Fiber-Optic Jungle Insect Traffic Taster

Here’s a device that can sense the movement of tiny animals simultaneously in several different parts of the rainforest and then shares this information in realtime by stimulating your tongue!

This how-to article will show you how you can build:

• Fiber Optic Insect Sensors
• Conductive Thread Switch Sensors for Insect Traffic
• Simple Audio Sonification for detecting changes
• Electro-Tactile Tongue Display

And you can build it ENTIRELY IN THE WILD. We built it on the side of an unnamed mountain in southeast Madagascar with the "Hacking the Wild" project: www.fisherlab.org/dissemination-lab.html. We use pretty basic materials found in an electronics lab or in nature!

Since we build this prototype entirely in the forest, it means you can pretty much make one anywhere! It also means we will share tips and tricks for designing and building in the wild along with incorporating natural materials in your design.

Why do you want to sense insects?

Scientists have all sorts of reasons for wanting to know this data for experiments. In general, though, we want to make technology that helps humans shift their attention from themselves, and focus more on all the other lifeforms (Somewhat like Boradkar’s sentiments in this “Design for All Life” write-up http://www.core77.com/posts/31264/Design-for-All-... Interacting with creatures that aren’t like you helps you appreciate them!

Ok, but why zap your tongue?

I wanted to be able to monitor several spots on a tree and the ground all at the same time. Usually we think to just look at things, but your eyes, and visual attention gets worn out REALLY QUICKLY. Also it’s difficult to closely stare at a branch of a tree, the trunk, and the base of the tree all at the same time. So I wanted to map this information directly onto your body in a more persistent, and maybe ambient way. Your tongue has great spatial resolution, and from reading about other tongue display projects (like in Kaczmarek K. “Sensory augmentation and substitution” 1995), more recent incarnations like Dublon’s “Tongueduino” (http://gershondublon.com/2012/04/tongueduino/) , and my own work on electro-tactile stimulation (like The Electric Eel https://www.comingle.io/howto/the-electric-eel). It also leaves your eyes and hands free to work on other things!

The jungle is also a noisy and dark place. this can make hearing a signal, or seeing an alert trickier. Zapping your tongue gives you consistent alerts in most environments.

What can you detect?

Our initial designs in the jungle were pretty rough, but with some refining of the design and the code, I think it could be pretty great! Still, we could totally detect larger insects walking over (like beetles, millipedes, bugs), and we had some promising, but sometimes fickle, results with smaller insects like large-medium sized ants.

Don't the fiber optic wires mess with the insects' movement?

They don't seem to mind at all. First, these lines are quite thin, and even the smallest ants walk over no problem. Second, in the jungle, the critters are used to manuevering over all sorts of impediments and vines are quite common, so really this doesn't seem that out-of-the-ordinary to them.

Oh, so you solved all the problems of sensing insects

TOTALLY NOT. But I think we came up with a really interesting design that could be applied in a lot of areas for science, art, and fun!

Why build stuff in the jungle, why not just build it in a lab? Won’t your things work better?

Possibly, but the rapid feedback you get from being embedded directly in the jungle helps find, and answer, all the tricky little problems that come up when designing. And designing electronics for use with animals in the rainforest is one of the most problem-rich areas I have found. Making in the field also forces you to pay attention and explore the environment around you and discover interesting details you may have otherwise overlooked.

