PollinOasis

PollinOasis

Story

In this project we were asked to design for two intelligences in order to establish a communication between them. The definition of intelligence that we were given was:

"The capacity of a system to perceive something from its environment, interpret it according to its own logic, act based on it, and adapt over time."

As we started reasoning on the brief, it was soon clear for us that we wanted to create something that would put in communication the human species as a first intelligence with other species as a second intelligence.

1. Context: Other intelligences care and communication

We decided to design a climate refuge for bees that could be distributed across the city landscape creating “climate relief” corridors (similar to ecological corridors). To design this, we wanted to create a system to communicate with bees and understand their needs and feedback about their use of the artifact across the city. As doing so directly was not possible because it would have required decoding bees' communication system, we needed to find a proxy for this communication, and see the artifact as a medium for the interaction of human-bees intelligences.

So we took scientific information of bees' habitat needs, thermal stress and perceptual skills to find foraging as a way to communicate with them. To the best of our capacities and resources we applied this information to the design of the artifact (passive and active design strategies), with the final goal of building an artifact that would make them thrive into the urban environment. We also considered how to map their use of the artifacts across the city (through an electronic system) as a way to communicate back to us their habitat preferences, which ideally would inform back the design of the artifacts.

This is the scientific evidence that guided our design:

Lack of habitat: The most known issue for pollinators is the lack of habitat in cities, but this issue is amplified in the context of climate change due to recurrent heat waves, which together with the heat island effect of our current urbanization system, creates an even more hostile environment for bees. A wide range of pollinators are under threat in urban landscapes: studies show up to a 43% reduction in pollinator species in highly urbanized areas compared with less developed areas. This decline affects bees, hoverflies, moths, and butterflies, all important pollinating insects; in urban landscapes, natural habitats get replaced by building and roads, taking away suitable environments for this species. Pollinating insects are vital for the reproduction of up to 90% of wild flowering plant species and many crop species. As urbanization causes more habitat loss, insect communities suffer and ecosystem become fragile.

Urban heat: Furthermore, temperature is a very important factor when considering pollinators' health and, as we know, cities are warmer than surrounding areas due to heat retention by buildings and roads. In some studies, wild bee abundance dropped by about 41% for every 1 °C increase in city temperature. Temperature strongly constrains bee physiology and behaviour: as ectothermic organisms, bees depend on environmental conditions to regulate their body temperature and maintain flight activity. Most species forage within a relatively narrow thermal window—roughly between 12–15 °C and 30–35 °C—with optimal activity typically around 20–30 °C. When temperatures rise beyond this range, bees experience thermal stress that can impair flight performance, reduce foraging efficiency, and increase dehydration and metabolic strain. Access to shade, water, and cooler microhabitats therefore becomes crucial for their survival in overheated urban landscapes.

General patterns of wild bee species richness as a function of urbanization. (a) species richness as a function of human population density; (b) species richness as a function of the proportion of impervious surface. Lines represent model predictions (dashed if not significant, α = 5%), and the shaded areas stand for the 95% confidence interval of these predictions.

Fragmented habitats: Another crucial aspect regards the distribution of green areas in the city. Parks and gardens are often isolated patches, this makes it difficult for pollinators to move between food sources and nesting sites. These patches are often surrounded by buildings, roads, parking lots, and concrete and because of this, the green areas are not connected to each other, which creates gaps in habitat. When green spaces are fragmented, one park may have flowers, but the next patch of flowers might be hundreds of meters away. Some pollinators, especially small bees, can only fly short distances (often 100–300 meters) and If the next food source is farther away, they may not reach it. In addition, flying across urban areas requires pollinators to cross roads, buildings and hot paved surfaces: these areas lack food and shelter, so pollinators use more energy traveling between patches.

All of this facts made us arise a simple but crucial design question:

How can we help pollinators to thrive in the city?

2. Purpose

PollinOasis is a green corridor of bio-based microhabitats for bees that provide food and shelter against hot temperatures in the city.

System structure:

Each artifact made of recycled cork has pocket habitats for bees with plants for foraging, shade and a cooling effect. The artifact will capture water from the rain and through capillary irrigation will distribute to each pocket. To ensure a wide distribution of the microhabitats in every street of the city, we plan to attach the artifacts to an urban element that is present everywhere: light poles. The artifact would be hung around 4 meters of height from the ground, in order to allow for a smooth experience both for the pollinators and for the humans, without one interfering with the other. As a way to make evident the return of bees to the city, the artifact detects bees' presence when they are foraging and emits a buzzing sound to the citizens walking by.

