Who wants to learn TRIZ and solve engineering problems like a Pro?

[technik3k]

There are many situations when you run into a problem with conflicting requirements and limited resources and it seems there is no solution.

You think that you need to be a genius or have an eurica moment to find an out-of-box solution and it may take years. But that is not true. Any competent engineer familiar with TRIZ methodology could do stuff that many would consider some sort of magik.

TRIZ Matrix to engineering is what Periodic Table is to chemistry, yet few are familiar with it and even fewer use it daily.

Usually, TRIZ is perceived as a boring stuff taken in one of the graduate level classes, but I found a perfect example to introduce you to TRIZ principles. It is just too good to pass up. As many of you are future engineers you will not regret learning about it. So here it goes:

TRIZ Engineering Riddle

(dramatic VRC edition)

You are at the state (or regional) Championship for 2016 Nothing But Net game.

You are a member of one of the top teams. Your club worked very hard all year long to get here. The first qualifier match begins in 15 minutes, but the rest of your team and the robot are stuck in the hopeless traffic. You run to your partner’s pits and see that they just fried all their electronics. Their robot is a cubic feet of a dead metal weighting over 20 lbs. You check on the opposing alliance - they are not the best but look competent enough to score 20-30 balls into a low goal.

This match is very important. Going to worlds may depend on it. Everybody looks at you. They know that you are some kind of a TRIZ Black Belt, but nobody knows what it is.

Your partners give you an old toolbox with random staff. You open it and see: empty tag plates, half-charged battery, cortex, controller, one motor with a sticker “PTC dies after 3 secs”, scratched piece of 3x8" plastic, quad encoder, potentiometer, some screws, a couple of bent c-channels, bunch of zip ties, rubber bands, 8 blue LEDs, 4 80k resistors, crayons, toy rubber duck, and a baby pacifier…

You have 10 minutes left. What do you do?

---

Let’s try to do a virtual class. There will be some questions, those who want could reply. I will answer questions, post more stuff. If everything goes as expected - then after a few iterations you all will solve the riddle on your own and will know what to do if it happens to you in the real life next year

The first step of TRIZ is to look at the problem without the interference the preconceived notions of what is impossible. It is not easy at the beginning. You need to write down the reasons why you think something couldn’t be done and try to spell out your reasoning. Sometimes you will see the path to a solution right at this step.

Those who think, that nothing could be done, please, post why you think that way. (if somebody knows solution, please, don’t post it yet, we need to go through a few more steps)

[Kelton]

technik3k,

I am very interested in what you’ve started here. When you say your partner for that last match has fried their electronics do you mean their cortex is broken, all of their motors just stopped working or maybe something completely different? Is there any chance that your partner could lend you parts for the match? Do you have a computer to program the cortex? I’m not exactly what the rules are on this but is it legal to have your alliance start on the same starting tile as you long as you and your alliance are not touching any grey tiles or outside of 18x18x18 in the beginning of the match? If it is I might have a possible solution.

Thanks,

Kelton

[technik3k]

It is great that nobody said that nothing could be done! There are a lot of things you can do even with 3 seconds of a single motor action.

@Kelton

When you say your partner for that last match has fried their electronics do you mean their cortex is broken, all of their motors just stopped working or maybe something completely different? Is there any chance that your partner could lend you parts for the match? Do you have a computer to program the cortex?

I like it wery much that you jump in and start thinking in terms of solutions. You can assume that everything inside your partner robot is dead and you don’t have time to debug it anyway. You have a computer, but your are limited to the parts found in the toolbox.

@Kelton

I’m not exactly what the rules are on this but is it legal to have your alliance start on the same starting tile as you long as you and your alliance are not touching any grey tiles or outside of 18x18x18 in the beginning of the match?

These are very good questions that everybody should be asking. In order to win this we need to know game manual very well. In fact the major part of solving this riddle will be to “hack” the game manual - it has a number of loopholes and we have to find them. Here is the answer to your question:

<SG1> At the beginning of each Match, each Robot must be placed such that it is touching one of its colored Alliance Starting Tiles, not touching any Scoring Object other than those permitted by <SG2> and not touching any other foam field tiles, or another Robot. No more than one (1) Robot may start the Match on any one (1) Alliance Starting Tile.

<SG2> Prior to the start of each Match, each Robot may use their four (4) Balls available as Preloads. A Ball is considered to be legally preloaded if it is touching the Robot, not touching any other grey foam tiles, and is fully within the field perimeter. (See figures 8 & 9) Any unused Preloads become Driver Control Loads.

The only deviation from the real rules in that we don’t have to worry about not being able to use real robot once it arrives. For the purpose of this exercise we will only think about doing something for this game that is coming up in 10 minutes.

The next step in the TRIZ process is to formulate the ideal solution. For us it is winning the game.