• E-Waste Plastic Bags
  • Don’t leave your garbage all over the rainforest!
  • (This includes all those little bits of plastic from when you strip wires that, when you are in the lab, you just kinda pretend disappear!)
• Sensors
  • Fiber Optic Sensors
    • Fiber Optic lines (Strings of hot-glue could potentially work)
    • Sandpaper or Emery Board
    • Photosensor
      • Standard Photo-resistor (easiest to work with we felt)
      • LED-as-sensor-and-display (Prettiest, but includes capacitive problems when in a wet jungle)
    • Double-sided Tape
    • Black Tape (or something black to tape over, like Velostat)
    • Heat-shrink Tubing
  • Conductive Thread "Wimper Switch" Sensors
    • 1ply conductive thread
    • Strip of conductive fabric
    • Stretchy, Non-conductive Fabric
  • A Dead ant glued to a fiber optic line (for testing!) (very useful!)
  • Thumb Tacs (For attaching to trees)
• Brains (Microcontroller and Audio Sonification Circuit)
  • Arduino (Could use an Uno, or Seeeduino like I did)
  • Battery (Lipo rechargeable 3.3V)
  • Wires
  • Wire-Strippers
  • Solder
  • Soldering Iron (We used a Butane powered one for the field)
  • Heat source for Shrink Tubing (Butane powered soldering iron worked!)
• Tongue Display
  • Two Leaves
    • About the same size
    • One slightly larger than the other
  • Conductive Thread
    • We used silver plated copper: #3981 kupfer versilbert 7x1 fach verseilt http://www.kobakant.at/DIY/?p=379
    • They are solderable too! Which is nice!
    • (Hannah says the other all-copper thread (instead of this silver plated one) is bad for your health)
  • Thin Ripstop fabric
    • To sandwich between the leaves to stop them from tearing with the thread
  • Wax
    • We used wax from honey found in the forest to seal the tongue sensor!
    • You could just use regular wax, or glue (but it’s not as cool!)

Making in the Wild

Hannah Perner-Wilson (Plusea www.plusea.at) and I made this as part of our “Hacking the Wild” project joining a biological expedition in Madagascar with the renown Ant Explorer, Dr. Brian Fisher (www.fisherlab.org). We wanted to collaboratively design tools for science with biologists and nature. We carried electronics prototyping gear deep into the rainforest, to an unnamed mountain, while hunting for a mysterious, yet-undescribed new specie of ant (read about it here: https://openexplorer.com/expedition/disseminationlabmadagascar). We wanted to figure out what sorts of tools would enable us to build in the wild, and it worked out great! We even built our own whole jungle workshop woven from sticks and stuff. It's a rewarding experience building electronics in such harsh conditions, and it is also quite empowering for design.

Sensing Ants: A Huge Challenge

Sensing Insects gets relatively little love

(This is just a tiny rant from Andy’s point of view! If you just want to start detecting critters and zapping your tongue, skip to the next step!) In studying living things there’s a hierarchy of things we are able to sense digitally. It all comes down to money and size. It goes like this:

1. Humans
2. Livestock
3. Large Mammals + Birds
4. ……
5. Tiny insects (like ants)

At the top you have Humans. We have gotten really good at making technology to track humans (either to extract money from them (Commercial industries), or hunt them down and kill them(military Industries)). We can use cameras, image recognition, all sorts of other biometrics, or put sensors directly on them or things they wear (or drive, or live in).

Next you have big animals that humans breed and use for food. Not only have big industries pumped lots of money into gathering metrics and moving these creatures more efficiently, they are also generally pretty big. We can put big radio collars and RFID Tags on them, and we can build the environment in which they live.

After that, if you look at animal tracking technology, it starts being a lot less fancy. Most of the projects you see are either for conservation projects, or laboratory studies. Also the animals used are, again, of the larger sort, like bears and deer and birds at the smaller end.

What I am trying to say is that comparatively little funding goes towards technology for insects, and they are even more difficult to work with because they are so small!

Sensing Insects

Some scientists have been making awesome software for monitoring insects (mostly ants! http://www.antracks.org/, bio-tracking.org), but usually this needs computer vision, and nice laboratory conditions. In the ants' natural environments, you have to deal with some tough problems:

• Arbitrary geometry: creatures can be on, or under leaves, around stems, sides of vines
• Tiny: can't put sensors on most little insects, they also don't move things that much, or take up much space, or have lots of pressure
• Fickle Lighting: Jungle is pretty dark, but sometimes light comes through in patches
• Insect Camouflage: some insects are the same color as their surroundings
• Insects are Dry: And generally not very capacitive or conductive
• Insects are room temperature: a lot of tiny (non-flying) insects like worker ants are pretty much the same temperature as the background
• Dispersed: Many insects are more interesting to study as a group, so you want a sensor that can be distributed over an area to monitor their traffic en masse. This means it has to be cheap and adaptable to different parts of the environment (ground, thick limbs, thin limbs)/

For years now, I have been trying to figure out ways to sense ants in the wild. I have tried: thermal sensing, sound, vibration, capacitance, electric field proximity sensing, computer vision, VCNL 4000, and optical mouse cameras (I think the most successful out of the batch), but none really worked that great. Now, with Hannah and Brian’s awesome help, we came up with this potentially awesome design!

