Hey yall! I’m currently the leading CAD advisor in my team. I have been working under the CAD dept for a while now. However, it is evident that our team has some struggles with organization and collaboration between different departments in our team. The conflict that I am most worried about is about is the conflict between how CAD and mech cooperate. How are team’s departments organized? How do the different departments, especially between CAD and mechanical, collaborate and work with each other on your team?
Our design & mechanical teams work together for design concepts & prototypes early in the build season. So in that regard, we don’t really recognize a difference between design and build teams during the first week or two of the season.
Once the high-level design requirements are figured out we run it like a semi-chaotic manufacturing firm. Our design kids CAD their parts and assemblies, create 2D drawings, and send it to the shop. Usually, they forget a dimension or something and our mechanical kids just go to the designer and ask them to fix the drawing.
It works OK for us; we just don’t have great organization around keeping track of what parts are ready for manufacture vs. what is actually in the shop once we start cranking out designs. At times that can lead to frustration between the groups.
Last season we started doing design reviews every few weeks (we probably had 3 or 4 deep dives all season) and we made the kids present us their design in front of the mentors and entire team. During those reviews we went into deep detail so by the time the drawings hit the shop the entire team had a good idea about what we were trying to build. That’s definitely something to consider doing.
On our team, CAD and design are basically the same group of people all working together. We prototype, concept, and then work together to CAD the robot. Although only a few people actually do most of the CAD work, it helps having everyone on the same page.
In terms of CAD and build (fabrication team), we make drawings of all the parts that need to be made. We ensure all the required specs are there for a manufacturer to make a part, and they handle the rest. It’s really quick and easy interaction.
Could you please elaborate on some of your problems? That might help people propose some more specific solutions.
We have separate CAD and Mechanical subteams, with some CAD members helping with part fabrication. I didn’t notice major conflicts between the two groups, but there were some small issues. A few times, Mech members ran out of parts to make and were waiting on CAD to pump out more drawings. There were also some drawings that didn’t quite line up or weren’t labeled correctly. Our part naming conventions also didn’t always make sense, and some of the assembly drawings got confusing. And, of course, there were some design arguments.
I think we lessened some of those issues last year by pushing toward reduced mechanical complexity, more COTS parts, a lower overall part count, and fewer different types of materials. Building up a better COTS part CAD library this fall and getting a better naming convention could help even more. Being aware of and using more COTS solutions will make life even easier for the CAD team, which might help them work faster and avoid some of the backups we’ve had in the past. Prototyping ideas really reduced the number of design arguments, because people could back their points up with evidence. GrabCAD has also been pretty helpful for organization and version control. Slack has been really helpful for communication about prototypes, ideas, designs, and updates.
Organizational structure and collaboration will depend on your approach to the robot design process. We are changing our approach a bit this season, but the key is to have your approach and structure communicated in advance. We will practice robot design and construction as a team this fall to make sure we are prepared for 2018.
Here are a few keys I suggest.
- Design / Build Leadership: You have listed CAD and Mechanical teams as departments. We do the same, but the distinction is each group is doing tasks within one design goal. We have one person that coordinates all mechanical activities and makes sure each group has the resources needed to make progress.
- Communication: Have a communication plan for coordinating the sub teams and reporting on progress and issues. We usually spend a few minutes at the end of each meeting for these reports and have a longer discussion at the beginning of Saturday meetings.
- Decision making: FRC build season is fast. Defining your decision making process is important. Getting input from the team is fine, but I recommend limiting decisions to no more than a few people - maybe your sub team leaders.
On our team, every step of the process is all different people, with usually 3-5 per sub-team.
Prototyping->CAD->CAM->Manufacturing. Manufacturing is branched into Mill, Lathe, Plasma, Router, and Welding. Everyone has their specific job and knows it well, and constant communication between sub-teams is key.
I would like to revive this topic as we prepare for the build season. Specifically, I have a few questions for any teams that keep their CAD and Mechanical teams separate:
How do you manage those first few days of build season? Who is pumping out designs? Who does prototyping?
How do you manage disagreements in design? Who breaks the deadlock? For example, if Mechanical wants to make a new design, but CAD doesn’t want it to change the design they already produced and works. Who has final say over the final robot design?
