There are overall, four phases in this project; 

1.  Observe the environment

2.  Translate the data and transport it to a program using satellite information.

3.  Make a neural network to predict any changes within the abysmal population

4.  Keep observing

For each phase, we will explain with scrutiny how it was processed and how we hope it will work in the future.

Observe the environment

*How are we going to know where to observe?*

We gathered, analyzed, and used satellite data provided by NASA (Earthdata - NASA) to identify the right spot to start our project.

The process of gathering the satellite data was the following;

·  First, as stated above, we gathered the information (pictures) and downloaded them.

·  After analyzing the satellite images, we thought it was better to put our little aquatic sensor ( part of the project ) in Monterey bay since it checks off both of the things we want, a deep place that has bioluminescent life.

Once we saw where the ideal place was to put our sensor, in other words, start the project, Monterey bay, it was time to really come up with how this sensor is going to look like and work like. We based some of this prototype out of another model published (Shibata & Sakagami, 2014). Although we based it off that prototype, we did add our special things for it to work even more efficiently under the conditions we needed.

THE AQUATIC SENSOR

We present to you, our dear judges… MICHIN

·  MICHIN - 

I want to describe the "anatomy" of this sensor with enough detail to understand its function and capacity.

 We wanted this aquatic sensor to go to depths of up to 600 m, so its design had to support pressure and it had to be able to move too. Overall its design had to be compact, ergonomic, and functional. It looks fish-like for the most part because we wanted it to mix with the environment so that the fish would not be scared of it. 

The "anatomy" of this fish-like sensor prototype is the following:

To be able to resist pressure it had to be made out of flexible material, in this case, a thin film of plastic. This film will be lightweight so that almost no additional weight is added ( to be able to float). This film will be sealed airtight.

On the inside of it is a more complex structure. First, we have an insulator liquid to help balance the outside pressure with the internal pressure. We also have a sponge-like capsule (lining made out of metal, filled with liquid and surrounded by sponge-like material) that helps the fish to stay upright and keep its position when it is not moving. Just beside those, we have our internal mechanism (surrounded by another thin plastic film sealed airtight).

This internal mechanism has a battery that helps to power the fin-like plate that it has for transportation purposes. It also has a microcomputer to help with the movement of this simulation of a fin.

Now, since we wanted it to observe what was going on with the abyss species, we wanted it to have a camera. We investigated and studied which type of camera was best for underwater purposes. However, due to time restriction, we tried to come up with the best solution at the time, although it might have multiple variables.

We decided it was best to use a camera that can see underwater, under pressure and can bring clear pictures. There is also a tentative idea of putting sensors inside this "fish" to help with the location. The concept is to build a more complex structure that can combine both a sensor and a camera for a more accurate underwater register of the creatures.

This sensor will be put afloat on Monterrey Bay and just observe quietly the environment while most of the action takes place on the neural network that will be described in one of the following points.

However, there is just one more point to mention here. The sensor will have a GPS made out of ceramic. A person with a phone near the area will have access to this information to tell approximately the live location of the fish-like sensor. It will also have a memory to store all the gathered data (pictures or videos). To recover this information, we will have to retract the sensor back to the surface and recover from the memory of the information. Afterward, it will be put back to the sea to keep observing.

Translate the data and translate it to a program using satellite information

Once we observed where to put the sensor, and came up with the idea of how it was going to work we now focused on the important thing, the usage of satellite data in our project.

We searched the coordinates of the place and other factors in that place that will help the system (neural network) to be more precise. Once we got all that information down to a database, it was time to come up with a program in this project, a neural web. As stated above, we translated all the information from the various satellite data provided by NASA with every image focusing on different things. This variation of topics also helped us to come up with a much more precise neural web. We downloaded those images using NASA´s worldview page and translated it into a python program. Now we got all the data we needed. We translated it to another data set and started building the neural network.

Make a neural network to predict any changes within the abysmal population

-This is the most fun and intricate part of this project-

I cannot stress enough the use of satellite pictures during this stage. It was thanks to all the different aspects of the ocean that were being demonstrated by that data that we were able to create an adequate and somehow precise neural network.

NEURAL NETWORK

The neural network was built by gathering many different pictures provided by the NASA worldview page ( https://worldview.earthdata.nasa.gov/) Those pictures were further expanded to pixels so that the overall picture that the network was going to analyze would be more precise. All those pixels were from different satellite pictures of Monterrey Bay. The different areas observed by the pictures were;

-Chlorophyll

-Open water latent flux

-Photosynthetically available radiation

-Sea Surface Currents

-Sea Surface Salinity

-Sea Surface Temperature

We chose to observe these factors to analyze any changes within the sea population and associate them with any of these factors and see how each might influential the deep-sea population,

To expand further the knowledge and accuracy of the neural network, any data gathered from the first trials of the fish-like sensor will be added to the information used by the neural network to help predict any changes.

Last but not least phase of this project…

KEEP OBSERVING

This is the last and continuous phase of the project, keep observing!

After the neural network was built and the sensor was put afloat, the only thing left to do is keep observing both the results that our neural network gives us and what our sensors are seeing underwater.

IXACHITLAN

I would like to make a little footnote here, to clarify the meaning of this name. We decided to give our team and project a name in Náhuatl since it is one of the mother tongues in Mexico.

Ixachitlan means abyss and michin mean fish.

We used satellite data for every single part of our project. It helped us to determine where to put our sensors, what are the factors that might contribute to variation in abyssal population and it helped us build the whole neural network.

https://www.youtube.com/watch?v=kxCrVW5B2fc

Satellite data - https://worldview.earthdata.nasa.gov/

                         https://eodashboard.org/?poi=W2-N2&sensor=Difference%20CO2&country=US

(APA citation)

Martini, S., Khunz, L., Mallefet, J., & Haddock, S. H.D.(2019). Distribution and quantification of bioluminescence as an ecological trait in the deep sea benthos. Scientific reports, 9(1), 1-11. DOI: 10.1038/s41598-019-50961-z

Shibata, M., & Sakagami, N. (2014). Fabrication of a fish-like underwater robot with flexible plastic film body. Advanced Robotics, 20(1), 103-111. DOI: 10.1080/01691864.2014.944213