This interactive sculpture, built by Andrew and Deborah O’Malley of Ottawa, ON, has 120 light panels each featuring a Shiftbrite and capacitive touch sensor. XBees for each side connect the sculpture to a controlling laptop.
More:
16 thoughts on “Interactive Light Sculpture”
Comments are closed.
That’s $450 just in ShiftBrites, $115 in Xbees, and that doesn’t include the touch sensors for each panel.
For something that could have been done for less then $100 maybe even $50 with raw parts and a multiplexing scheme. Ironically it was for an event to raise money for the homeless.
I usually defend things that about the art and not the technology, but this is a major waste. If they had only talked to one engineer….
Yeah, not to rag on the Arduino, which I love. But people who start with the Arduino sometimes do things the easy way instead of proper way.
I’m a little biased since I *sell* ShiftBrites (and distribute through Adafruit, Sparkfun, Polulu, etc). However, this seems like a perfect application. You might think an LED matrix is easy to make, but try doing 10 bit PWM at full brightness (no 1/8 multiplexing here!) on each color channel of 100+ RGB LEDs using just one microcontroller/Arduino. Not easy, and definitely not friendly to any deadlines that may be involved. This project will be viewed at several more shows, and the ShiftBrites, cables, and Arduinos are all 100% reusable for other projects down the road. A custom LED matrix would probably rot in a corner for years…that would be a true waste.
“but try doing 10 bit PWM at full brightness (no 1/8 multiplexing here!) on each color channel of 100+ RGB LEDs using just one microcontroller/Arduino.”
I have, it’s not that hard. So have other people: http://rgbledcoat.blogspot.com/ . 1/8 multiplexing does not reduce brightness very much if at all done right. And a ‘custom’ job could be easily designed to be reusable in the future. The link I posted has simple RGB LEDS on a breakouts.
I own a handful of shift-brights and they are useful. But for a project meant to help raise money for a cause to have overspent $500+ doesn’t sit right. A little basic engineering and more planing could had made this a cost effective solution for a good cause.
Maybe they only had two days to build it? Then shiftbrights seem right, but the my point of the waste of money is still valid. Just the blame shifts from the designers to the organizers that put the designers in such a time crunch.
For starters, I have an electrical engineering background, so I’m quite familiar w/ the alternative approaches for the components (lighting, touch sensing, control) used for this project. The specific components chosen for this particular project struck a good balance between the budget, time line, and manpower available for the project.
Regarding Shiftbrites vs. a scratch-built matrix, Shiftbrites were chosen for their ease of assembly and control, certainly w/ a deadline in mind, as Garrett alludes to. The ability to reuse and re-configure the Shiftbrites is also a major plus. Why re-invent the wheel? I’ve hand-built RGB LED arrays, and never want to have to do it again if I can avoid it. I’d rather spend time on programming animations/interactions then soldering/populating boards, connectors, and cable assemblies.
As for the charitable part, all the money for the charity came from the ticket sales for the event; the installation portion was a budgeted expense, so the actual cost of it has nothing to do w/ how much money was raised in the scheme implemented by the event organizers.
At the end of it, we have a working piece of hardware that the crowd really enjoyed and tied in effectively w/ the event and the benefiting organization’s cause, which must mean we did something the “proper way.”
I work on and routinely invest my own money into an education outreach program to visit elementary schools to get kids interested in Science and STEM careers. See http://www.sciencebrothers.org/ (Make and other blog editors that might see this, please contact me before posting anything on it.) The program is billed as federal grant funded, though little money has actually come from the delayed federal budget. So all the demos ( tesla coils, LED labcoats, POV displays, chemical reactions, etc) are built at home using mostly our own dime. As you can imagine, that makes me a little jaded regarding the cost vs results discussion. Just because it works doesn’t make it the efficient way. That’s extra money that could be used else where, no mater where it came from. I’d love to save $400-$500 on something and use that money to get more demos for the kids.
This display is unique and interesting. But it could have been achieved using a much cheaper method of construction. Wiring would have been similar and even reduced, LEDs could have been re-used, There’s very little waste a custom job could create if designed properly. It’s naive to think otherwise. A little extra time devoted would have yielded a much more efficient result. Where you getting paid to do this? Paid labor could definitely kill a budget and then this design makes perfect sense.
Like I said I’m usually the one defending the over-engineered designs of new makers and artists, but this struck a cord with me. I guess it’s bitterness looking out from my program’s lack of funds to another’s seemingly endless funds; both for a good cause. For that, I apologize.
But I’m definitely not a “Maker” that’s “gonna hate”.
We received an artist fee to work on this, but that along w/ our own funds all went into supplies and fabrication, so we didn’t take any profit or payment for our time on this project. When I say, we, I refer to a two-person team consisting of my wife and I.
I still stand by our decision to use Shiftbrites, as they do exactly what we need. Yes, a custom solution could have been made with less material investment, but the labour investment would have been huge and beyond our timelime — I don’t consider making and wiring four 30-node RGB panels as a trivial exercise. (Re)engineering a lighting system from scratch was not a priority or something we had a little extra time for, considering all the other aspects that needed to be completed, such as control software, touch sensing — there is no equivalent touch-sensing solution for the windows, so that used more than its fair share of our time — and fabrication of the structure.
Back to Shiftbrites, though: we were indeed fortunate to have a budget to work this way, but I still disagree that there was “waste” in this project, considering we can easily re-use the Shiftbrites. I’ll even vouch that Shiftbrites *are* an efficient way to get large RGB arrays up and running, since they reduce the complexity of building a custom array (or designing/fabbing/troubleshooting your own reusable RGB nodes). They strike a good balance between make-your-own and fully commercial solutions (such as light strings from Color Kinetics, for example).
Science Brothers seems like a great project; being part of a technology-focused artist run centre up here that does youth programming, I can appreciate the constant chase for funds for such initiatives. I wish you guys all the best in making science/electronics cool and accessible for kids.