(By the way this is an awesome challenge open to everyone! And if anyone has suggestions they want me to try out, totally let me know!)

The basic concept is that we are making a flexible photosensor ground wire. Fiber optics are pieces of extruded plastic that can carry light from one end of a flexible tube, and emit it out the other end. Try it out, put an Led on one side of a bundle and see how the light moves through it!

If you scratch up the sides f the bundles however, the fiber optic wires will also emit light from the sides. The same works in reverse! If you put an a light sensor at one of the bundle of fiber optics, and the sides are scratched up, you end up with a long, flat surface that can sense the amount of light over it.

1. Grab your bundle of Fiber Optics
2. Arrange the fibers so they are FLAT
3. Tape both sides (near the ends, but not AT the ends) with the double-sided tape
4. Attach black material over one end (to block light coming though)
5. Bundle up the wires on the other end into a cylinder
6. Put a small piece of heat-shrink over them to keep them bundled
7. Shrink it with your soldering iron/heat gun
8. Grab a 5mm LED, put it at the end of this heat-shrunk tube and fit another, larger, BLACK heat-shrink tube over them both
9. Heat shrink this larger bundle to finish the OCULAR CAP
10. Turn the LED on (with a coin cell battery) (you shouldn't be able to see much)
11. Take a piece of sand paper and start sanding them on a flat surface
12. While the LED is on, you should be able to start seeing the light shine through
13. Continue until it glows pleasantly
14. Replace the LED with a photoresistor

Then lay it on the ground where insects like to go, or wrap around a tree and secure with a tac, and test it out. You can even just hook up a multimeter to it and see if the resistance changes when bugs walk over.

This interface can be extra beautiful, but might not cover as wide of an area as the full sanded band. The basic idea is to string the regular (unsanded) fiber optics through leaves and mushrooms to make pretty glowing points of light (or sensory positions).

Hannah and I were thinking of a couple different ways this could be used:

Simple Gate + Display

You could have an led in a certain position with a gap between it and the base of the fiber optic bundle. Then when a creature passes through this gap, the light will flicker as pretty dots on the leaf. The problem with this one is that the insects movements are limited to this very little area of where the sensor gate is. Would work well if placed near a nest entrance, or other place where you knew where the critters were going to be.

Point Sensor

A cool thing about this project is that you could potentially use an LED as the light sensor. With that, you could detect AND display the results with the same light-tubes. This method has an advantage over the other in that the area that insects could be detected is somewhat more dispersed, but the change in the light hitting the sensor overall will not have as much impact from one little point being covered by an ant.

Our Tests

Our initial trials didn't function that great as a sensor, but it made for really gorgeous displays, especially at night in the jungle.

These are pretty simple to make.

1. Grab your fiber optic bundle
2. Melt one side of them so that the tips form little balls (this stops them from pulling through the leaf or mushroom)
3. use a large needle to make an initial hole
4. thread the fiber optic through it
5. bundle them together (like with the band-sensors) using heat-shrink tubes at the other end
6. Attach an LED at the bottom (or photoresistor)

When Hannah came up with this idea, she referred to it as a basic "Wimper Switch." I thought this was some sort of electronics engineering term, but actually she was just meaning "eyelash switch" but didn't really translate it from the german. I like the name anyway.

The concept is just like her Stroke sensors: http://www.plusea.at/?p=2236

But made as a band for insects to try to trigger as they walk past.

You don't even need to program anything, just complete a connection between the lashes and the band and hook that to a piezo buzzer or LED.

The initial prototype we made didn't actually really work well at all because the thinnest conductive thread we had was way too thick. Hannah is working on a new version of them with super thin fabric she had at home that is truly like an "eyelash" in thickness.

