology, let's see what the electrical vs mechanical vs pneumatic wizards can REALLY do on a level field! I'd be excited to see the robots created when the various team & sponsor combinations can REALLY tap their design strengths!

Just my $.02...

Can anyone from FIRST please comment on why we have these artificial limits to the electronics and pneumatics now? There may be an excellent reason that I simply don't see from this perspective. Thanks!

What do you guys think? Should the contest be open to all technologies equally? (Or maybe this should be another week's question???)

- Keith

[Nate Smith]

Originally posted by kmcclary
Can anyone from FIRST please comment on why we have these artificial limits to the electronics and pneumatics now? There may be an excellent reason that I simply don't see from this perspective. Thanks!

I'm not an "official" word on this, but from what I do know, it comes down to one thing: safety. The pneumatics kit is not a random set of parts like much of the rest of the kit is, but rather all of the fittings, etc. are hand selected to work together. As for the electronics, if an "anything goes" solution was allowed, FIRST would lose much of the safety/fairness/rule enforcement controls(emergency cutoff/timer shutdown) that they have with the current system.

However, in both of these cases, there are possible solutions. I propose the following:

Regarding Pneumatics:
I agree with FIRST that some limitations need to be placed on pneumatics for obvious safety reasons. Because of that, I propose that the current PSI limits and connector methods provided in the kit and rules remain. However, I propose the following: on the operating side of the pneumatics(60 PSI max), ANY pneumatic device whose primary purpose is not destruction(no pneumatic saws) is allowed if and only if the connection components(quick connects, elbows, etc) provided in the kit OR IDENTICAL PARTS purchased by the team were used to connect the device. NO MODIFICATION to off-the-shelf pneumatic components in regards to range/type of motion or pneumatic connections would be allowed.

Regarding Electronics:
In an attempt to open up the electronics capabilities of teams, I propose the following:

The existing control system be overhauled to consist solely of the following components:
-Radio modems(similar or identical to current)

-Operator Interface Block(OIB) - A provided electronic component which would connect to the radio modem and competition control cable, then providing a documented standard I/O stream on one or more ports, allowing a choice of interface options. An operator interface module similar to the existing system could be provided if teams did not want to design a complete custom system.

-Robot Interface Block(RIB) - Connecting to the radio modem and the PCM on the robot, this would provide a standard interface for the robot side control/logic systems. It would provide a standard means of connecting whatever logic systems a team decided to use. If a team did not have the resources or the desire to create a custom system, an addon to the RIB would be available which would expand it into roughly the equivalent of the current Robot Controller.

-Power Cutoff Module(PCM) - Mounted on the robot, and connecting to the RIB device, all speed controllers/relays on the robot would be required to recieve power from this device. This would allow for a safety or timer cutoff by FIRST, still giving them control over when you can control your robots on the field.

What do you think?

[Andrew]

At least they didn't require that the custom electronics must reside in a project box this year! That's progress.

I agree with the "ANY PRODUCT, ANY VENDOR, $3500 max" idea and hope we keep going in that direction.

[Kevin A]

I think that the kit of parts is good. It includes everything to make a basic remote controlled robot.

Any vendor should be allowed, especialy with the autonomous mode coming into play. $675 for a digital compass? Please...thats just pitiful.

I agree that safety should come first, and if FIRST lets the control systems out of their hands, they just can controll it.

[ChrisH]

On the contrary, I'd much rather see a more restricted list of components. Maybe it's just the industry I work in, (aerospace vehicles) but if it ain't in the spec book, it doesn't go on the bird.

Only tested, proven, well known technologies or materials are allowed. Any testing pretty much has to have been done by my company or one of our partners or it is ignored. Which means that often worthwhile devices or materials are excluded because we don't feel like spending the money or effort to test them. Qualifying a new material, or even an old material made a new way, takes a minimum of $1M. So our material list is pretty darn short.

So much for throwing it wide open so "we can be like industry". In real life there are often restrictions on materials etc. that are even more stringent than those in FIRST.

