Check out the Latest Articles:
deep data prototype

A step by step guide to the construction of the prototype device for the ‘Deep Data’ project by Andy Gracie.

Making the deep data prototype

a step by step guide of how to make the v1 Deep Data prototype device for subjecting tardigrades, or other organisms, to changing magnetic fields.

step 1: collecting organisms

Baermannn funnels for nematodes:
Collect soil samples from just about any location.
Construct Baermann funnels using plastic funnel with top diamter of approximately 8cm. Attach standard aquarium silicone hose to the bottom of the funnel and clamp shut. Mount funnel on stand so that it remains vertical with the hose downwards. Create packet of soil sample by placing a small amount inside filter paper or strong kitchen towel and wrap around to enclose.
Place the packet in the Baermann funnel, cover with distilled water and leave for 24 – 48 hours.
Open clamp to allow a few drops of water to collect in a petri dish or similar.

Moss samples for tardigrades:
Collect sun-dried moss cushions from white rocks, from natural stone walls and from terracotta roof tiles. Many tardigrades prefer calcitic stones as they need some calcite to build up their stiletto teeth. Mosses from forests are less appropriate – many tardigrades prefer mosses which become completely dry every few days and this is not the case with forest mosses. You will come across tardigrades from time to time also when screening freshwater samples from ponds. But as a rule the population density there will be much lower than in mosses. Avoid smelly and permanently damp mosses. Water bears like mosses which are free of bacteria and fungi. As a consequence it is advisable to store the collected mosses in a way that they can dry completely, e.g. you might expose your finds some time to direct sunlight or keep them in paper bags at a dry place.

method one:
find moss
soak in water overnight
squeeze out water
find under microscope using at least 20x magnification

method two:
Choose a moss cushion which just fits into the petri dish you have. Remove most of the loose earth particles if possible.
Place the moss cushion upside down into the petri dish and fill with tap water, rain water or de-ionized water. After some time the moss will be soaked and you will have to add some more water. When the moss will be fully soaked after a few minutes the water level in the dish should still be a few millimeters. Refill some water if necessary. Leave the dish with the moss like it is for at least some hours or overnight.
Take the wet moss cushion out of the dish und look out for the tardigrades by means of a dissecting microscope at a magnification of at least 20 x. In case you should have no dissecting microscope you might try a good mineralogist’s loupe (10 x). A black underground gives the best contrast and the light, e.g. from a torch, should come from the side. Look out for little animals which seem to move like puppies. Some of the water bears are red like bricks, most of the bigger ones whitish or transparent.

Searching mud for magnetic bacteria:
Body of water, preferrably natural, with black coloured sand, or a region with black coloured sand just underneath some other type of sand. It’s bad smell may help you find these spots, however the soil does not have to smell bad for it to be suitable. Experience tells us that the bacteria seem to prefer salty, undisturbed water. The presence of plants or photosynthetic algae or bacteria (as indicated by their green colour) is usually a bad sign, as these organisms produce oxygen. If the local oxygen level of the water is too high, it may kill the bacteria.

To collect the samples, it is best to use a small glass jar or glass fruit-juice bottle with a lid. A container that can hold 300 – 500 ml is ideal. Scoop soil and water into the container, ensuring the sample contains soil from all parts of the first inch or two of the sand. We are not entirely sure exactly where abouts the bacteria reside. You should aim to fill the bottle 1/3 to 2/3 with soil, with the remained being the water around the sample. When you take the container out of the water, fill it to the brim by using your lid to scoop more water into it. Place the lid on the jar loosely, to enable a small amount of oxygen to interact with the ecosystem. It is a good idea to label your sample and record some information about it- you may wish to return to a good site, or monitor the effect of seasonal or environmental changes on magnotactic bacteria populations.

Using a pasteur pippette, suck up a few millilitres of liquid and a little bit of soil from your sample, near the soil/water interface. You want a little soil but too much soil will make viewing difficult.

Rest your pippete on the bench with the end of the pippete facing one pole of a bar magnet and wait for a few minutes to allow the bacteria to migrate towards the magnet. Bacteria typically travel between 0.1 and 1 millimetre per second.