Who learns how to use the machines, especially the CNC’s or other automated machines that require CAM software to run?
In response to your questions, LordPeaches
We have around 10 or so mechanical students and only 2 CAD students at the moment. Some of the mechanical students know enough CAD to design a prototype mechanism. In the past, we’ve divided up into groups with 1 CAD student or mechanical student who knows CAD per group, often leading that prototyping group.
The main break in a deadlock where CAD has one design that works but Mechanical wants to change is either the electrical team or programming team, depending on which design works better for them to complete our strategy.
CAD students are highly encouraged to learn CAM, but currently most of the CAM is done by either the mechanical mentor who has a CNC Mill in his garage or our mechanical lead student, who also knows CAD.
As to how our CAD team interacts with other subteams in general, that's under here.
The CAD subteam on our team is currently made up of two people. As mentioned, some of the mechanical people on the team also know CAD, and they do some, but most of their time is spent fabricating parts. When it comes to a difference of opinions between the CAD team and other subteams, it usually resolves in favor of which opinion is more simple (KISS).
For the first few days - the whole team. Our early build season meetings focus on reading/understanding the rules, setting game strategy, and early brainstorming/prototyping. Every group on the team participates in this process.
For the continued prototyping past the first few meetings, the mechanical team is iterating and refining prototypes while the design team starts locking down the drive base and key dimensions for potential manipulators. The products and key dimensions of prototypes are fed to the design team as we refine designs and down select.
We strive for consensus on most major strategy and design decisions, and solicit input from all technical officers at the very least. Additionally, we use a weighted decision matrix to help inform our decisions. While we’ve certainly had cases of individual officers who have strong opinions contrary to decisions we’ve reached, the bulk of the officers will typically manage to get most of the team onboard with a given design. Ultimately the design subteam has the authority to execute the design in CAD and mechanical has the authority to fabricate and assemble the design in reality, so in practicality they often get a “final say” when it comes down to the real-world design conflicts that can occur as you attempt to realize your conceptual design.
The design group is responsible for learning the CAM software and interfacing with the control software on the router. Mechanical students assist with setting up the machine.
I’m a big advocate for combining your CAD and mechanical subteams into one. Over the past two years I’ve found that the distinction lends itself to a lot of dysfunction, and this year we’re merging them. Instead of dividing into “this team does CAD”, “this team assembles everything”, and trying to figure out which of them should be building prototypes, divide the students by mechanism instead. A couple kids responsible for the drive train, a couple kids responsible for the elevator, a couple kids responsible for the intake, a couple kids responsible for the climber, etc, with one student captain responsible for coordinating those groups and keeping an eye on final integration/packaging. Among each of those groups, you’ll need at least one student who’s comfortable with CAD and one who’s comfortable using hand/power tools.
In my experience, when CAD and mechanical are separate, the 3-4 kids in CAD designed the whole robot, which is a huge amount of work, pressure, and exhaustion, and the other 20 kids on the team felt like they were just carrying out the CAD team’s orders and weren’t part of any meaningful designing or decision-making. Plus when it comes time to actually assemble something, the CAD team are the ones who know what it is and how it works, and 7 times out of 10 it gets put together wrong if CAD students don’t supervise. And the CAD team had no interest in (or energy left for) improving and iterating - once the CAD was done, they considered their job complete. There was also an undercurrent of resentment that ran both ways - mechanical students felt like they weren’t smart enough to do CAD, and the CAD team felt like they were always blamed for anything that went wrong with the robot.
Whereas on my previous team, where there was just one Mechanical team broken into subsystem groups, we’d have 10-15 students meaningfully involved in the design, and kids felt ownership over their subsystems and excited about continuing to improve them after they were built. Some students did more CAD than others, and some didn’t do any CAD, but they were all invested in building a great robot and moved naturally through the stages of brainstorming to prototyping to CADing to machining to assembly to testing and improvement.
Your CAD team should also include some members of the electrical/pneumatic team. At the very least, they should have input (maybe veto power) into the CAD and get to review the design frequently and regularly. Otherwise, you can easily end up with a design that is difficult to manufacture and/or service i.e. your control system buried under major mechanisms and the battery in a location that makes it a PITA to swap.