Following on from some thoughts of yesterday, what would it look like to have a “platform” that would make it easier for people to use their computer and their 3D printer together to create prototypes for new kinds of useful devices?
A microcomputer (like the one in even an inexpensive cell phone of today) is very good at computing, but it’s not very good at making things happen in the physical world. For that you need components like motors and solenoids.
But there’s a problem. While a modern microcomputer contains many millions of electronic switches, the cost of motors and solenoids is not going down. So any mechanism that needs to use lots of motors and solenoids for physical actuation is going to hit a serious wall of cost.
I’m thinking that we will need to create a general prototyping architecture which will allow lots of different parts to be activated using only a small number of motors and solenoids, by using the same motors/solenoids to actuate different things. For example, imagine an animatronic prototyping board (APB) that looks like a phonograph turntable, carrying a solenoid which travels around in a circle. As the moving solenoid comes into contact with different parts, it pushes each part up or down (all under control of a microcomputer).
Since our future home 3D printer will be able to print any already-assembled arrangement of gears and mechanical relays, the right 3D printed mechanism should be translate the simple up/down movements made by this APB into pretty much any robotic action we need.
This is also the layered way that computer software is built — a general purpose compiler at the lower layers, and then more special-purpose application software built on top of that.
I’m not sure that an animatronic prototyping board will really look like a turntable carrying a solenoid. But I’m pretty confident, just as we now take general purpose microcomputers for granted, that future made-and-designed-at-home robotic devices will be built using some kind of APB.
Things like toys, old VCRs and inkjet printers have some amazing mechanisms inside. They use a single motor, coupled with lots of gears, cams and levers to get different things happening at different times. I’m sure the calculus is just as you describe:
$(motors + controllers) >> $(plastic parts)
I’d love to see one of these mechanism designers interviewed. I’ll bet it’s a bit of a black art.
Yes, good point. The question I ask in my post is whether there is a difference between (1) making a single fixed device configuration that will be mass-produced in the millions (ie: the toys of today), and (2) makig a platform for trying out a new electro-mechanical device idea, where you want each prototype to be economical in cost even when you’re only making a few.
I think the difference may have a lot to do with the metal parts, which are much cheaper than motors, but also are not easily produced on a 3D printer. The architecture I described in my post localizes all of the metal parts (as opposed to plastic parts) into one universal prototyping platform.
Your Animatronic Prototyping Board sounds a little like puppeteering the robotic mechanisms (under microprocess control), if I understand what you’re suggesting.
It would be really helpful if you had a picture of an example of your board controlling a mechansim. Extra points for a java animation where we get to move the parts 🙂
Oh no, homework assignments! 😉