Then we will iterate backwards from the end and make it more and more specific, until we have a solution. Only if we fail to find the solution after some time we may make the goal less ideal. But I think we could win.

Now, I would like you to post your ideas on how to make statement of the ideal solution more specific.

As an example, here is an idea I just heard from across the room from technik jr:

Lets build rubber band catapults for four preloads, motor will trigger them at the beginning of autonomous. They hit high goal and get us 20 points.

I like his idea, I didn’t think of that, but there are several problems:

First of all, it gives us at most 30 points (4 * 5 + 10 auton bonus). The opponents will likely score their 30 balls, some of which will be 2 pt orange bonus balls, so it is 30+ pts and they win.

Second problem is that it is very hard to build such a contraption in 10 min. Scoring 4 preloads into the high goal, without any testing. Really? We need something simpler.

The final problem is that he jumped a little bit ahead of himself. He was thinking from the beginning, and we need to think back from the end.

So lets rephrase the question: what are the ways we could score to win, regardless of how we might do it?

The correct answer for this TRIZ step would be “4 preloads into high goal + 10 auton bonus => 30 points”.

This is not enough, we need better. Are there any other answers?

[Kelton]

Well to only ways to score are the auton bonus for 10 points, putting balls in low goals, putting balls in high goals, low lifting for 25 points, and high lifting for 50 points. I think an important thing to note is that if you are lifting your alliance you only have to do it once. What if you use your 2 c-channels to make sort of a 1 stage “scissor lift” and use the rubber bands to push the 2 pieces of c-channel together which would push up the scissor lift. Put a pin where the 2 c-channels cross and mount the motor so it can pull or push the pin out. The motor would probably be able to pull the pin out in less than a second so it wouldn’t matter that the motor burns out after 3 seconds. When the pin would be pulled out the scissor lift would spring up.

Next all you would have to do is talk to your alliance and have them just bolt a backup battery and plates to just their fried cortex nothing else. That way under vex’s rules they would have a “robot” that could get through inspection even though essentially it does nothing and would weight less than a pound so it could be easily lifted. Then just attach your alliance’s “robot” to your scissor lift and high lift them for 50 points beating your opponents 30 point score. The only problem with this idea is that I’m not sure to make the 1 stage scissor lift from tipping over. Since we don’t have an axle it may be difficult to have the motor pull out the pin. Anyone have any thoughts on those two problems?

There is probably a better solution but that’s what I came up with. I apologize if my explanation is unclear. If you have any questions let me know. I’m curious to see what other will come up with.

Thanks,

Kelton

[Tim]

I don’t see us building a launcher or even a drive base with any of those parts so I think the simplest way is somehow high elevating ourselves over our dead alliance partner. I can’t think of a way to self-propel ourselves 12" high though.

[puzzler7]

I don’t have a solution yet, but LEDs and resistors are useless, because you don’t have time to debug(LEDs) and you can’t modify Vex electronics(resistors).

The crayons, rubber duck, and pacifier can’t be in any way functional, so those are useless.

The quad encoder and potentiometer are next to useless, because you only have 3 second of one motor, not much time to measure anything, certainly not enough time for a PID. You don’t have much time to program anyway.

That leaves us with a cortex, plates, controller, battery, 3 sec motor (assuming the sticker doesn’t lie), 3x8 plastic, c-channels (bent into shape by being hit with the battery), screws, zipties, and rubber bands.

I like the idea of pulling a pin, and rubber band powered jump onto the other robot, but I’m not sure how that would be done.

Also, do we get nuts? Because otherwise the screws don’t screw anything. What about a toolkit, or an axle, so the motor can spin something?

[technik3k]

@Kelton

Well to only ways to score are the auton bonus for 10 points, putting balls in low goals, putting balls in high goals, low lifting for 25 points, and high lifting for 50 points.

Ok. This is what we are looking for. There is a winning strategy. We need to get (50 pts high elevation + 10 pts auton bonus) => 60 points total. This way we win even if the opposing alliance scores all 30 balls, including 10 bonus balls for a total of 40 pts.

@Kelton

I think an important thing to note is that if you are lifting your alliance you only have to do it once.

This is very good observation. We will need it later.

Now we are ready for the next step of the TRIZ process: brainstorm and write down all possible paths from the beginning to end goal.

The crucial difference from the usual brainstorming process is not to limit ourselves to paths that seem to work, but to include any paths that have some “debris” in the middle (those “debris” are the conflicts or contradictions between elements of the system). Learning not to throw out such paths is one of the most important skills of TRIZ master because TRIZ works by analyzing and then resolving or moving those conflicts somewhere else where they will be less of an obstacle.

Luckily FarceSolid already found that for us:

@FarceSolid

I don’t see us building a launcher or even a drive base with any of those parts so I think the simplest way is somehow high elevating ourselves over our dead alliance partner. I can’t think of a way to self-propel ourselves 12" high though.

