Main | March 2008 »

February 19, 2008





Molo Design - soft seating

This soft seating element is fabricated with a non-woven, tyvek like material.

It is elegant, durable and unusual.

It is constructed with a honey comb design that provides light channels for the project that we have in mind.

The design is such that you can join multiples together with magnetic fasteners and change the topography of the seating.

the 9" high x 24" diameter seating element is $580 and the 12" x 24" is $690 and the 12" by 36" is $1140

I recommend the 36" diameter one.

Alternatively there is one that is 18" high x 24" diameter @ $980.

Soft Seating is my 1st choice alternative to the less than translucent sofa.

Cube thing

Here's the cubes I found, which is one of the possible alternatives to the couch: They come with an incandescent bulb in them, so at the least, we know light can get through, and there's already some amount of room inside to work with.

February 9, 2008


I ordered 2 of the PIR sensors which will be here in 1-2 weeks. I seen that the delay time is 5 sec., is that going to be fast enough? Probably for detection of people coming to the couch but not for making it do things when the people are next to the couch.

Arduinos Ordered

I ordered two USB Arduino boards which will ship to my apartment via US First Class Mail. They should be here in this coming week so we can tinker with them.

As another note that I will address in class on Tuesday, I will be absent on Thursday because of a job interview which will require me to fly out to the east coast. We can figure out what adjustments to make in the coming week.

I'll see you on Tuesday.

February 6, 2008

Continuing with ideas...

Well, I guess I'll comment on other people's ideas here and then put some other ideas as well:

•Matt's idea of using a LV-EZ1 ultrasonic rangefinder: It looks like a pretty good idea to me for sensing human movement near the couch. Each one is only $25 (of which we'd probably only want one or MAYBE two), they could be very easily implemented with an ADC in the uC seeing as their output is just 10mV/in the object is away. We could use them to tell if people are moving towards or away from the couch from either direction and change the couch's mood or other response accordingly.

•Phototransistor: It could also be a good idea, but in my mind, more for short-range to see if someone is sitting on the seat blocking the light it should receive. Mounting it could be an issue. It uses reflected IR light, but the couch may block that, or at least make it tough to use. If used we were to use something like this, we may have to put holes in the couch to mount them (probably not a great idea).

•Jason’s PIR sensor might have the same issue as it uses IR light as well. However, we will probably end up having to mount at least a few things through the couch that use light like this. And the PIR sensors would be especially easy to implement since they can just use digit GPIO pins as inputs to the uC.

•Fiber optics: At least looking around on, it seems like the cabling to do the fiber-optics is relatively inexpensive, but the light units to produce the light can be very expensive (we’re talking in the range of hundreds of dollars). If we can find a cheaper light source and a good way to secure the fiber to the inner couch surface, perhaps a little epoxy or something, we could possibly work with this.

•Fluorescent lighting: This is another possibility to explore once we have the couch since using fluorescent lighting, we could potentially mount a small number of these inside, flip a relay to power the lights of a certain color, and light the whole couch one color (or multiple colors depending on how they’re placed) pretty evenly. Again, we’d have to see how fluorescent lights look inside of it once we have it. Also, there is a delay with them turning on that could be problematic. But one cool idea is maybe putting a few black lights inside of it for fun.

•Clustering: Keith mentioned one type of “clustering? being the clustering of LEDs. Another kind of clustering that we could do for controlling the LEDs or other systems in the couch is basically have a different microcontroller control a different section of the couch. If we just ordered a bunch of samples, this would still be fairly cheap, and we could save a LOT of wire by doing this. Plus it would probably look a lot cleaner when we’re done.

•Prototyping boards: With the idea above about clustering, we’d have to think about whether we want to get demo boards for each cluster or just get a prototyping board and solder parts to it. Personally, I’d say all we’d need is the proto-boards, which I found one here that is pretty cheap and probably about the right size - for a small $2.95 one.

•It’s been a while since anyone mentioned anything like this, but I remember a couple people saying something about how we could have it query some kind of a web server to get weather information. I have a feeling if we tried to do anything even close to this, we’d end up just spending a bunch of time and money and end up with nothing. I don’t mean to smash anyone’s dreams, so if it’s a lot easier than I’m thinking, great.

So that's what I've got for the time being. I'll keep looking into stuff and post new things or improvements on other ideas I find.

February 1, 2008


Here's what I can remember with respect to things that have already been mentioned. If anything's redundant with Matt and Jason's posts, just ignore.

1) The possibility of placing white LEDs around the "hard" edges of the couch. They could be addressable for extra fun, or simply on one circuit to light the whole sucker up.

2) LED "clustering". Basically we can take one white, one bipolar red/green/yellow, and one blue LED and place them together in a single cluster, then place clusters in a grid organization around the visible portions of the couch. Granted, it would be difficult/impossible to make purple or orange hues in this sort of configuration, but it would be much more cost effective, and as I understand, RGB may not be too good at coloring mixing anyway.

