The Idea of the Museduino
I never thought I would get caught up in the idea of building my own open source microcontroller, but lately I’ve been dreaming about it. Last week I dreamt that I met some women at a party in Brooklyn (waking life: I live in New Mexico) and they showed me this beautiful new board they had built. As they were explaining all the cool I/O , I woke up, super bummed that it was a dream and that hardware didn’t exist.
Over the last few months, we’ve been compiling data on expanding field of open source microcontrollers, not with any expectation that my love of the simple, perfect arduino Uno would be displaced, but because I’m curious about this emerging arena. Who is it, putting so much time and money to bring $15 boards to the market? Also: many of them suck and just don’t work. (When I read Emile Petrone’s post , “We Need More Embarrassing Hardware”, I could kind of see his point, I also want to say No.We.Do.Not.) There is plenty of sloppy hardware out there, and later on I might name names.
We were in the lab reviewing the our growing microcontroller taxonomy the other day, and I suddenly realized where all this was going. We were going to build our own microcontroller, optimized to do what we are always being asked to do- that is, design it specifically for museum exhibit designers and developers.
I said, “It looks like we will have to build our own board.”
Stan said, “I think we can build shields instead.”
I said, or I thought, “If it is a set of shields, they will have to be optimized for certain sensors, and they will have to communicate with each other”.
Miles, our research assistant, was like, whoa, okay, when can we start. Also, can we talk about how bad certain microcontrollers are? (yes, yes, later)
And so the idea of the Museduino was born (a word I think I can use until told otherwise) , a set of linked arduino Uno-based shields, (although they will fit on other boards — see below) optimized for the work of exhibit developers who know what they want to do, don’t mind a minor learning curve and assembling components, but would prefer to not be daisy chaining multiple boards and IDEs.
Museduino: sensing; designed for multiple sensors such as IR, PIR, PING, Audio levels, temp, light; basically, for controlling actuators based on people in a room, associated noise levels, time of day.
Museduino: AV: designed for output of high(ish) audio and video quality.
Museduino: 120V: designed to trigger relays for the use of wall power, good for controlling lights and higher voltage devices such as projectors,
Curious to know what microntroller works best for your project? Take our PlayBuzz quiz, “Which Microcontroller Are You?”.
Hardware is Hard (this is part 2)
But then! I started thinking about all the careful customization that I do, with motors out, PWM, specialised sensing like RFID, and I started to think I was way off base. There isn’t one way that exhibit designers employ interaction and responsiveness. Maybe delivering these tools would just insert another kit in to a process that cannot be effectively kit-ified.
But we decided to do it anyway. Build a system that allows for effective, robust buildouts of interactive exhibits, one that allows designers and developers to engage with microcontrollers in the steps past the prototyping/proof of concept stage. For me, planning for robustness has been a challenge.
So what the hell, here’s the Museduino system:
Museduino is an Electronic Exhibit Development Kit.
This kit addresses the deployment of a microcontroller-based system with sensors and actuators that is robust enough for a public installation. We will use open source hardware and software with locally available, inexpensive components, whenever possible.
The goal: scalabilty, extensibility and the use of ubiquitous hardware.
The foundation is a Arduino-pin-compatible shield with a connectivity system and remote sensor/actuator modules using standard CAT5/RJ45 patch-cord cabling.
A set of pins on the foundation board has jumpers so that the dedicated serial-communication pins can be used for inter-microccontroller communication if desired. Power is distributed via the CAT5 cabling to all sensor/actuators. Cables can be as long or short as needed. Satellite boards will have a color-coded solder mask for analog, digital, motors, relays and A/V.
Many microcontroller boards have Arduino-compatible connection headers. To name a few:
Beaglebone Black
Freesccale K-series boards
STmicro Nucleo boards
Netduino
ChipKit boards
Seeed boards
Adafruit and Sparkfun variants.
Intel Galileo
We’ll be test driving these in our museum first, and then looking for interested developers to check them out as well. I’d love to hear from anyone in exhibit world who wants to be a Museduino beta tester.
