Our team this year had a horrible process of designing parts.
it was a think of it build it and put it on type thing and we were never organized.
so next year I will most likely be a team leader and I was wondering how most teams go through designing parts.
right now we thought of something like this:
-Rough Design
-Discuss with others
-CAD Design
-Prototype in VEX
-Discuss with others
-Make changes to proto/CAD
-Create
we also had our whole team swarming over our 'bot all season and could barely get in to program or check if somethings would fit. my idea for a fix of that was take a piece of ply wood, lay it over the robot and cut out holes where ever there was a place that had nothing in it. so we could just cut more of these then give them to people who wanted to design a part.
any suggestions to my ideas?
and what does your team do?
Our design process this year was kind of simple (thank god) and pretty quick. After we figured out the basic chassis design, everything sort of developed from there from āwhat would be nifty to have but not too complicatedā thoughts. We ended up with milled motor mounts that could slide for chain tension, an evenly-balanced chassis with tons of torque, and an extremely robust design.
EDIT:
Next year, however, I know we want to Prototype with a VEX robot. I donāt know if we will, though.
we started of the build season with a pugh analysis to decide what our robot needed to do.
then we prototyped different designs that fit what we needed
We then made the parts and put the robot together.
AFter the build season, I had the chance to draw everything up in Inventor
hopefully, next year, we will draw parts first in Inventor and then build them
This season, we probably did things the āwrongā way. Vague concept of scoring on all three levels, followed by some measurements taken from the field drawings (and its accompanying trigonometry) yielded a PVC arm capable of reaching the top spider. Not nearly enough attention was paid to prototyping the gripper, which is something I hope we fix for the off-season.
The finer details we hammered out more or less on the fly, such as tensioning the chain on our shoulder while providing enough chain wrap. (For that, we ended up using a short piece of UHMW channel weād bought off McMaster-Carr and running it another way.)
Only the drive system was done up in CAD, although things the model left out caused a couple of issues (mainly dealing with chain running through where Iād put mounting plates).
Next season, however, Iām hoping we can improve our Inventor work to allow us to prototype more on our limited budget. (Electrons, as far as the team budget is concerned, are free. ;))
Weāre in the same boat. I really hope that next year we will design in Inventor first to get out all of the kinks and then make all of the parts needed. That is after all, one of the things Inventor was designed to do.
179 typically meets the day of unveiling to build a mini field and go through strategy and game play. We then determine requirements of game play (things that allow flexibility in strategy). There is then a few days off for people to work on designs and concepts. We rarely prototype. Schematic like designs are laid out in CAD to determine kinematics relationships between bot, playing field and game pieces. These sketches take on life as you work in the details. The devil is always in the details. Lots of hand sketches are made of conceptual modules. Each component is considered modular for revisions, replacements and ease of installation. Electronics are always separated from the chassis to allow tandem builds and not cause succession delays. We look for as many off the shelf components as possible without sacrificing worthy customization. The full robot is modeled in AutoCAD with every component considered. Then holes are drawn into every piece of surface area not being used for structural support or mounting. Materials are ordered and the chips start flying.
This year we started with the traditional arm bot design with a claw like many teams had. However we performed physical tests with our rack attempting to score on it with an arm while being defended and found it was near impossible. So we decided then that a ramp was more important and that scoring on any 1 level was all that was necessary. So about a week into the design period we started over. But we still finished the bot with 2 weeks before shipping. Each year we start with a fresh piece of paper, no 2 bots are the same.
Dan that was a great break down obviously something Swapthing does works they have the trophy case to prove it. 1251 usually starts out after watching the kickoff video by breaking up in to groups with various mentors and comming up with what each group wants the robot to do and some come with a rough sketch. After this there are many brain storming sessions on various designs that come up and finally a compromise is reached. Then with the final rough design we get together with our main mechanical students and mentors and start to CAD out every aspect we can. Next, is usually the prototyping stage and then we usually rebuild it strong enough and light enough to where we want it. Finally, everything is finailzed about the design and production is started. The electrical team works seperate for a bit on the board which is usually a bolt on piece and it all comes together once the frame is in the building. Thats basically a rough overview of what goes on. Usually we build two robots one being a prototype/practice bot the other being a final product in general we spend about 2 weeks cading and protoyping.