3) Placing cheap white LEDs in random places to represent small points of light. Lighting all the blues and all of these random LEDs could produce a starry sky effect, depending on how the couch diffuses light. If it requires drilling for these to actually look like points, it's probably not worth it.

4) The possibility of fairly complex designs and patterns that could resemble something, like a sprouting/growing tree. Clearly this also depends on the diffusion of light through the couch.

5) Phototransistors would probably be the best way of checking the time of day, if we don't have some kind of internet system hooked up. The couch could fairly easily tell the difference between day and night, even if it was an extremely cloudy day.

6) Some links I found that had some interesting products that we may want to consider:
Bipolar LED, can display red, green, or yellow. $0.33 each.
Bipolar LED, can display red, green, or yellow. $0.20 each.
Basically any type of resistor we need, $3.50 per 100. Probably the easiest way to get them, since digikey requires you buy thousands of them at a time.
7) Spencer mentioned consulting with the stockroom to see if we can get resistors or LEDs for significantly cheaper.

8) Scheduling. Since the Doodle module only had a few availability slots, I think it would just be easier for us to post our schedule, and we can figure out any time that 4+ of us can meet and do meaningful work on the couch.
Monday after 3:00pm.
Tuesday between 1:00pm and 2:30pm, and after 3:45pm.
Wednesday after 1:00pm.
Thursday after 1:00pm.
Friday after 1:00pm.
Saturday and Sunday I am usually free, save for weekends that I leave to go home or am occupied with other things.

If you would please reply to this post in a comment with your availability, I would appreciate it.

9) In the upcoming days I'll be working on making a Gantt chart to better organize our upcoming tasks. I think once we get the couch and can confirm measurements and exactly how it diffuses light, we'll have a better idea of what will be feasible and what won't.

Some ideas.

     I'm just going to throw out all the ideas I can think of right now, so it will probably jump around a lot:
     The PIR sensor I found was at (one of the $6.90 ones). From the looks of it, it will be able to provide a 1 (digital) when it detects someone moving, and with the small size they could easily be mounted discreetly on the couch. They have 140 degree range, so using 2, we can cover every location near the couch.
     Fuhs suggested that we could use 3 to detect where someone is located near the couch. We could do something where it would detect someone approaching and get "happy", and if it detected them leaving without sitting on the couch it would get "sad".
     We could also try out using fiber optic cable to create a large amount of smaller point lights. These have the advantage of easily being able to create lights of varying intensity (for if we wanted to do the star idea), but might be impossible to use with the couch.
     At this point, we can't really figure anything out regarding specific lighting ideas, since we still don't know how the couch diffuses light, or how we can mount lights inside it.
     For driving any LEDs we might use, we can pretty easily run a near-infinite number of them using some LED drivers (and there is already arduino code for this, which could easily be expanded to however many LEDs we want). It's pretty easy to get a driver that can be dimmed, but it will dim each LED hooked up to it (so there would be groups of 8 or 16 that have to be dimmed the same). This might be fine if we just want to create a certain RBG color over the couch or a certain area (run all red LEDs together, and dim them together, etc), but might not be sufficient if we want every LED to be completely independant.


Hey guys, lets start organizing some ideas for this thing. One way to go about doing this is to categorize ideas. The way I look at it, the couch has to act like a living thing. That is, it has to perceive it's environment through sensors, and respond through..... responses :] .

"Receptivity & Responsiveness" is our theme here.

-Optical sensors (including PIR, and/or smaller IR sensors)
-Pressure/Force Sensors
-Vibration Sensors (accelerometers?)
-Weather (through internet query)
-Time of day
-Bluetooth devices/traffic
-Wi-Fi traffic

-Changing colors and other LED lighting (I assume this will be the primary means of communication for the couch.)
-Displaying usage statistics (display this on a webserver?)
-Animations (on a projection display, or laser display?)
-Changing Moods (affects other responses)

Here are some links to equipment we should consider evaluating for use on this project:
-PIR Motion Detector from (Jason, I think you found one that might work better if we want a smaller range... could you post that?)
-BlinkM module from, see the BlinkM Website for more info. Also see this website for a neat video of the BlinkM in action.
-The Arduino and Arduino accessories.
-RGB LED light strip I have one of these that I can bring in, although I am not overly impressed with it's red component or color mixing performance.
-A nice big 7-Segment LED display. Mike, I stole that one from you.
-Ultrasonic Range Finders I forget who's idea this one was, but the LV-EZ1 page has example arduino code.
-Infrared Detector/Phototransistor Module These would be a cheap & easy way to detect nearby objects.

I also have a copy of this book that I will bring in. It has a bunch of projects utilizing the arduino and example code for interfacing it with your computer and other objects. It also has some pretty cool project ideas :].

I will also bring in a few different RGB LEDs that I'm not using, we can evaluate their performance in lab. (I have both 5mm LEDs and 3W high power LEDs. I'm also waiting on some samples from Kingbright that I will bring in later on)

That's all I can think of for now. If anyone knows when the couch will arrive, please post that. I will try to get a hold of an arduino ASAP so we can start evaluating that.