3. How do you fabricate? Key fabrication processes and steps

3.1. Geometry

We started approaching the geometry design by investigating what are the elements that could make the best habitat possible for the pollinators. We specifically looked for what types of adjustments would allow for cooling and giving a solution for resting and we found out that important factors were geometry and thermal mass of the structure, which was given by the thickness of the walls, pockets distribution and distance between the structure and the metal of the light pole.

We started designing many different shapes, getting more and more towards complicated ones to make them visually appealing, but at some point we realised that in order to make this system the most efficient and replicable, we should have taken a simpler path with an easier shape.

While thinking this we stumbled across a project made with 3d printed clay-like material in the hall of the university and we understood that such shape would have been very suitable for what we wanted to achieve. We so came to the conclusion that if we designed a structure made of 4-5-6 layers replicated we could have just rotated the layers of 90 degrees every layer in order to obtain some pockets where we would put our soil and so our plants.

The shape we obtained is this modular and scalable geometry, with layers stacked on top of each other; the plants will be placed into the pockets created by the rotation of the same shape layers. The top layer is slightly different as it has some higher walls to be able to catch rainwater and convey it into a central chamber where it's stored.

3.2. Material Recipe: cork

The criteria for material choice was the following. A key criteria that we included to consider bees perceptual system was to make the artifact attractive for bees by tapping into their ability to perceive UV light patterns in flowers that signal bees where to find food. We wanted our artifact to reflect UV light and in this way signal to bees they can find food there. See below other factors for material choice.

According to this, we discussed two options:

• White clay (kaolin) mixed with other elements like sawdust to increase micro-texture OR
• Recycled cork mixed with other elements, which we ended up choosing because it was water repellent and we didnt want water to overflow the artifact when it rains. Also this material is much more light-weight and this is good to attach the artifact in the light poles.

We adapted an existing recipe developed by IAAC to be used with robotic printing machines (and thus a bigger nozzle) to match a smaller nozzle of an Ender printer. The original recipe was:

However we modified: the quantity of water, instead of 1 dry mass to 2 of water, we did 1 of dry mass to 3.5 of water. Also we increased the % of xanthana gum to 30%, we replaced casein with sodium alginate, and we did not use potassium sorbate in this test.

3.3. 3D printing cork

We decided to create the object with a hacked 3d printed we had in our university, extruding cork paste from the nozzle.

This are the steps we followed to print the material:

-Placed a 2mm nozzle.

-Filled the container with the cork paste

-Air compressor pressure about 1.2 to 1.6 bars

Software parameters:

We used various softwares to be able to have the right communication with the hacked clay 3d printer to obtain what we wanted. In detail we had to use Grasshopper, Mamba and Repetier Host to be sure that the machine would do the right path while printing, without unnecessary and “dangerous” jumps between the parts.

All the detailed steps of the software adjustments are exactly the same as the one explained in another project we did printing clay, which you can find in chapter 3.2 in the file at this link: "https://www.hackster.io/living-tiles/living-tiles-64efbf"

Printing clay is not the same as printing cork and we had to adjust the material water percentage many times in a trial and error process.

After some tries looked like we got a bit our hand on the process and we obtained a pleasent shape that we wanted to save and bring to show in the presentation.

The print would have still continued going for some layers but we were satisfied with the amount of layers we already had and a bit scared that adding more layers would make the structure instable so we decided to stop the print in order to put aside a kind of finite part. When we stopped the print, for whatever reason the nozzle started to go down the z-axis and crushed part of the artifact, panicking we tried to lift it back up with the software but probably pressed the wrong direction and the printer dragged the nozzle all the way through the x-axis, destroying almost all of the object that we had developed.

After this little dramatic moment, we realized we couldn't rely on this process, because it was very slow and hard to get the hand on; unfortunately we couldn't have access to other printers that we had in our university that would have surely made this possible and sort of simple.

So given the time costraints, we decided to manually print one real size layer in order to show the possibility of the material and the shape.

We had to add some infill in order to give stability to the structure, so we decided to proceed with this star looking infill because it allowed us to have more contact points between the "petals" of the object and thus more strenght.

In order to give a better comprehension of the shape, we decided to 3d print a small scale model of the object to better showcase it. We were debating this decision for quite some time since printing something that would add more plastic to the environment was going a bit against our design direction. Since it was a 4 day intensive workshop, having the possibility to quickly build a shape that would help the others visualize the project was quite important, so we decided to do it.

3. Electronic system: sensors and actuators

We planned an electronic system into the artifact to:

a. Count bee presence

b. Record the count with hour mark

c. Emit a buzzing sound every time a bee is using the artifact

d. Record air temperature and soil moisture to match if there is a correlation. Due to time constraints we were not able to do point "d".

