There’s a time in the development of many makers when they switch from making things purely for themselves to making things for other people to make. The point at which you start giving your designs away and other people start making them is when things get really interesting. The transition from lone maker to sharing maker can be difficult, as it underlines to how easy or difficult your design is to build. When you’ve designed something yourself you make certain allowances, but it’s a different story when someone else is building your design and they find part of the build difficult, or worse stupid.
For the past few years I’ve been making robots for other people to build. First with Mirobot and now at Mime Industries with the MeArm. We’ve recently launched a new product (shameless plug alert: check out the Kickstarter) which we’ve put a lot of effort into making easy to build. There’s also an additional factor in this case, which is that we also want children to be able to build it. Making something that can be built by children is particularly challenging because you can’t rely on the base set of skills that most adults have developed. So here are a few things we’ve learned and techniques we’ve used in the MeArm Pi:
1. Screws are difficult
I know, screws are one of those basic things that you, as a maker, take for granted as being simple to use and easy to understand. But you’ve probably never watched a class of 30 kids trying to screw together the design you made (and failing miserably).
I was amazed by how hard kids found it to use a screwdriver and don’t underestimate how long it can take to screw in a single screw. In the latest design we’ve been able to reduce the number of screws down from 42 to 14, and take the build time down from a few hours to 30 minutes, by using a few tricks:
Elastic bands are one of the simpler ways of holding an assembly together. They provide a nice even tension and if you design your kit right a single elastic band can replace an awful lot of screws. They generally create a surprisingly robust construction because the tension is applied in a number of dimensions. For the base of the MeArm Pi we’ve used some bands to hold the assembly together around the Raspberry Pi, replacing what would probably have been at least four screws. They also serve a dual purpose as non-slip feet.
I’m not sure if this is the right name for this, but I haven’t seen it in too many places. It’s a useful trick that I started using in the v2 Mirobot. If you have two pieces that you want to fix to each other you can hook one end in, rotate it into position and then secure the other end. By holding in the other end with the next piece you can build up the construction in such a way that each piece you add holds in place the previous piece. For the grip assembly in the MeArm Pi we’ve moved from a screw together assembly which used nine screws to this new latch-style assembly which only uses a single screw, that locks them all in place. It’s tricky to explain, so the best way to see this is in this video showing the assembly of the part. Look particularly at how the orange sides hook on and hold everything in place.
Combining the latch with elastic bands
We used to use a couple of screws on each of the side servos to hold them in place but by making clips that hold them in place and then elastic bands to hold the clips we’re down to zero.
2. Finding the right pieces is difficult
When you have a lot of parts of a robot in front of you and you need a specific part they all start to look very similar. There are a few tricks to reducing the likelihood of build errors by the way these pieces are designed.
Similar should be the same
If you have two very similar pieces in the design, see if you can make them identical. That way you reduce the number of chances for the maker to pick the wrong one.
Different should be very different
The flip side of this is that if they can’t be the same, then you should try to make them as different as possible. This way it’s more obvious when looking for a piece which is the right one.
Symmetry is good
This is particularly relevant when laser cutting acrylic as the pieces will often be easy to put on in one of two rotations (or four really). So if you can make a piece symmetrical then the maker won’t be able to put it on the wrong way round because both ways are fine. By reducing the number of possibilities you reduce the overall complexity of the build.
Asymmetry is also good
This is the same idea as the previous one; if you can’t make a piece perfectly symmetrical then you should try to make it so that it can only be attached the right way round. It should be really obvious if you try to put it on the wrong way round that it’s on wrong.
An example of these ideas in action is the base of the MeArm Pi:
- The central piece has different size tabs so it can only slot over the servo one way
- The two white end pieces are identical and can also only go on one way round
- The two orange side pieces are different shapes but could be put on flipped by mistake. To avoid this the slots are sized so that they will only go on one way round and on one side only
By making sure that pieces can only go on one way round it means that when (and it is when, not if) someone tries to put it on the wrong way, one of the tabs won’t fit in the hole. At that point the natural thing to do is just try flipping it.
By using these ideas, you should end up with a kit that is not only simpler for children to build, but also makes lives much easier for any adults involved too.