Definitely have your strategy laid out first, we normally have a brainstorming session where we talk strategy onlyā¦ no details at allā¦ then after we decide, āwe want to do this, this and thisā, we have our design team come up with the best ways to do this, while other members prototype these designs using scrap parts and last years robot. Although i must say we got very sentimental with Gloria since she was a World Champ and didnāt use her for prototyping until a lot later than most robots get scrapped. Very sad but Iām sure she would have agreed to it if the mentors would ever let us install that artificial intelligence and speech chip. We use SolidWorks for our design because we do all our designing at FANUC and thats what they have (Iām sure Paul has some arguments for it other than that).
Cyber Blue starts our design by choosing how we want to play the game (tube scoring this year). This can take anywhere from a few days to a couple weeks, depending on the game.
After we decide on a strategy, we start prototyping different types of gripper designs and drives. We often look to past years and some concepts explored by other teams when designing arm and manipulator parts. This year, we designed a chassis during the off season, and it worked well. We modified the design for Rack ānā Roll.
We CAD several basic overall designs, then we pick the one that suits our purposes best. We modify the parts in this design in Inventor until they are ready to be manufactured. Next, we look back to our prototypes to see which manipulator works best for the game. We CAD the manipulator and add it to the final robot design.
Once we are pretty sure that a part is ready to be made, we send it to the manufacturing lab downstairs. If the part requires fine measurements, we use a CNC mill.
Finally, we assemble the parts into their various subsystems. If a system has a problem that we did not see in the design, we modify the CAD file and remake the necessary parts. This continues until the robot is completely built and ready to be shipped.
We always meet with another team after kick-off to get out-of-the box ideas. It helps to bring in totally new people and concepts into the discussion.
-Pick a way to score.
In my opinion the best and safest way to design a robot is to pick one strategy and stick too it. Although there are plenty of teams that āDo Everythingā there are far more teams that fail to āDo Everything.ā Decide what you think will be the best way to score and design your robot to do it.
-Prototypeā¦ A lotā¦
Professional engineers have a great deal of experience and often donāt need to do a lot of physical testing. They can see problems on-paper before they get built.
Most students lack this experienceā¦ On team 100 we brainstorm what we feel will be the five best designs and try them out. Doing this takes a lot of time, but it guarantees that the final design will work. Our prototyping is very in-depth, we built an entire fully functioning prototype of the Tube-a-saurus before it got built.
Prototype is on the left
-Perfect the design in CAD
Take what you learned from the prototype and put it into the CAD. We went through several CAD revisions before we perfected our design.
After you have done all of this, you will feel very confident in where you are going. It will just be a matter of building.
Together in discussion, students decide what kind of bot we want(usually itās the veteran seniors who pick because by the time itās senior year, all we do is talk on IM, text messages, and e-mail and the next day we report back to our coach and say exactly what we want.).
Eventually the group falls to 4 or 5 people(usually senior veterans) on what the actual bot will look like and discuss why. - We do not start anything on the robot until the students come up with an all student designed robot. One of our engineer mentors came up with something for us but we didnāt like it because the robot wouldnāt really be ours.
No autoCAD(yeah i know) of entire bot, only time we use AutoCAD is to make protoypes with our laser. Nor do we go to a Machine Shop, completely backyard mechanics-measured and fabricated by the hands of the students.
whatever we design, we make sure we K.I.S.S. it(Itās āKeep it simple stupidā, but we say āKeep It Simple Smartā.-Judges liked it a lot at Colorado Regional).
After a failure the design process for the 06 year, I decided it was time to change the way we did it.
Originally, we simulated the game, then broke off into groups and designed the manipulator (in this case, a shooter). After eliminating most of the designs, we voted on the final one. However, no one followed through with that design and something completely different was created.
This is how things went for this year.
Read over game rules
Break up into groups and determine overall functions of the robot for that group (ie. score on bottom and middle, lift both robots 12 inches). We then simulated the game with these 6 teams.