There is a conflict right there: we need to be 12" high but there is clearly no way to lift the robot with resources we have.

Speaking of resources, knowing the rules of the game is extremely important when you face such a dire situation. I would say that perfect knowledge of NbN Game Manual is worth two additional motors, maybe even four.

Another, important skills of TRIZ master is not to get caught in the “language trap”. You need to understand what exactly needs to be done - not what everybody assume needs to be done based on the language used to describe the problem. Here is the hint:

High Elevated – A Robot is High Elevated if all of the following criteria are met:

1. It is touching the other Robot on its Alliance
2. The Robot that it is touching (see criteria #1), is entirely within the Climbing Zone
3. It is not touching any Field Elements, excluding the field perimeter.
4. The entire Robot is completely above the plane parallel to the foam field tiles, formed by the top of the field perimeter.

Please, tell me what is the language trap that everybody gets caught in?

What could we gain by escaping it and where would the conflict move?

---

@Kelton

Since we don’t have an axle it may be difficult to have the motor pull out the pin.

@puzzler7

The quad encoder and potentiometer are next to useless, because you only have 3 second of one motor, not much time to measure anything, certainly not enough time for a PID. You don’t have much time to program anyway.

That leaves us with a cortex, plates, controller, battery, 3 sec motor, 3x8 plastic, c-channels (bent into shape by being hit with the battery), screws, zipties, and rubber bands.

Also, do we get nuts? Because otherwise the screws don’t screw anything. What about a toolkit, or an axle, so the motor can spin something?

You are jumping ahead of our exercise, but I love the way you are prioritizing resources, including your own time. It is very important to think in terms of what is available here and now instead of what could or should have been.

Btw technik jr tells me to charge the battery first. That’s a good idea too.

Also some clarifications: there are no nuts, but you can borrow them if necessary.
Partner robot is game legal. But the only thing it is capable of doing is to weight 20 lbs.

[StoweBots]

@puzzler7

The crayons, rubber duck, and pacifier can’t be in any way functional, so those are useless.

I don’t know the humour value alone of seeing a rubber duck fly into the high goal must be worth something.

@technik3k

You run to your partner’s pits and see that they just fried all their electronics.

When you say all electronics, do you mean motors as well, if so then there is I think only one useful solution (beyond spitting the dummy/pacifier).

[technik3k]

@StoweBots

I don’t know the humour value alone of seeing a rubber duck fly into the high goal must be worth something.

When you say all electronics, do you mean motors as well, if so then there is I think only one useful solution (beyond spitting the dummy/pacifier).

The pacifier is the backup object of the last resort.

As I said before the partner robot is the game legal cubic feet of the dead weight. Hint: use it as it is, all parts that you will ever need are in the toolbox.

[StoweBots]

@technik3k

The pacifier is the backup object of the last resort.

Joking aside I have a solution, but not sure if you want solutions yet?
It’s a one hit wonder attempt for them to score goal points, then just use their ‘dead weight’. Pretty sure I could slap my idea together in 10 mins.

As I see it, in that last 30 seconds all partner robots (working or not) are ‘dead weight’ and if you intend to elevate anyone you better build to lift anyone. Which I now feel a little silly for not weighing our skyrise robot before we pulled it apart, but I could guess from the bits we used what it weighed. Anyway my point is, if you intend to just use the other dead robot to score points by elevating it you should already be capable of doing that no matter their condition.

[jpearman]

I’m so tempted to post a solution, but I won’t.

High Elevated – A Robot is High Elevated if all of the following criteria are met:

1. It is touching the other Robot on its Alliance
2. The Robot that it is touching (see criteria #1), is entirely within the Climbing Zone
3. It is not touching any Field Elements, excluding the field perimeter.
4. The entire Robot is completely above the plane parallel to the foam field tiles, formed by the top of the field perimeter.

To get the 50 points we need two robots. There’s enough material to build one, but 10 minutes before the first match inspection will already be closed, oops

[StoweBots]

@jpearman

I’m so tempted to post a solution, but I won’t.

To get the 50 points we need two robots. There enough material to build one, but 10 minutes before the first match inspection will already be closed, oops

Me too!

Although I don’t see the point in building a new robot, dead weight is dead weight. I want to use some of the parts to create a way for them to score points during autonomous, before you simply lift the dead weight.

[Kevin_Boenisch]

In addition to the parts in the box, are we allowed to borrow other parts from other teams or your alliance partner? There are always teams happy to loan a few parts, and if worst comes to worst you could salvage parts from your partner’s dead bot (not that they’d like that!).

@jpearman

I’m so tempted to post a solution, but I won’t.

