Fabrication

[EricH]

Originally Posted by Brandon_L

Try telling your potential boss "I'd like the job, but you're going to have to hire my mentor as well".

Usually, they already did. Most companies will have a time period where new hires have a mentor to help them adjust to the company.

If you have an infinite supply of scrap material, that's one thing. But, not every team will have that. There are teams out there that have to buy all their material. And, some of those teams are on a shoestring budget (as in, before they buy material, they have to look at can they afford it, or can they find a donation). For those teams, making it right the first time--or screwing it up so it CAN be used--is important.

Now, regarding it being cheaper for you: I also said that you should factor in labor. With materials being a non-issue, let's assume that each mistake or miscalculation costs you 15 minutes of 2 students working to fix it, on average. That's about 1/2 man-hour (that would otherwise not be used), and if one man-hour is 8 bucks, that's 4 bucks for an average mistake if everyone was paid. I'm not going to ask how many you make per season, because there are a lot--but let's assume that over a full build/competition season you make 100 mistakes to make calculations easy. $400 for fixing mistakes, in the entire season. Not too bad, unless you're on a shoestring budget. (Yes, I AM AWARE that all time in FIRST is volunteer time, so all labor is free unless you're dealing with an outside shop. I'm also quite aware that I didn't factor in major "Oh great that part will not work at all" redesigns, which take substantially longer.)

This isn't the real world: but it is meant to reflect the real world. Your team is a business, or can be looked at that way. If your team was a business, with paid employees, you would have far less room for error. Some planning before you build is necessary.

[MichaelBick]

Originally Posted by Brandon_L

Its cheaper for us. We don't have the money to send parts out to be made. We do have a pile of scrap metal and nearly infinite wood from our schools wood shop. Its what works for us. This is not meant to be a "Real world engineering experience". This is a robotics competition with teams from high schools made up of high school students that are mentored by real world engineers (be it we don't even have those). Try telling your potential boss "I'd like the job, but you're going to have to hire my mentor as well". In a real world engineering situation, CAD and then sending parts to be made would be the best choice. But this isn't the real world, and not all of us have the luxury of money. There is no right or wrong way to build a robot during build season. It depends on a teams situation what is cheaper. Don't assume everyone is in the same financial boat as you.

Still, CADding should be done. Unless you also have an infinite amount of students, you still have time constraints. How much time do you waste when you have to remake parts? Also, you don't always need to pay a shop to do things for you. Many will give you sponsorship.

[nerdherdmember]

Originally Posted by DinerKid

Because we need our machines to be running all the time and they are often the rate limiting step in the robot construction process we have started using quite a bit more ABS in our designs, it machines like butter and is incredibly strong.
~DK

Could you explain how you use ABS? Team 422 uses aluminum C channel, Angle, and 80/20 for almost everything, and I would be interested to hear about what different materials teams use, and how. I also spent a lot of time talking to 1829 about their use of carbon fiber at the Virginia Regional, so my interest in odd build materials can be considered piqued.

[Anupam Goli]

I'm trying to encourage my team to use more aluminum tubing and composites. 80/20 is really heavy and though easy to build with, does not make the best frames. Aluminum tubing, when proper weight reduction holes are drilled, is very strong and can be very light.

[Brandon_L]

Originally Posted by MICHAELABICK

Still, CADding should be done. Unless you also have an infinite amount of students, you still have time constraints. How much time do you waste when you have to remake parts? Also, you don't always need to pay a shop to do things for you. Many will give you sponsorship.

We do CAD, I wasn't saying we don't. My point was sending parts out to be cut doesn't work for everyone.

Now, regarding it being cheaper for you: I also said that you should factor in labor. With materials being a non-issue, let's assume that each mistake or miscalculation costs you 15 minutes of 2 students working to fix it, on average. That's about 1/2 man-hour (that would otherwise not be used), and if one man-hour is 8 bucks, that's 4 bucks for an average mistake if everyone was paid. I'm not going to ask how many you make per season, because there are a lot--but let's assume that over a full build/competition season you make 100 mistakes to make calculations easy. $400 for fixing mistakes, in the entire season. Not too bad, unless you're on a shoestring budget. (Yes, I AM AWARE that all time in FIRST is volunteer time, so all labor is free unless you're dealing with an outside shop. I'm also quite aware that I didn't factor in major "Oh great that part will not work at all" redesigns, which take substantially longer.)

Again, this isn't the real world. We don't pay for labor. If it were real world, I agree 100%.

[EricH]

Originally Posted by Brandon_L

Again, this isn't the real world. We don't pay for labor. If it were real world, I agree 100%.

If I have to tell you to read the whole post, I will. And in the part you quoted, I specifically stated that I was aware that paying for labor was not an issue, because it is entirely donated. Do I have to make it stand out like a billboard?

