Shadow Dance (Mikko Lipiäinen-Finlândia)

De nuvem
Ir para: navegação, pesquisa


Shadow Dance is a tool that helps adjusting dynamically the lighting conditions for garden plants to perform optimal photosynthesis by luring humans to cast shadows in garden space. A device to be developed measures the amount of light energy during the day and expresses the plants need of shadow/light by an intuitively understandable signal. The lesser the need for shadow/light the weaker the signal. The intuitive and playful nature of the signal lures humans to adjust lighting conditions.

The project was proposed by Mikko Lipiainen and developed with collaborators Debora Leal, Raquel Rennó and Thiago Hersan.



Mikko presented the general idea of the project to ¿Interactivos? group.



Mikko presented to the collaborators a list of devices to be used for data generation:

  • CO2 sensor (Parallax shield #27929)
  • LDR (light sensor)
  • Wheatstone Brigde circuit to measure resistance
  • Arduino Uno
  • 2 XBEE shields and explorer for wifi connection between computer and the sensors

Sensors will be controlled by Arduino and data will be analyzed using Pure Data


After a meeting with ¿Interactivos? tutors Leslie García and Ricardo Brazileiro was suggested the galvanic resistance of the plant as another input to generate data. To achieve that a Wheatstone Bridge circuit should be built.


C02 sensor


Working with CO2 module and gas sensors. Trying to understand how to preheat the MG811 CO2 sensor when attached to CO2 module. We are using Parallax CO2 module that uses MG811 sensor.

With 5v Sensor's maximum reading in arduino is 600 and when getting co2 data goes easily to 0, then recovering almost the maximum value slowly (30 seconds). TIP 1&2 3.86V, TIP 3&4:treshold voltage value can be adjusted by hand using small screwdriver.Control pin is sending voltage (WRONG!) TIP 1 is giving sensor reading.

Sensor has to be calibrated each time the computer/hardware is restarted.

Important: PREHEATING TIME! Give couple of hours to the sensor to get heated.

Wheatstone Bridge


The circuit was built according to the design presented by Leslie


    • Firmata (from Arduino Software - used to get data from the sensor)
    • X-CTU, Coolterm or similar to configure XBees
    • Pduino (connects Arduino to PD)


A continuous source of CO2 has been generated mixing warm (not hot) water, sugar and yeast in a separate bottle connected to another bottle by a rubber tube (to track the CO2 production) that is connected to the bottle with the plant and the CO2 sensor



Small test gas chamber built and CO2 sensor added in It´s still very hard to calibrate and calibration must be adjusted all the time

Problems with CO2 sensor:

  • troubleshooting
  • testing straight from the breadboard
  • red light should turn on when gas limit exeeded, not vice versa
  • added line to arduino code to give CNTL input a HIGH level value
  • strange conductance sensing phenomenon with sensor
  • sensor giving 10 times lower values from TP1 and TP2 (0,30-0,70 V), then lower values from TP3 and TP4 too
  • let the sensor to cool down

XBees were tested and work fine


Discussion about the concept of the project:

  • What kind of interaction we want between plant and the audience?

Leslie presented many different options on PD for mixing and producing sounds. After a discussion about the concept of interaction we would like to have it was decided that pre-recorded voice would be better to generate reaction from the viewer than music or other sound effects that would imply musical composition (adding another conceptual layer to the project). We´ve done some brainstorm to think about phrases the plant could "say" based on the input of light/shadow and CO2 levels.


We´ve started to hack a common battery radio to use it´s speaker for the plant.


CO2 sensor troubleshooting continues We´ve discovered that the recommended voltage to a good performance of the sensor is between 6,5 and 12 Volts, so instead of using the USB connection to the computer as a power source we´ve connected the shield to a 11v battery. As a result, the reading of the sensor became more sensitive.


CO2 sensor and LDR sensor testing:

10:30 plugged in, TP1+2:0,40 V
10:54 TP1+2 1,90 V
11:08 TP1+2 1,98 V
11:18 TP1+2 1,96 V // 1,98 V seems to be the maximum voltage
11:23:58 RECOVERED, does adjustment of gain affect to the recovering time?
11:27:36 RECOVER TEST 3 (lesser amount of CO2 than earlier)
11:28:17 RECOVERED (seems to recover faster)
11:45 TP1+2 2,04 V
11:49:45 RECOVER TEST 4
11:50:10 RECOVERED! // 1 V = 100 points in arduino
11:52 battery pack: 11,05 V
12:11 TP1+2: 2,01 V
12:11 battery pack: 10,60 V
12:19 TP1+2: 1,96 V
12:19 battery pack 10,50 V
12:25 TP1+2 2,15 V
12:25 battery pack 10,50 V
12:41 TP1+2 2,21 V
12:44:43 RECOVERY TEST 5
12:45:29 RECOVERED
12:48 TP1+2: 2,04 V, TP3+4: 1,89 V
12:48 battery pack 10,50 V
13:10 battery pack 10,30 V
13:10 TP1+2: 2,25 V, TP3+4 1,94 V
13:46 Battery pack 5 V, sensor gives strange readings,
TP1+2: 0,40 V, usage time: 5 hours
14:20 started with new batteries, battery pack 10,04 V, TP1+TP2 raising very fast to 2,04V, TP3&4 2,09V
14:46 Battery pack 6,58V, TP1&TP2: 1,48V, TP3&TP4: 1,37v, AnalogInput steady 356 (not good)
14:50 Battery pack 7,4V, AnalogInput 324-391 (working)


SG811's EMF:

    • 325mV=335ppm-400ppm
    • 315mV=600ppm
    • 308mV=800ppm
    • 305mV=1000ppm
Parallax CO2 module: TP1&2=190mV-225mV=335ppm-400ppm