1. Cut a strip of stretchy non-conductive fabric (With some thickness, like the neoprene she had)
2. Cut two strips of conductive fabric
3. Sew one conductive and one non-conductive band together via lots of little strings that you let dangle over the edge (Remember use the thinnest conductive thread you can find. Ours didn't work that well because even at 2ply it was still too stiff)
4. Wrap the extra piece of fabric around the target zone of a tree
5. Wrap the eyelash band around this one so the eyelashes dangle over the other conductive strip
6. use aligator clips to make a circuit between each strip of conductive fabric and something like an LED or piezo buzzer so that when the lashes touch the strip, you get an alert!

Your tongue feels disproportionately big in your mouth, so start putting different sized leaves in to get a good fit. I actually traced my own tongue on my notebook to find out how big it was.

Ideally for a body electrode, you would have a circle surrounding another circle made of two different electrodes. This can give you the broadest sensation. We tried to sew this by hand.

1. Find an appropriate sized leaf*
2. Cut a piece of ripstop fabric the the same size
3. Sew 3 dots for the ground wire (through the leaf and the fabric)
4. Sew 3 separate threads around each ground wire dot that will be your three different cathodes
5. Make sure to not short any of your lines out
6. Solder the 4 lines (3 electrodes, 1 ground) to some header pins
7. Connect some wires and make sure your tongue display is actually zapping correctly!

*WARNING: the jungle is full of all kinds of plants that are so used to getting attacked by creatures that they produce all kinds of different defenses. This makes it not the best idea to just go putting random leaves in your mouth. The smart thing to do would be to test the leaves first or know what you are putting in. I didn't actually do that, but i'm a little stupid this way.

Whew! You worked really hard for a while in the jungle. Have some jungle popcorn to get your energy back up! Lots of salt!

Ok! Back to work

Connect your photoresistors to the Analog Inputs on your arduino.

Connect your positive electrodes on your tongue display to the PWM pins on your arduino.

Hook the ground electrodes to the ground.

You can also have an extra output going to a speaker for more feedback

Organic Breadboards

Originally, Hannah came up with this fun idea to use these mushrooms we found as breadboards. They work pretty well, hold things in place, don't get burnt. But for sensitive applications, especially ones that involve capacitance (like using an LED as a sensor), these screw up the electronics a bit. (But they sure are cool looking!)

Get the code for your Traffic Taster right here:

https://github.com/quitmeyer/MadFieldCode/blob/mas...

Tips for Jungle Coding

We carried expensive, power-hungry computers all the way out into the wild with us, for really one main reason: programming microcontrollers. Battery life is previous, we were only able to recharge the computer 1.5 times during the 2 weeks we were out in the bush. By rationing our computer use though, we perfectly used it up for several projects.

Some handy tips for programming in the wild:

• Get huge batches of simple test code for lots of projects on hand
• Write pseudo-code, and then actual code, out by hand on paper. This will help you think through it, and when you are ready, you can just hurry up and write down everything on the computer.

Tech Documents in the Wild

You probably don't have internet in the jungle! and you need to conserve energy. It's a good idea to get a cheap-e-ink PDF reader, and get offline copies of any technical documents you might need to know how to hook stuff up properly! The amazing folks at Sparkfun gave me a huge data dump of ALL their PDF's for all their products! And it's here for free for you and your next jungle adventure! https://dl.dropboxusercontent.com/u/7533057/Tech%2...

https://github.com/quitmeyer/MadFieldCode/blob/mas...

In a crazy place like a rainforest, every device needs to become super site-specific. Figure out exactly where you want to install your sensor, and start building it there! This will make sure all the lengths of the wires and everything are correct.

Sensor

We were able to sense beetles, and large ants. We got fickle results with small things.

Sensing at night, with a constant light-source worked better.

Future ideas would be to use constant IR lighting. Even better would be modulated IR lighting.

I think some of the greatest improvements could come from refining the software to do smarter stuff with how it detects the high-frequency movements of the insects against the low-frequency changes in environmental light.

Tongue Display

This worked quite well and was very noticeable (and unlike sound or vision, was not affected by the changing noises or lights in the jungle).

The wax tasted good.

It made me drool extra.

I could have embroidered the electrodes more nicely. Ideally i would have circles of conductive fabric cut out, and much higher resolution.