I did note that FIRST is a reflection of the real world. As such, real world analysis and skills are a highly useful tool. However, they should not be the sole driver. If FIRST was real world, you and your team would be paying for that infinite supply of scraps, and for the labor. You probably wouldn't be functional financially, or would be having a lot of trouble paying bills. Do the analysis sometime, using the FIRST BOM accounting methods for the materials, and minimum wage for the labor costs. I think you'll be surprised at just how much your robot is actually worth. (I conservatively estimate any given Kitbot+stuff box-on-wheels built by a small team at about $9K. The bigger the team and the fancier the robot, the more it'd be worth. I'm not factoring in any awards earned by said robot, either.)

But when you get to the real world, you will have to worry about that. Engineering is a balance of money, time, and quality. (In some fields, use weight in that balance as well.) You can pick one, or maybe two, but the third will suffer. Reduce the cost, the time might go up or down, but the quality is almost certain to go down. Reduce the time, the cost goes up (for expediting) and the quality goes down. Increase the quality, one or both of time and money goes up. This is a real-world exercise, and this is engineering--learn to balance them now so that you can deal with the balancing later.

[sanddrag]

Eric, I call it the triangle of manufacturing.

You can have it good, fast, or cheap. Typically one of the three, if you're lucky two of the three, but never all three.

And the way I describe this idea I may have just made up, but I feel I heard it somewhere before.

[Tristan Lall]

Originally Posted by EricH

We just had this discussion in one of my senior-level college classes today actually. Apparently, back in the 80s, at least one company designed by fabrication--they focused entirely on ease of manufacture. This cost them a lot of business because once built, the product could not be serviced easily. Fabrication drove their design process, and this damaged the company. They have since refocused to designing with all the other factors (and believe me, there are a LOT of other factors!) in mind, and are producing better products.

What's the company? That sounds like an interesting story (to an engineer, at least).

I think it's fair to say that the relative value of manufacturability vs. serviceability depends greatly on the application, the industry and the capabilities of the enterprise.

[Peck]

One problem I have seen is that some teams have a really nice bot but don't even know how to make a simple rod on a lathe (manual or cnc). They ship out even on stuff that your grandma (assuming your grandma isn't skilled in fab) could make in a few minutes.

If you (the team) want to ship it all out, fine. Just make sure your students (as a collective) don't look like idiots when a rod snaps at competition and someone brings you metal stock.

[EricH]

Originally Posted by Tristan Lall

What's the company? That sounds like an interesting story (to an engineer, at least).

I'll PM you the company. They've since changed to a design process that isn't driven by ease of fabrication.

It's one that builds a lot of heavy equipment, particularly for farming and construction.

[MrForbes]

Originally Posted by Brandon_L

Its cheaper for us. We don't have the money to send parts out to be made.

I think you might have some false impressions about how other teams work. I only have first hand knowledge of how two other teams work, both had some neat CAD (student) designed, out-of-house fabricated parts. The rookie team 4183 which my brother works with had most of their robot structural and shooter parts made by a sponsor. The parts were free, there was no monetary cost to the team. But the cost in time was kind of heavy, first they needed to learn how to use CAD and then they had to figure out how to design the robot, and the parts. By the time all the parts were made, build season was almost over, so they had to work really hard to get the robot built and working, and there were some design issues that required them to make some "hack" changes to the robot at their first regional to get the balls to move through it.

The other team I know about is 842, they had some (student) CAD designed transmission plates made by a sponsor, again no monetary cost to the team.

Instead of worrying about the cost of having parts made to your design, you might want to spend some time finding out if there are any companies in your area that can do this type of work, tell them what your robot team is all about, and ask if they'd help you next season.

[aldowyn]

3507 is only a second year team, but this is how we've been doing it.

We haven't even been using CADs, since no one knows how yet (working on fixing that next year), but we DO use drawings (obviously). Most of the time this ends up being just a rough draft, but we do make sure to conceptualize EXACTLY how each piece goes where, how it's supported, etc. etc. We do most of the fabrication ourselves. I believe this year the only thing we had machined was our shooter, which was rolled somewhere else - it had to be exactly right. (It was. I honestly believe ours was the most precise shooter at OKC, if we could have known how hard to shoot it. We had a 91% accuracy in the qualification rounds)

In any case... just designing? Where's the fun in that?

[FenixPheonix]

Team 751 does all our own manufacturing for one simple reason: we don't have the resources.

Let me explain. Our team has, in terms of manufacturing, a manual mill, lathe, drill press, four hand drills, a drop band saw, a vertical bad saw, a sander, a grinder, and a TIG welder that we got half-way through this season. We have a small team of fifteen students, and a highly devoted core of less than eight. As such, we have to balance both man-hours and machine hours. For every day-long mill job we set up, we sacrifice another, and lose design time for some area of the robot. And that is precisely why we, the students on the team, choose not to outsource our work. We've found that the lesson in practicality and trade-offs is more valuable than the functionality that we would gain from outsourcing. If a person on our team has a great idea that would need three days of manufacturing time, we all sit down, and run cost-benefits analysis on it. Is one day of mill-time and a student day worth half a pound of cheese-holing? We run into this one a lot. Most of the time, we say it isn't. But before SVR, we realized that, for one part, it was, because that half pound let us change out our corner wheels from Plaction for Performance, letting us solve the wheel breakage issues from Sacramento (cracked one in half and dented three to the point of being unusable crossing the bump.) So we spent the time, and ended up below the weight limit.