Is it easier to be able to use a wide variety of "stuff" to build your dream robot? Heck Yah! But a major thrust in FIRST is innovation. Innovation is a sister of Invention, which is born of Necessity.

If you really need that function on your robot, you will figure out a way to make it happen. But you may have to make innovative use of a component to do it.

In my opinion, if the rules don't pinch, they aren't doing their job.

[Adam Y.]

We can't assist a relatively weak Robot Controller in its Autonomous Mode with outboard CPU assist designs, unless we're both VERY frugal financially, AND all of the parts used happen to be stock items with only a few specific vendors.

Actually that should be fairly easy knowing what I now know about the pic microcontrollers. I am almost certain that I have heard of teams this year of using pic's on there robots to help there Robot Controller out. And considering the fact that Digi-Key, which was one of the suppliers this year, carries over 4000 of them which are usually no more expensive than a few dollars. The only real expenses with them are the programmer and the compiler (unless you like assembly) which should not count toward the final amount since they don't go on the robot.

[MisterX]

I tend to disagree with the whole pneumatics thing cuase what my team did last year was reverse the pneumatics so it created a sucking for our suction plate which worked great in practice and is in it self completly safe however with somethign being used in a manor it wasn't designed for there is a some safty issues an issue that was easily solved by a $.15 piece that we found however none of the allowed companies sold it so we had to use a different piece in yet another manor not intended so what could have cost $.15 (one little valve, weighing minimum ) turn into a $5.00 piece that weighed 1/8 pound , but we needed 4 that added up to 1/2 pund which doesn't sound alot but is when tettering on the edge like we were. ANy way at regionals the make shift valve caused this huge hub-bub at regionals between the judges and the kid and engineer that designed the elaberate sytem and had we actually been a contender at regionals we wern't we would have had to take them off therefore destroying our robots dedfense ( and it was a ramp defender). Luckily the judge was nice enough to let it slide ( THANK YOU!) and it made no difference at all and we finished 2nd to last but that doesn't matter. The put is if we had addecuate supplies none of that would have happened!!!!!!!!!!!

P.S. the whole scheme right there is pretty much shortened down a lot and I could go on for ever discussing that one piece and all of its hasle but I dont want to waste space on the forum so just message me with any questions or comment that are directly related to the piece.

[kmcclary]

Originally posted by Nate Smith
I'm not an "official" word on this, but from what I do know, it comes down to one thing: safety. The pneumatics kit is not a random set of parts like much of the rest of the kit is, but rather all of the fittings, etc. are hand selected to work together. As for the electronics, if an "anything goes" solution was allowed, FIRST would lose much of the safety/fairness/rule enforcement controls(emergency cutoff/timer shutdown) that they have with the current system.

Done right, safety should not be a problem.

Pneumatics

With air, all you need is the rule: "All added pneumatic components must be rated for the pressures involved". No sweat. Considering that FIRST already literally uses hardware store parts (such as the basic brass T-blocks which as air couplers are rated WELL above the working pressures we're working at), not allowing ANY extra fittings seems an excessive restriction.

A more reasonable one may be something like: "all additions will either come from [these vendor(s)' line(s)] or else be shown to pass hydro static pressure test <air industry org spec #xxx>". Heck, at least let us add more identical VALVES from the SAME vendors... We have more pistons allowed than valves right now!

Electronics

You don't need a new "Power Cutoff Block". You can use a Spike for that.

I instead propose the following safety rule changes:

Rule [X]: "Custom electronics may move motors, actuators, and RC servos, or interrupt the path between [Victor/Spike] and it's driven component, PROVIDING that ALL of the below are met:
A) ALL motive or actuation power for devices or interfaces driven by custom electronics must be DIRECTLY DERIVED from either the output of a Speed Controller or a Relay Module; AND
B) If custom electronics are used as a current controller or interrupter, the current and voltage specifications of that circuit shall meet or exceed that of the equivalent IFI control normally used. Documentation showing compliance to this spec shall be provided to the inspectors upon request.