To get the 50 points we need two robots. There enough material to build one, but 10 minutes before the first match inspection will already be closed, oops

At many regional competitions (at least here), inspection is open for a while longer than the first match, and robots that haven’t had enough time to be inspected are allowed to play a match and be inspected as soon as it’s possible. It technically isn’t allowed, and I’m sure it wouldn’t be done for robots that are decent to good or questionable in terms of legality, but this here probably won’t be an amazingly strong robot.

I want to post a solution as well, but won’t yet.

[jpearman]

@StoweBots

Me too!

Although I don’t see the point in building a new robot, dead weight is dead weight. I want to use some of the parts to create a way for them to score points during autonomous, before you simply lift the dead weight.

My point was that to have a scoring elevated robot, you have to have two robots, one for each of the two alliance teams. So you have to build one, a motor, cortex and C channel can technically be called a robot, but I was pointing out that even if you build this then it’s unlikely to get inspected close to matches starting. But lets ignore that for this exercise.

[FullMetalMentor]

@jpearman

… So you have to build one, a motor, cortex and C channel can technically be called a robot…

Ditch the motor and plug in an LED instead, and *voilà *you have a photon drive!

[StoweBots]

As far as I can tell the point of this TRIZ exercise is to go for a solution that is not considered simple, but one that is the best possible outcome and can work with a few basic moves.
I think it is possible with the parts in the tool box for their robot to score balls in the goal, but for now I’m off to bed. I’ll check in tomorrow to see if solutions are ready to be excepted.

[anon50541363]

@technik3k

TRIZ Engineering Riddle

(dramatic VRC edition)

You are at the state (or regional) Championship for 2016 Nothing But Net game.

You are a member of one of the top teams. Your club worked very hard all year long to get here. The first qualifier match begins in 15 minutes, but the rest of your team and the robot are stuck in the hopeless traffic. You run to your partner’s pits and see that they just fried all their electronics. Their robot is a cubic feet of a dead metal weighting over 20 lbs. You check on the opposing alliance - they are not the best but look competent enough to score 20-30 balls into a low goal.

This match is very important. Going to worlds may depend on it. Everybody looks at you. They know that you are some kind of a TRIZ Black Belt, but nobody knows what it is.

Your partners give you an old toolbox with random staff. You open it and see: empty tag plates, half-charged battery, cortex, controller, one motor with a sticker “PTC dies after 3 secs”, scratched piece of 3x8" plastic, quad encoder, potentiometer, some screws, a couple of bent c-channels, bunch of zip ties, rubber bands, 8 blue LEDs, 4 80k resistors, crayons, toy rubber duck, and a baby pacifier…

You have 10 minutes left. What do you do?

Forget TRIZ, or whatever, in CS we basically get a degree in math and problem solving so here goes…

First I strip out all the useless crap from this problem…

@technik3k

The robot [is] hopeless. This match is very important. Partner’s …] just fried all their electronics. Their robot is a cubic feet of a dead metal weighing over 20 lbs.

[You have]
empty tag plates
half-charged battery
cortex
controller
one motor with a sticker “PTC dies after 3 secs”
piece of 3x8" plastic
quad encoder
potentiometer
some screws
a couple of bent c-channels
bunch of zip ties
rubber bands
8 blue LEDs
4x 80k resistors
crayons
toy rubber duck
baby pacifier

You have 10 minutes left. What do you do?

I build a mechanism to get my robot on top of the other one inside of 3 seconds in auton. Win the auton bonus and claim the high lift points.

IF I HAVE TIME, I attempt to create a launcher that can function exactly once, firing one preloaded ball hopefully into my goal.

As for the mechanism, since I’m only lifting a Cortex and battery it could be a simple c-chan arm that is designed to hook onto the other robot by you know… bending it more (the chan’s already bent, who cares what happens to it now, right?).

[jpearman]

@FullMetalMentor

Ditch the motor and plug in an LED instead, and *voilà *you have a photon drive!

We derailed the thread, I should have kept quiet, and blue LED’s are illegal as they are not official VEX parts.

[anon50541363]

@jpearman

We derailed the thread, I should have kept quiet, and blue LED’s are illegal as they are not official VEX parts.

This is a thought experiment anyway, without a backup battery (which was not provided), your robot would never pass inspection.

Also, I could pass off 8x LEDs as “non-functional decorations.”

As for the photon drive, that’s pretty freaking funny.

[jpearman]

@cody

I build a mechanism to get my robot on top of the other one inside of 3 seconds in auton. Win the auton bonus and claim the high lift points.

IF I HAVE TIME, I attempt to create a launcher that can function exactly once, firing one preloaded ball hopefully into my goal.

As for the mechanism, since I’m only lifting a Cortex and battery it could be a simple c-chan arm that is designed to hook onto the other robot by you know… bending it more (the chan’s already bent, who cares what happens to it now, right?).

You spoilt the fun Don’t forget those team ID plates and the flag holder (and backup battery, do we have one of those?) and we don’t have VEXnet keys, or do we?