And it's precisely because of our in-house work that we CAD everything. We design all our parts, do time estimates, and schedule our machine time, going for maximum efficiency. We've got folders of part layouts that students just churn out, organized by sub-system and machine utilization. Our drivetrain was fabricated (not assembled, Andy-Mark shipping held us back) in three days, and everything else took us about nine. Everything was CAD'd, everything was prototyped. When you know how to run design well, you can streamline manufacturing. CAD is vital for optimizing machine schedules and ensuring that we can make everything, and producing accurate drawings our designs (very few of which are made by the person who designs them. That's usually for the few that we decide need the machining time the most for optimization, as we want the person who understands the design intent and manufacturing process, having designed both.)

We're looking into expanding our manufacturing abilities with another manual mill, and possibly a CNC. We are also hoping to get access to a local waterjet (at the TechShop. We've also done a bit of laser cuttig there.) This will eliminate some of the urgency of the scheduling (though I have no doubt that our designs will become more complicated to saturate our machines again.) However, we will continue to design and build all of our own parts, as we find it promotes a culture of responsibility and efficiency.

My point here is: we build in-house to teach our members about time and resource management, at the cost of some machine functionality. Is it worth it? Depends on what you want to get from it.

[Anupam Goli]

Originally Posted by FenixPheonix

Team 751 does all our own manufacturing for one simple reason: we don't have the resources.

Let me explain. Our team has, in terms of manufacturing, a manual mill, lathe, drill press, four hand drills, a drop band saw, a vertical bad saw, a sander, a grinder, and a TIG welder that we got half-way through this season. We have a small team of fifteen students, and a highly devoted core of less than eight. As such, we have to balance both man-hours and machine hours. For every day-long mill job we set up, we sacrifice another, and lose design time for some area of the robot. And that is precisely why we, the students on the team, choose not to outsource our work. We've found that the lesson in practicality and trade-offs is more valuable than the functionality that we would gain from outsourcing. If a person on our team has a great idea that would need three days of manufacturing time, we all sit down, and run cost-benefits analysis on it. Is one day of mill-time and a student day worth half a pound of cheese-holing? We run into this one a lot. Most of the time, we say it isn't. But before SVR, we realized that, for one part, it was, because that half pound let us change out our corner wheels from Plaction for Performance, letting us solve the wheel breakage issues from Sacramento (cracked one in half and dented three to the point of being unusable crossing the bump.) So we spent the time, and ended up below the weight limit.

And it's precisely because of our in-house work that we CAD everything. We design all our parts, do time estimates, and schedule our machine time, going for maximum efficiency. We've got folders of part layouts that students just churn out, organized by sub-system and machine utilization. Our drivetrain was fabricated (not assembled, Andy-Mark shipping held us back) in three days, and everything else took us about nine. Everything was CAD'd, everything was prototyped. When you know how to run design well, you can streamline manufacturing. CAD is vital for optimizing machine schedules and ensuring that we can make everything, and producing accurate drawings our designs (very few of which are made by the person who designs them. That's usually for the few that we decide need the machining time the most for optimization, as we want the person who understands the design intent and manufacturing process, having designed both.)

We're looking into expanding our manufacturing abilities with another manual mill, and possibly a CNC. We are also hoping to get access to a local waterjet (at the TechShop. We've also done a bit of laser cuttig there.) This will eliminate some of the urgency of the scheduling (though I have no doubt that our designs will become more complicated to saturate our machines again.) However, we will continue to design and build all of our own parts, as we find it promotes a culture of responsibility and efficiency.

My point here is: we build in-house to teach our members about time and resource management, at the cost of some machine functionality. Is it worth it? Depends on what you want to get from it.

Right. I'm not against in house fabrication at all, I think it's great to know the process and be able to make good time estimates. If people can do it properly, let them. I love being able to use our team's CNC mill, and being able to mill manually is very useful. As a freshman and Sophomore, I loved being able to make parts and place them on the robot, as I got proud that my own part that I created was being placed on the final machine.

The point of this thread was to point out that designing is needed before fabrication, and everyone here seems to agree that detailed drawings or CADs are the way to go before any fabrication. My team likes to jump into fabrication, so a lot of times we end up finding solutions while we are creating parts. It's terrible. Parts cut too short are somehow placed somewhere else, parts too long are cut, and then found out they are not needed, and the result is an entire robot that is one swift kick away from being destroyed. In 2010, we used a lot of 80/20, but we didn't do CADs or detailed drawings. By the end of CMP, which was our second event that year, our Frame was offset by more than 30 degress because of the irregular placement of parts and the use of subpar parts because of a lack of design.