Comments:

Rule X.A - You don't need a new "power interface component" from IFI for safety. A Spike is already a great part for this. This new rule would INSURE that electronics CAN'T move the bot under any shutdown condition. This is also easily verifiable by power wiring inspection, without inspectors having to understand the custom electronics at ALL.

Rule X.B - Allows for things like a cutoff circuit to be placed in series with a motor or solenoid and it's IFI driver [Spike/Victor], for faster toggling than the RC can provide (important in some applications).

This, and/or the "additional air parts allowed" rule, could allow things like pneumatic servo valves to be either purchased, or created out of the existing valves with PWM techniques. Either way, it opens up a WHOLE new design arena!

I do like your thinking about creating a general control architecture for ANY robot controller to be used in place of the IFI one, but I'm not sure if that's politically viable, at least in one step.

I do feel we need a MUCH better RC CPU (or opening up the electronics budget) next year, to best accomplish Autonomous Mode.

Yea "Adam Y.", we thought about PICs, and you can do a lot with them. But given only six weeks from the time we found out we needed a PIC expert, it turned out we didn't have one on hand nor time to GROW one. So, we were basically stuck with the RC. Add to that we were a Rookie team with limited experienced personnel, IMHO our Autonomous Mode control system design was painful, crude, limited in scope, and the path was prone to "drift".

Originally posted by ChrisH
On the contrary, I'd much rather see a more restricted list of components. Maybe it's just the industry I work in, (aerospace vehicles) but if it ain't in the spec book, it doesn't go on the bird.
[...] In real life there are often restrictions on materials etc. that are even more stringent than those in FIRST.

Well, let's be fair here. You're talking about "life critical design" applications, an EXTREME design case. This is normally ONLY seen in places like medical, military and aerospace industries. It's one thing when you're where a hardware or software breakdown is fatal either immediately or because you can't get spare parts for the entire life of its mission. That's IMHO an extreme case of "real world", and is not how MOST industries work. (Besides, if you read the fine print on virtually ALL of our part spec sheets, every one of them has something about "not authorized for life critical applications", so we couldn't design for that extreme even if we wanted to.)

IMO, This is a LOT different design situation. I feel this is an "educational exercise for the design and creation of a 'consumer level' product". After all, it only has to run for minutes at a time, and a hour or two at the most over its ENTIRE lifespan. Spare parts, tools, and workers are available on site for maintenance, after every two minutes of operation if necessary.

IMO, here quick repairability in a student's design is much more important to stress than MTBF. As Team 71 did last year with their file card arm crawler system, you may even DESIGN IN and EXPECT wearout of parts virtually every round!


IMO, this contest's model is: "build it on time, within budget/size/weight, make it good enough to perform (and survive) at least two minutes at a time, have it be VERY repairable 'in the field', and don't let it hurt anyone". As long as its breakdown or failure modes are not a threat to participants and viewers, and we meet the above, I'd still go for open part sources.

Besides, I feel that "a student often learns more by something that breaks under usage than by an immediate success."

Also consider: In most industries, if it NEVER breaks, it MAY have been seriously overdesigned for THIS application, and may have room for cost/weight/labor saving redesign to remain competitive. Yes, huge MTBFs are critical in your industry, but with MOST automotive, consumer (et al) products, the OPTIMUM is NOT to design to perfection, but to design to "lifespan expectations and for repairability, without incurring excessive design and/or manufacturing costs".

BTW, At the first major company I worked at after school, I had to have my UM college training of "pour time and money into it, and hold onto it until perfection is achieved" literally beaten out of me as some kind of an "attempt to bankrupt the company". (...and the floggings shall continue until morale improves...)

The painful real world lesson I had to learn was "Working well is one thing, but if it already far exceeds it's design parameters and expected use lifetime, to heck with perfection. It's time to LET IT GO so we can get it into production!"

Honestly, I've STILL found myself dealing with THAT last one in every FIRST contest to date! (Hey, YOU try to take the robot away from the Mech Assembly crew so we can wire and program it... <chuckle> It's always a tug of war. If you let them, I'm sure they'll keep completely redesigning the mechanics on you until ship date!)

