Some more goodish news

Switching the RX and the TX and connecting it to the XBee has sort of worked. I no longer get errors from ZTerm and RX light is on. I also got the signal down to 3.3 volts by sending it to a 10K resistor then to the XBee from the to ground via two 10K resistors. I found this out here. The only problem is that the XBee is not responding and the terminal just parrots back what I have written.

Some good news

I switched the TX and the RX on my USB programmer and it started working again. I have the mini testing voltage through one of the analog pins. I have to play with the static resistance to get a wider range of readings. With a 10K resistor many things cap out at 255 and I have one been able to get 4 different readings. I have many more resistors at my apartment and will play with those tonight. Hopefully I will have a model done late tonight as well. I will try the XBee with the switched RX and TX as well.

When It Rains...

It doesn't pour, the gates of hell open, fire and brimstone rain from the sky...so the project is going about 6.02*10^23 steps backward. I think my Arduino Mini USB programmer might have bit the dust as I can no longer communicate to my mini. I get the Arvdude error and the RX light never lights. That or my USB Serial drivers are shot. The mold is not going too well too. I might have to redo the whole thing. I am trying to salvage it by sealing one half in plaster pinning the other half and using plaster fabric to temporarily seal it. This should work according to many people I have asked. But since everything is failing what is one more thing. If I have to redo it that will not be that bad it is just the time frame shortens. I will probably have to reduce the number from seven to four.

Behold Mold-zilla!!

So here is a picture of the mold. I will have the other half finished tomorrow and the actual cast done on Tuesday. Tonight I will try to get the XBee working and see how much voltage the Arduino can sense. In regards to the XBee, as far as I can guess it is a power issue with the Arduino needing to be powered by a 9V adapter and not the USB. Mainly because the Arduino is not following the laws of physics with the electrical output and what the resistors should change the voltage to (its output is inconsistant from 5V to 1.6V--the Arduino might be broken as well). I am missing an order because FedEx would not let me change my delivery to a weekend delivery. But it will arrive tomorrow via motherly transport from NJ.

An update...the arduino I borrowed from the closet does not work correctly if I power it from the 9V adapter.

Xbee Madness

My first problem was the XBee received power but its power light was not on. Then I grounded the second ground rail and the LEDs lit up. I could not get the Mini programmer to talk to the XBee. Then Jamie suggested that it was communicating at 5 volts when the XBee understands 3.3 volts. I powered it down using a resistor and a LED but the power was too low (1.7 volts). I asked around and the problem was the USB mini programmer cannot both power and send messages at 3.3 volts. I have now connected a standard Arduino to communicate and the Mini Programmer to power it. And...it still does not work. I played with the baud rate and figured the Xbee was set to 115200 instead of 9600. Now I get a prompt but I still cannot send.

Updates

I have ordered the accelerometers from Sparkfun. I need to get 9 to 3 volt voltage regulators and protoboard. I will get the protoboard tomorrow from my favorite shop in China Town. In other news I have finished soldering the breakout boards for the Xbees and have gotten one to power up. Now I need to program it. The Arduino Mini USB programmer seemed to fail at this. I will try a standard Arduino.

Picture of breakout boards:

Yes, Yes that is 5 Xbees in a container...and now we can geek out more with 16 ARDUINO MINIS IN ONE CONTAINER: (nice packaging Sparkfun)

Hopes, Concerns, Etc.

I have little worries about the programing and putting the electronics together (just hope everything come in on Friday). My current concern is with the mold which Steve will help me with. My main concern is with the location detection. I heard from one person that conductive foam does not conduct very well. Also creating the correct apparatus to connect the creature to the table and detect location is still not solid enough. Also still searching for coin/pancake vibration motors....

Note I just checked UPS, my JAMECO package will be stuck in NJ for several days. It is currently in NJ and will not arrive till Tuesday. I will call them and see if they could speed things up. Also Sparkfun lately has not sent me tracking numbers but the order should arrive tomorrow or Friday.

Theoretical Program

See if I am connected to the table
   Send out my location and current behavior.
   Receive Data from others.
     How many are doing want behavior?
     How far away am I?
       Adjust variable accordingly
   If I am connected am I doing anything?
     If I am not doing anything, see if I am influenced to do something,
       If I am then do it.
     Else every now and then beep and vibrate
   If I am doing something see if I should be doing something else.
     If I should be then switch
     Else continue my current Behavior
 

   If I get disconnected from the table
     Stop Behavior, vibrate nicely,
     Detect a gesture
       Apply a new behavior.

Foam is In

The conductive foam came in. The sample is TINY but long!! I need to experiment with it and see if it works. Then I can order a larger one.

Orders Away

Here is what I have ordered and bought so far:

Molding and Casting materials from The Compleat Sculpture

• I will use Dragon Skin Q for the mold.
  
  
  From Sparkfun
  
• Breakout boards for the ZigBee
  
• Arduino Minis
  
  
  From DigiKey
  
• ZigBee Series 1 (did not order from maxstream for they never called me back)
  
• 9V Battery Leads
  
  
  From Jameco
  
• Solderless Breadboards
  
• Piezo Speakers (I have some small headphone speakers to play with as well)
  
  
  Still needed:
  
• Coin Vibration Motors, these are oddly hard to find
  
• Accelerometers, I am looking for a better way than spending $40 a piece.
  
• Gloves, Lacquer thinner, spray paint, Pink Foam. These I will get from Home Depot. (luckily down the street from me)
  
• 9V LI Batteries (Radio Shack)
  
  
  Still Waiting for:
  
• I am still waiting for the conductive foam. However I heard it is not that good from someone so I am looking into another way.
  

Readings 5

"Androids as an Experimental Apparatus: Why is There an Uncanny Valley and Can We Exploit It?"
by Karl F. MacDorman

Oddly a few weeks ago in my Real Time/ Unreal Space class we were talking about the Uncanny Valley. No one brought up the subject of death. We were also talking about 3D modeling and not robots. However that in of itself gives little validity to the experiment. I did not like the experiment they ran. To say that a series of pictures changes a test subjects view is a lofty and conceited thing to say. Yes images do guide us to different opinions but to say that these people chose the speech because of an unconscious fear of death because another studies said they would prefer that speech is off. The correlation is off. A does not equal B, it might be an interesting relation but it is not an equal one. The whole experiment seemed to well guided as it set up the answers it wanted to hear. There are a lot of factors in the uncanny valley and to list it to one is preposterous. Death is a factor for some people as seen in the interviews but is it a major factor probably not. However the uncanny valley is something that most artists must deal with. Hollywood for some reason is embracing it with movies like Sky Captain and Beowulf. I think it is a poor choice but that is my aesthetic opinion. In robotics and some forms of 3D this is a problem. I like the idea of making the obviously unreal real like Chico MacMurtrie's work.

"Interactive Theatre"
by Cynthia Breazeal and crew

First off I like the way they spelled theatre. Also the comments from unnamed folks in our department would say about this..."robots are not art." I agree that the key to social interaction in general (not just robots) is moving away from standard input devices. You can see this with my fascination with gesture recognition. The fact that things must communicate in a way we can understand with avoiding the uncanny valley is very important to design.

Orders and Models

I finally found some sulfur free plasticine and started to create my model. I need to buy some more for the legs of the creature. I will order most of my other materials tomorrow as it is payday. The Arduino minis for me and the class should arrive tomorrow. I plan to create the behaviors for the gestures by next week and incorporate them into one of the models. After that comes creating the table (not too worried about that), linking them, and making them social (this I am slightly worried about). If all goes well I should be finishing up the last week in November.