Poll: Reasonable and achievable, or shooting for the stars?

[Joseph Smith]

This is something I've been thinking about for a while.
In a lot of threads, I read comments like "too many teams are going to try to do everything and fail" and "teams should focus on doing ONE task really well". While these are fair points, would it really be better if all but the powerhouse teams only attempted to build "reasonable" robots?

So here is my question:
Are students on a team more likely to be inspired and learn about the engineering process if their team decides to build a reasonable, achievable robot that performs some aspects of the game strongly?
OR
If their team attempts a more ambitious, challenging design that they feel BEST meets the objectives of the game, and ends up struggling to finish and/or perform at the level that they wanted?

At the end of the day, is it better to know your limits and not push too hard, or step up in the face of the challenge and try something awesome? After all, even if a lot of teams struggle, there will be teams out there who pull off something they never previously imagined possible.

[Jon Stratis]

Why not both?

Especially this year, if done right teams can add on mechanisms to address each challenge one at a time. You can spend the time needed to figure out shooting and get that functional, then add on a gear mechanism, and then a climber (or whatever order you prefer) as time allows.

Achieving an "everything" robot is pretty inspirational. But so is achieving a successful robot that gets you into the playoffs. I would argue (based on experience) that going after an "everything" robot and missing is worse for a team (in terms of moral, getting members to return the following year) than going after a simpler robot and being as successful as that strategy would allow.

[EricLeifermann]

Having been on teams in every spectrum of skill and resources. I can say that a robot that moves and meets its objectives is way more inspirational than a robot that doesn't do anything because the team overreached.

Set reasonable goals and expectations and do everything in your power to achieve them. Elite teams are elite because of what they do in the off season not what they do during the build season. If you want to shoot for the stars the time to do that is May-December not January-April.

[lenny8]

It all really depends on a teams ability and resources. For some teams shooting for the stars is another teams reasonable and achievable.

[Valkonn]

For the past 3 years, team 3495 had a very small design team and made very small and incremental improvements. During that time, I would have agreed with you in the "reasonable and achievable" approach. There is no point in overreaching and attempting to fulfill goals that are way out of your experience level.

For example, building a turret without any successful vision experiences is likely going to end up pretty badly. Building a fixed shooter would be the more reliable option.

However, if you stick with this mindset forever, you're going to be pumping out bots without any advancement and will always perform at a sub-par level. Speaking for my team personally, last year was the year we decided to switch things up. We used vision processing (it failed). We switched our drivetrain belts to chain (that failed, they were incredibly high maintenance). We went with an innovative drivetrain design (it failed) rather than the basic pneumatic wheel setup.

Now, those failures were pretty rough. We got crushed in the Las Vegas quarterfinals and let down our teammates. But by the end of the season we had a greater understanding of vision, chain, and drivetrains. Those failures allowed us to make huge improvements and apply them to this robot.

My point is that it's OK to fail, so long as you learn from your mistakes. You should never be content with your level of performance. The point of this program is to learn, and that means taking leaps into technical areas you know very little about. Whether you succeed or not is irrelevant.

[Chris is me]

Originally Posted by Jon Stratis

Why not both?

Especially this year, if done right teams can add on mechanisms to address each challenge one at a time. You can spend the time needed to figure out shooting and get that functional, then add on a gear mechanism, and then a climber (or whatever order you prefer) as time allows.

I was expecting this game to play out in this way, but the extreme packaging constraints this game imposes really hinders a piece-wise development strategy like this. It's very difficult to make room for everything without planning for it all at once in advance... and then you become a do-everything bot that's falling behind in build season.

---

To address the OP: I've been on teams that shot for the stars and failed. I've been on teams that did the reasonable and achievable thing. The latter is much more inspirational and rewarding, every single time. No matter how reasonable and achievable, every robot will have lessons to learn, things to tweak, etc. Maybe if you're a top tier team with your act together, you'll get bored and your kids won't learn as much, but I really think those are the teams that don't struggle to shoot for the stars anyway.

It's not just about competitive success that makes reasonable robots better - it's just a smoother process overall, and people are more inspired when they make something that works.

[Kevin Sevcik]

Originally Posted by EricLeifermann

Having been on teams in every spectrum of skill and resources. I can say that a robot that moves and meets its objectives is way more inspirational than a robot that doesn't do anything because the team overreached.

This. Aiming to over-achieve is fine if you know you have an achievable fall-back position, which is a really easy prospect this year, since a gearbot with a winch will be a viable 2nd pick. Years where the game really forces less capable teams to pick a hard, flashy, probably high points task vs. an easier, less flashy, lower points task are often painful to watch. We've had less of those as the GDC's gotten better, but ball launching in 2014 or climbing more than one level in 2013 come to mind. There were better things to do with your time and space than either of those, but committing to them didn't leave you much of a fallback.

Also, real world engineering is very much about knowing what you, your team, and your materials are capable of, and not exceeding that. Promising a client the moon and only making it half way doesn't usually end well.

[JABot67]

Knowing what is reasonable and achievable for your team or organization, and planning a great design and strategy around it, is essential to the engineering process. It would be my opinion that if members of FRC teams do not consider in their design process past experiences and educated guesses, they are not learning real-world engineering.