Step Two: Culturing the organisms

Algae cultures for tardigrades:
small cultures of tardigrades in 60mm glass Petri dishes in commercial bottled spring water at
room temperature in a shaded location. Feed Chlorococcum sp. algae, adding about 1 vol algal culture to 4 vol tardigrade culture. Algae and water changed once every ten days by allowing animals to settle and pouring out most of the water and algae and adding back bottled water, 4-5 times, then adding back fresh algal culture. Cultures kept in a shaded location. Hundreds of tardigrades per Petri dish can be reared continuously in the lab at room temperature in this manner.

CH (Chalkley’s Medium)
Freshwater Protozoa
Stocks per 100 ml
(1) NaCl 2.0 g
(2) KCl 0.08 g
(3) CaCl2 0.12 g
Stock solutions 1-3 5.0 ml each
Make up to 1 litre with deionised water. Autoclave at 15 psi for 15 minutes.

Soil extract for nematodes:
Site selection for a good soil is very important and for most purposes a soil from undisturbed deciduous woodland is best. Sites to avoid are those showing obvious signs of man’s activity and particular care should be taken to avoid areas where fertilisers, crop sprays or other toxic chemicals may have been used.

A rich loam with good crumb structure should be sought. Stones, roots and larger invertebrates should be removed during an initial sieving through a 1 cm mesh. The sieved soil should be spread to air dry and hand picked for smaller invertebrates and roots. It should be turned periodically and picked over again. When dry it may be sieved through a finer mesh (2-4 mm) or stored as it is prior to use.

Soil is prepared as above. 105 g of air-dried sieved soil and 660 ml of deionized water are placed in a 1 litre bottle and autoclaved once at 15 psi for 15 minutes, then again after 24 hours. The contents of the bottle are left to settle (usually for at least a week) and then the supernatant is decanted and filtered. The final pH should be 7.0 – 8.0.

SE (Soil extract with salts)

Into a beaker put garden or agricultural soil (preferably this one which seems to undergo no or minor treatment with chemical nutrients) and natural or tap water so that the supernatant water occupies approximately four/fifths of the depth. Autoclave for one hour, than filter. The liquid is a soil extract. Combine with water and stock solutions of salts.

* Soil extract: 100 ml
* K2HPO4 0.1% w/v: 20 ml
* MgSO4x7H2 0.1%w/v: 20 ml
* KNO30.1% w/v: 20 ml
* Distilled water: 840 ml

Step 3: Collecting and Processing Probe Data
Text file from cohoweb
go to:
select the probe you wish to use
select ‘create file’
select hourly or daily averaged
enter start and finish dates (in our case it was the entire mission)
select the data you want to use (in our case it was B Field Magnitude (average of fine scale magnitudes), nT
click ‘submit’

An application was created in Max/MSP/Jitter by Martin Kern which reads the data from the text-file generated from CohoWeb into series of values.

The data contains information about the probe and the conditions in outer space; such as
year, day (0-366), hours, probe’s distance to the sun, probe’s angle to the solar plane, magnetic field value (nT). Based on this information the application creates a logic to apply the B-Field value on the 3 magnets in the device. The visual feedback and the values from the magnetic sensor are manipulating this logic as well.

Step 4: Connecting to Arduino

The Arduino board is loaded with Pduino for Max/MSP.

Step 5: Building the device

Making the magnets:
The magnets were made from coils of 0.2mm insulated copper wire wound around 4mm steel bar cut into sections approximately 15mm long. After several attempts at 100% hand winding we discovered that the best approach was to fix the bar in the jaws of a hand drill, keep it at a steady speed and wind the wire onto the core that way. 100mm or so of wire was left as a leader before superglueing the first 2 or 3 winds in place, the wire could then be wound onto the rest of the length of core before being glued in place and returned to the beginning for the next layer. Normally we wound the coils six times.

Contrary to popular wisdom we found that there was no negative affect on the power of the magnet if the coils were not wound strictly parallel.

Casting and curing silicon:
To facilitate the combination of electronics, such as magnets, sensors and LEDs and the cultivation of biological microorganisms we developed a number of prototypes by casting of a transparent silicone rubber. Polydimethylsiloxan PDMS, which can be ordered from Dow Corning with the product name Sylgard 184.