We ended up connecting a proximity sensor (ultrasonic distance sensor v2.0) to the Barduino board to:

• Detect when a bee is between 0 cm and 15 cm
• Count it as 1 bee visit
• Avoid counting the same bee multiple times while it stays there
• We placed the sensor behind the flowers to increase chance is a bee and not another random insect.

Moreover, we wanted to match the bee presence in the habitat pocket of the artifact to a buzzing sound broadcast from a embedded speaker. We thought of this because the absence of buzzing in our sonic landscapes is a sad signal of bee decline globally, and we wanted to bring back and augment this sound. So the buzzing of the artifact is intended for people rather than bees. Because bees are guided primarily by, floral scent, UV visual cues, microclimate, chemical signals, this buzzing sound is not a major attractor in foraging behavior. To trigger human-audible buzz with bee presence, we matched the bee count with the sound. Every count will result in the sound being only active while bee is there. For this we used a DFminiplayer with an audio track 0001 of a buzzing bee, and a little speaker connected to the miniplayer.

The final successful code is the following:

#include "DFRobotDFPlayerMini.h"

// -------------------- PINS --------------------

#define DFPLAYER_RX_PIN 12 // DFPlayer TX → ESP32 RX

#define DFPLAYER_TX_PIN 13 // DFPlayer RX → ESP32 TX

#define ULTRASONIC_PIN 1 // SIG of ultrasonic sensor → GPIO1

// -------------------- SETTINGS --------------------

#define TRIGGER_DISTANCE 15 // cm

// -------------------- STATE VARIABLES --------------------

bool isPlaying = false; // Tracks if sound is playing

bool interactionActive = false; // Tracks if someone is in range

unsigned long interactionCount = 0;

DFRobotDFPlayerMini myDFPlayer;

// -------------------- SETUP --------------------

void setup() {

Serial.begin(115200);

delay(2000); // give DFPlayer time to boot

// Start serial1 for DFPlayer

Serial1.begin(9600, SERIAL_8N1, DFPLAYER_RX_PIN, DFPLAYER_TX_PIN);

if (!myDFPlayer.begin(Serial1)) {

Serial.println("DFPlayer initialization failed!");

while (true); // halt if DFPlayer not found

}

myDFPlayer.volume(20); // set volume (0-30)

Serial.println("System ready");

Serial.println("Waiting for presence...");

}

// -------------------- LOOP --------------------

void loop() {

int distance = readDistance();

Serial.print("Distance: ");

Serial.print(distance);

Serial.println(" cm");

// -------- PERSON IN RANGE --------

if (distance > 0 && distance <= TRIGGER_DISTANCE) {

if (!interactionActive) {

interactionActive = true;

interactionCount++;

Serial.println("----");

Serial.print("Interaction START #");

Serial.println(interactionCount);

myDFPlayer.loop(1); // play 0001.mp3 in loop while present

isPlaying = true;

}

}

// -------- PERSON LEFT RANGE --------

else {

if (interactionActive) {

interactionActive = false;

Serial.print("Interaction END #");

Serial.println(interactionCount);

Serial.println("----");

myDFPlayer.stop();

isPlaying = false;

}

}

delay(150); // small delay to reduce noise

}

// -------------------- FUNCTION: READ DISTANCE --------------------

int readDistance() {

long duration;

int distance;

// Send ultrasonic pulse

pinMode(ULTRASONIC_PIN, OUTPUT);

digitalWrite(ULTRASONIC_PIN, LOW);

delayMicroseconds(2);

digitalWrite(ULTRASONIC_PIN, HIGH);

delayMicroseconds(10);

digitalWrite(ULTRASONIC_PIN, LOW);

// Measure response

pinMode(ULTRASONIC_PIN, INPUT);

duration = pulseIn(ULTRASONIC_PIN, HIGH, 30000); // timeout 30ms

distance = duration * 0.034 / 2; // cm

return distance;

}

4. Further development

In general we would like to test different typologies of microhabitats to find out what typologies are more preferred by bees (this data would come from the sensor data). The different typologies would be made by slight variations of geometry, material and choice of plants.

Sensors: We would like to include the sensors that measure microclimate conditions on site, like the air temperature and wind speed, as well as sensors that help us monitor the soil moisture to maybe embed another actuator that can increase water flow to the pockets on demand.

Water system: In a next iteration we would incorporate the water system with capillary irrigation into the artifact, right now this stayed in the render phase. Another aspect we could incorporate is a little fun to activate when the water is flowing to the habitat pockets to create extra cooling effect when bees are passing by.

Other pollinators: We could also study the behavior and thermal needs of other important pollinators in the city to make the artifact useful for more than bees.