You do have an excellent point though. If this WERE a life critical application, I too would be worried about component specs and if the design rules were not stringent enough. If a student's life or safety hinged on their machine working, this would be a whole different ball game!

Hey... Let's simulate a "life critical app" next year... Let's put all of the HPs in carnival 'water dunk tanks', and have the robot perform (and defend their HP against) a release ritual on the push plate!

- Keith

[Daniel Brim]

Originally posted by Jeff Waegelin
but as Dean said at Kickoff, life is not fair. There will always be teams with more resources, and trying to limit them to make things more fair never did much to even things out, anyways. Now, with less restrictions, teams are more free to use what they want, but the $3500 cap puts a reasonable spending limit on deeper-pocketed teams.

This gives me an idea (tell me if you think this is ok). Why not have teams donate to FIRST dollar-for-dollar on what they spend. If they want to spend more than 3500, why not, but this way they know that spending to much would not be a good idea. Certain expensive items, such as victors, should be left out. The money that FIRST gets would be spread out to all of the rookie teams (and second year teams if there is enough).

[meaubry]

The philosophy and rules didn't really change that much from before. Back when we had a kit of "stuff", we still had an addional hardware list. Every year it changed in content - some new, some fell off, new limitations on quantity and such. We were more restricted as to where you could get the "stuff" from. It was determined that the additional hardware kit should be opened up to include more "stuff", the decision where the "stuff" could be purchased from was also revised.
I've always liked the fact that the list was expended to include more "stuff" and where you can get it from - as it allowed for more flexability in developing a working robot within the 6 week build time frame. I doubt that limiting the "stuff", where you get it from, and even the K3 Rule from this year, will ever level the playing field as some thought it could/should/would. It is a different challenge to build a competitive robot from a more restrictive kit of parts while maintaining some aceptable level of divisity among them. In other words, a more flexable kit of parts along with a game that doesn't lead to a single design solution and will encourage teams to be more creative. That is what I love - different creative solutions to the same problem. So in conclusion, I love what FIRST has done. In fact, this year demonstrated what happens to robot diversity when the kit includes so much "stuff" that far too many end up looking exactly the same - oh, and the game design didn't help that either.

[Wetzel]

Also keep in mind that rules have to be enforced. The inspectors have been volunteers, and will continue to be volunteers with little 'offical' training other then probally worked with a team. I inspected last year at San Jose and LA, and even though I had been on a team for 4 years and was active in discussing diffrent designs and the rules with people on many teams, there were still quite a few things I had to ask teams to explain/justify.

It would be easy to say have a class the day before to teach the inspectors, but then you wouldn't have very many inspectors.



Wetzel
~~~~~~~~~~~
More then one side to a story

[sevisehda]

It seems like I have one of the first decenting opions. Pretty much now a team can put just about anything on there bot except for depleted uranium and a few other exceptions. This doesn't necessarily mean better bots and different bots. If you take a look at some of the 'old school' bots they are built to the same quality as there contemporary conterparts. The uses of materials on older bots was creative and imaginative. Although today teams CAN use alot of materials it seems most bots use the same materials(cough 8020) as everyone else. In the 'industry' most of the time there are limits on materials. Often companies only do business with certain suppliers. Or like someone pointed out only use tested materials. On the electronics side often there is a technology cap that stops you from upgrading components after a set time. FIRST is somewhat mimicing the real world by emplying bounds.

Pneumatics: Hands down its mainly a safety issue. If teams could make there own or modify components there would be a risk of breakage. My suggestion is FIRST have a catalog of approved pneumatic components in which teams can buy more of items.

Electronics: Increase the budget. Maybe something like this, 1000$ Max for electronics, 3500$ Max for materials, 4000$ Max Total. No I didn't do the math wrong, this way there is a 'shared' pot between materials and electronics. While still maintaining a lower budget for teams with limited resources.