I spent some time recently at the Fab11 conference, a gathering of the people behind the Fab Labs that are springing up all over the world, where entrepreneurs, hackers and the curious can learn about making things. So, it was no surprise that this was a great place to pick up some tips on designing, building and hacking things. Here are a few of the lessons I picked up at this fascinating gathering of the fabbers.
If you can make something in an hour, you’ll make it better in a day
said [Joris Van Tubergen]. He knows something about making unusual things because he 3D printed a full-sized Elephant. To do this, he worked out how to hack the Ultimaker 2 3D printer to print to an unlimited Z height by flipping the printer upside down and moving the Z motor to lift the printer rather than the print head. With a few tweaks to the software, he could then print full-height elephant slices to speed up the process. He is absolutely right: while it is tempting to endlessly fiddle with a concept on paper, you learn more by building a prototype, even if it doesn’t work.
In the session on making robots, [Marc Raibert], founder of robot company Boston Dynamics described his design philosophy quite simply: “Build it, break it, fix it. We go round that loop ten times a day in the early days of a robot. We build our whole team philosophy around the idea that it is not going to be right at first. Or even if it is right, it is going to be more right if we take it out there in the world and beat it up a little bit”. He encourages his engineers to build prototypes, push them until they fail, then figure out how to fix the bit that failed. Then rinse and repeat. That’s a great approach, especially for big problems such as the Big Dog and Atlas robots his company is building.
Sometimes Good Enough is Good Enough
Endlessly reworking designs is a rabbit hole that many hackers fall into, though. It is way too easy to get caught in this cycle of endlessly tweaking rather than stepping back and deciding that something is good enough, and getting on with other parts of a project. [Joris] spoke about this when discussing his printer hack. As part of his hack to make the Ultimaker produce unlimited Z height prints, he had to work out how to get the printer to reverse the axis of the print and not worry about the larger Z heights. Rather than hack the firmware of the Ultimaker, he made the changes in Cura and manually tweaked the G-code of his test models. That manual (and rather inelegant) approach meant he could focus on the main problem of making sure that his concept would work as he hoped it would, then he could go back and work on the technical details afterwards. And that is exactly what he did: after he decided to do a Kickstarter to offer the hack to users, he created a custom firmware image for the Ultimaker 2 that solved the problem. But if he had done that first, the entire project could have floundered.
Simplicity is Key
In the making robots session, [Mick Mount] of Kiva Systems discussed how his company builds robots that shift shelves in the warehouses of companies like Amazon to bring products to the people who pack them. Although you might think this was a great excuse to build an all-singing, all dancing, hi-tech robot, he keeps his minions simple, building them using car batteries and off the shelf actuators rather than cutting-edge technology. He described his robots as being like Lego Mindstorms compared to what other robotic researchers were demonstrating.
We make a very simple robot…we try to keep it cheap so we can do thousands of them in a warehouse
he told the conference. But this simple robot was enough to convince the notoriously thrifty online retailer Amazon to buy his company for $775 million in 2012, so there is obviously something to his idea. He makes a great point: when you are building something to fulfill a specific task, the simplest and most direct solution is usually the best. We all want to use the latest and greatest technology to build something, but it is usually more useful to focus on the task at hand rather than the technology that might get you there.
Hacking is Fun, But Educating Other People is More Fun
One of the aims of the Fablab movement is to create spaces that can educate people in digital fabrication by showing them how easy it is. This is a laudable aim that we, as hackers, should remember. While creating interesting hacks is fun, it is also important to teach others, to show them that they can build stuff themselves. The fundamental aim of the Fablab movement is to create spaces where people can build their own products, either to sell or just for fun. But whatever the motivation, you have more to gain by teaching others than by keeping knowledge secret.