After going over the simulations, the team votes on which function we want the robot to accomplish (we donāt even look at the possibility of doing it all, anymore)
Monday, we split up into groups again and physically design how the robot completes the tasks we voted on previously. (ie. arm or elevator, wide or long base orientation)
A small group, determined before Kickoff, that includes the team captains and all of the sub-team leads, plus a few students who put in an application to be on this group meets in one of the rooms with any mentor that wishes to attend and we discuss pros and cons of different things (ie mecanum wheels vs omni wheels, arm mounted on front of the robot or back of the robot, shoulder joint or no shoulder joint)
The team got thinned out to maybe 4 or 5 students. Two kids go to town on Pro-E after this smaller group goes into more detailed design. This design is presented to professional BAE Systems Engineers who come for this Preliminary Design Review. Action items are given to design members to figure out certain parts of the robot (ie torque required for the arm, what type of motors, chain vs belts). After all of the little details have been worked out and after everything has been put into Pro-E, we hold a Final Design review (CAD Software)
The gripper was prototyped just to see what the best way of gripping the tube would be.
Then, drawings were printed for all of the parts 3 times. One for our design notebook, one for the machine notebook that goes to the machine shop, and one for the inspection notebook. All parts fabricated go through inspection. The inspector makes sure the measurements are correct and weighs the part. The weight of the part is then added to the spreadsheet. We werenāt too far off with our weight this year, which is good.
We wanted to get rid of the huge design team because arguements come up and a lot of time is wasted. Also, you get people who donāt really want to be there (ie. the entire animation team)
Thatās it. If you have any questions, PM me. I actually started a thread around this time last year to learn how other people design their robots.
Similar to what other people have posted, this is how we design:
Immediately after the kickoff, we print off 6 or more rule books
We meet immediately after kickoff to discuss the game, the possibilities.
Our first 2 official meetings are, āDecide how you want to play the gameā. Absolutely no designing is allowed, just strategy. We break off into small groups, given the objective āPick one thing that you think we can do better than anyone elseā The team then agrees on what strategy weād like to have.
Our next two weeks are designing to implement the strategy. Parts are prototyped, ideas are tossed about. A base drive train is made, and given to programmers. During this time, our only concern is implementing the strategy.
Building a robot for a specific strategy has worked well for us, and has led to many engineering awards over the past years.
well, our team at kickoff has our own meeting, that spans over 3 days, with each day doing something else:
Day One - students, and mentors, are dissuaded to think about robots, we think about what we want to do and what we want to accomplish, and what will win the game (this year we decided to not have ramps since we assumed that there would be ramps aplenty to choose from, and I think we werent far from the truth). One key thing that people tend to overlook is logistics (time spent moving across field, under defense, etc). Always try to factor that in.
Day Two - Rough design of what the robot can do to complete our objective, minor sketches. When dealing here, remember the wonderful acronym KISS (Keep It Simple Stupid).
Day Three - Specific design features of what the robot can be built to do, detailed sketches (Dont get overzealous, there is a construction deadlineā¦).
Then, for the remaining build season, we try to construct a complete replica of our robot on Pro-E CAD, which helps when making parts cause it gives you a sense of assembly. We waterjet and CNC most parts, with the talented machinists finishing off the work in their own way. This year, since our 2k5 and 2k4 robot had a similar design, we could use those robots as a real-life test to our design, since most of us know what looks great on paper is sometimes misleading.
We use a very simple process:
At kickoff, hold open discussion with any other teams we happen to be with.
Read rules.
Play game with humans as robots. Done multiple times, focusing on different sections/strategies. Day 1-end of Week 1
Come together and decide on a general strategy. Week 1-2.
Develop requirements to fit said strategy. Some are from previous years, some are game-specific. Week 1-2.
Start figuring out designs that can play the game the way we want. Week 1-2.
Decide on a rough design. Sometime between Weeks 1 and 2.
Detail the design in Inventor, starting with the drivebase and working our way up and out. Usually around Weeks 2-3, depending on a number of factors.
Build. Any time after the detail work is done. May prototype before detail is done.
Sometimes, parts of the robot are detail designed after the base is built. Also, we call in outsiders occaisionally to help us see if thereās something weāve overlooked.