The way to shoot for the stars is not to dive into the depths of uncertainty, or even worse, go forward with a plan that you know has a high chance of failure. The real way is to gather up resources and knowledge and experience to the point that you are actually comfortable about tackling that next challenge. For now, the way to inspire your team to grab that next rung in the ladder is to do what you can with what you've got.

Play to your strengths. Go for awards. Have your driver learn defense. Make a simple mechanism that possibly doesn't work as well as a complex one but is much easier to repair and make replacements for. Compared to other competitions, FRC makes it relatively easy to do well with limited resources. Why not make the most of this fact?

I have too many FRC anectotes about this topic to possibly talk about in a post like this. However:

Consider 3322 in 2013. We decided to climb the pyramid, and based our entire robot dimensions on our design for a climber. This resulted in sacrifices for disc loading and shooting. The result was that we didn't get a climber on the robot and we experienced disc jams the whole season because our dimensions disallowed us from designing the shooter we wanted to design. Next, consider 2930 in 2015. We had a modified kit frame with a simple elevator and simple pneumatic grabber, and a PVC can grabber for PNW. This was by far the simplest robot design by a team I've been on. The result? We seeded 5th, 8th, and then 2nd at our district events and won the Pacific Northwest Championship. Looking back, we have learned from our past seasons and we are a stronger team that can afford to try more things, as long as we don't neglect our priorities.

By building a simple, reasonable robot, you are not "giving up". Knowing your limits proves you are not weak, but strong.

[JesseK]

I think most teams have a presumption that doing 1 thing well is less of a "challenge" than trying to do everything. Any team can ferry a few gears in a match and climb a rope after many seconds of line up and capture. It is quite difficult to build a robot which will consistently ferry MANY gears, account for ALL 10% occurrences and THEN capture/climb a rope in a minimum amount of time. It's the difference between a true expert and someone who 'dabbles'.

This year I've learned a lesson lesson that would seem obvious given that it is in plain sight, but isn't put to practice by the majority. Average teams who do not understand what it takes to optimize and test a design for realistic on-field scenarios will never be at the same level of power house teams. It is the perfect year & game design to learn this lesson, as optimization of the "easier" challenge is still quite valuable.




Now for the metaphor...

Don't shoot for the stars. It's ambiguous, has many unknown unknowns (including purpose), and is generally lonely up there. Eventually you can't even see where 'home' is and lose all perspective. Shoot for something more plausible, like the Moon or Mars. If you're particularly ambitious, orbiting Venus or Jupiter would be pretty cool too. And if you aim for those targets, don't miss - even our solar system is vastly empty and quite lonely.

Besides, I don't think our FRC batteries have enough juice to get the balls into orbit, let alone beyond orbit.

[Joseph Smith]

Some very interesting and thought provoking responses here.

In my experience, the times I've learned the most and felt the most inspired as a student on the team were when we chose an ambitious design or mechanism and put in a lot of time and effort to refine it and make it work. Even if it means struggling at the first competition or working long hours before bag, it's always been immensely satisfying to see all that work pay off.

I agree that there is nothing fun or inspiring about performing poorly at an event because you couldn't get it working in time.

In the end, maybe the best route is to choose a design that you know you can achieve and be successful with, but with some aspect that pushes your boundaries a little bit- a new mechanism or technique that you've never attempted- but have a fall-back in case it doesn't work out.

[Joseph Smith]

Originally Posted by JesseK

Besides, I don't think our FRC batteries have enough juice to get the balls into orbit, let alone beyond orbit.

But that's the beauty of it, we're burning FUEL as we go, and thanks to the Tsiolkovsky rocket equation, the more FUEL we burn, the lighter the load we have to lift!

[Caleb Sykes]

Originally Posted by JesseK

Besides, I don't think our FRC batteries have enough juice to get the balls into orbit, let alone beyond orbit.

FRC batteries store about 1MJ of energy, and it takes about 1MJ of energy to get a 100 gram object to escape velocity, so one battery has enough energy to get about 1 fuel beyond earth's orbit.

[Kevin Sevcik]

Originally Posted by JesseK

Besides, I don't think our FRC batteries have enough juice to get the balls into orbit, let alone beyond orbit.

18 Ah @ 12V = 12*18 Watt-hours * 60 min/hr * 60 s/min = 777600 Watt-s or Joules. Escape velocity is 11186 m/s. So 777600 = 1/2 * m * (11186)^2 = 12.4g. Balls weigh 70g, so you'd need 6 batteries charging your ball launching system to reach escape velocity. Not counting drag, etc. A properly designed charging system could probably send a penny into solar orbit, though.

EDIT: Wikipedia also tells me delta-v to Low Earth Orbit is 9.4km/s, so getting into orbit isn't any easier. Not to mention you need an onboard motor for a circularization burn. Unless you're using ground based lasers ablating your fuel as a "motor".

[RoboChair]

Originally Posted by Kevin Sevcik

18 Ah @ 12V = 12*18 Watt-hours * 60 min/hr * 60 s/min = 777600 Watt-s or Joules. Escape velocity is 11186 m/s. So 777600 = 1/2 * m * (11186)^2 = 12.4g. Balls weigh 70g, so you'd need 6 batteries charging your ball launching system to reach escape velocity. Not counting drag, etc. A properly designed charging system could probably send a penny into solar orbit, though.

That's the most interesting math I have seen all year.

[Kevin Sevcik]

Originally Posted by RoboChair

That's the most interesting math I have seen all year.

I bet I know what someone's planning for an offseason project....