PDMS preparation:
Mix the two components at a ratio of 10:1. a plastic cup is useful for a container.
Stir and mix thouroughly.
Leave it sit for 2-3 hours to get rid of the bubbles. If you have access to a vacuum you could do it in a few minutes.

PDMS curing:
The PDMS will be processable for about 24 hours if you leave it at room temperature. For curing its best to put it in an oven at around 60 to 80° C for about 1 h at least. Full curing should be achieved after 4 h.

Pour a thin layer of PDMS (1 or 2 mm) into the bottom of a 90mm plastic petri dish and leave to cure.
Arrange and glue the electronics and wiring onto this silicone base. Pre-cured silicone ‘bricks’ can be used to vary the height of components within the device if desired. These bricks will be completely invisible once the silicone has cured.
Pour more silicone on top until all the electronics and wiring are covered – making sure that the pin sockets are kept clean.
Once it is all cured, punch a 2mm diameter organism well in the centre of the device. This can be done with a 2mm biopsy punch or a very sharp scalpel.
The silicone device can now be placed back into the dish and a tiny drop of silicone added to seal the bottom the well.

Step 6: Electronics

The whole electronic systems consists of an Arduino board connected to a computer running MAX/MSP, three H-bridges, a power supply , a push button and four LEDs on the interface, an audio amplifier with speaker inside the interface, and the in-silicone bio-electronic device, ASTROBIO Mark I, which is placed on the microscope and contains LEDs, electromagnets, a hall-effect sensor and a well to habitate the bioorganisms.

Each elecromagnet is driven by an individual H-bridge (L293 B) to be able to run high currents through the coils controlled by the Arduino. The H-bridge is connected to the Vcc of the Arduino on pin16 and pin1 (chip enable) and the common GND. On pin8 the H-bridge is connected to the external power supply (addjustable from 3 – 12 V). The control inpupts (pin2 and pin 7) are connected to two of the arduino PWM pins and smoothed out by a large capacitor (100 µF) to GND. This allows to control the current flowing through the coils and switch the direction. To power the magnet only one of the PWM can be used at a time, to switch the polarity the PWM has to be turned of and the second can be turned out. The H-bridge takes charge of routing the current to GND

The Arduino (Duemillenova, ATMEGA168 microcontroller with USB2serial) is working as an I/O board to interface the computer to the electronics. SimpleMessageSystem was used for communication, which is a simple ASCII based serial protocol to communicate from MAX/MSP to the Arduino hardware and vice versa. The computer can control the magnets and the interface LEDs and read the sensor data from the magnetic field sensor (Hall-effect sensor) and the push-button. To srew and fix the wiring robustly a break-out shield was used.

To increase the sensitivity of the hall-effect sensor a rail-to-rail op-amp could be used, such as LM324 or TLC274.

Step 7: The Controller

The control interface was made from a slightly retro style plastic project box and served as housing for all the electronics (Arduino, H-bridges, lighting control pot, audio amplifier and speaker, space probe LED display and push-button). For the sound (radio emissions from Jupiter and Saturn) a mono amplifier was used (LM 386, 0.5 W) powered by the Vcc from the Arduino and connected to the computer line-out by a mini-jack and a speaker (8 Ω, 0,5 W) glued to the inside of the lid. Several ribbon cables connect to the ASTROBIO Mark I, one for the electromagnets and one for the magnetic field sensor. 4 red LEDs were mounted on the lid and connected to the digital pins of the arduino in series with a resistor (1 kΩ). They were used to display the choice of the NASA space probe data sets, which could be switched through using the push-button. The control of the LEDs and the detection of pushing events were done by MAX/MSP. To guarantee that only one step was done by a single push, a readout delay (500 ms) was performed in the MAX patch.

Step 8: Credit where credit is due

This prototype device for the Deep Data project was developed during Interactivos?09 at Medialab-Prado in Madrid
It would have been impossible without the effort and expertise of;

Georg Kettele
Martin Kern
Marc Dusseiller dusjagr labs
Yasser Bigay
Anders Restad tinkertank
Varvara Guljajeva varvara’s “hyperreal” zone

  1. […] some Baerman funnels. While Andy Gracie is an expert in these DIY funnels and used them in various hackteria workshops and his own artworks, there seems also other interesting options to add light and heat on top of […]

Leave a Reply

Your email address will not be published. Required fields are marked *