July Full Moon Update 🌕
According to the "moon updated timeline" of the project, by the July Full Moon I was expecting to "Have a functional SymBioWare end-user friendly version." This is not the case 😅
So many things have been happening that it is challenging to keep track and report. And even though the initial planning schedule was definitely subjected to the Planning Fallacy, progress is at a continous pace with new ways that were not even slightly considered at the beggining. But the direction and goals reamin the same, with a couple of additions/refinements. In terms of a f"unctional SymBioWare end-user friendly version." we can expect it for Q2 2024 ;)
So here are some news about what's been happening! In my last lab note I mentioned the key aspects of what happened at the Zuzalu experience, which was definitely a turning point in the research - to a rich horizon I'd say - that of the so called Artificial Intelligence.
Since then I've been engaging and building a lot around the concept of AI avatars for non-human entities. And I'm happy to share some of the milestones there, like the demonstration of a LLM interacting with a live dataset while roleplaying a plant (a special one that I have nurtered and cared for. It even makes plots, thanks to the amazing PandasAI project.
eCanaris has a special place in my heart forever. This is because of a final experimentation we did together. As she was demonstrating being at the end of her life cycle, I initiated a harvesting process, taking it out of the earth carefully so as not to damage its roots structure. I was curious about the interaction between the roots and the wires we had manually weaved in it.
Right after taking it out of the earth, I put it into a cointainer with water. There was importante observation. the wires I had used connected to the roots had corroted, I guess because of the water in the soil. This explains some very drastic variations in the signal over time that I had been observing throughtot the almost 2 months of total experimentation. Still, having constante access to the data via a spreadsheet increased even more my intution to the InterComUnit sensor. The one I've been using since the beginning of my experimentations.
Anyways, I reconnect the same sensor to the same roots and started monitoring again these finals moments of eCanaris. The signal was weak, but alive (even stronger that the final signals I was getting when she was on earth). I left the device capturing data for some time and when I left the lab, I decided to take her outside of the water, so as to begin her drying process. But I left the sensor wires connected to her roots. The day after, I went through the data and was it was a memorable experience. A constanst decrease is visible in the data, where the signal gets weaker and weaker over the course of approximately 1 hour. Until...
Sunday, July 02-2023 at 17:29:36 the signal went to 0. No conductivity was taking place in the between the tow ends of the sensor. What that means? A very curious and unique emotion to me. Something I never really felt before in my life. It reminded of movie scenes where a patitent at a hospital has their vital signals measured (beep... beep... beep...) and it suddently turns into silence. Life left that body. This is what I feel that was captured here. An impressive observation in my perspective.
I'm grateful for what eCanaris has been teaching me and the portals she has helped me open!
"May her life not be taken in vain. May her spirit find their way!"
One of the big lessons is that I can't rely on G Spreadshets as my database, for many reasons. Thanks God I've enrolled at IFPR where I met prof. Gil Eduardo de Andrade. He plain created a way for a ESP to send data to a MySQL server and helped all the steps along the way to implementing it with my own case. So now the project has a proper database to receive data (and to read from). Amongst many advantages, this allows for the creation of a Monitoring Site for the live data (whenever I turn on the SBWPCB board (powered by Adafruit Feather ESP32S3). So that's a thing! Major infrastructure step working ☑️ Obrigado, Gil! Vamos escrever esse artigo!
A key milestone
Alright. Now I want to take proper time to disclosure a key milestone is this process/work! Since the beginning the goal has been to a direct perceptual connection accross the biodata and my own perception - via haptic sensory substituion. The Vibrotactile Platform for On-body Application (VHP) is the enabler of this! I have been so happy to have been granted one (2 actually) to experiment with. But for around 2 months I've been struggling to send data to it. In this case, via Bluetooth (or BLE) interface! The VHP has been designed with custom configuration and code and one part of it is a custom ble_com protocol, built on top of the adafruit bluefruit library, running on a nordic nrf52840 chip. Wow. It took me a lot of time and tabs just to discover and basically understand these terms. So much learning!
I tried many approaches to connect both the Nicla Sense and the Feather ESP32s3 to the VHP via BLE, but it was not until I realized (thanks to GPT-4 - thanks Giveth) that the ble_com sets the VHP as a peripheral device that I found a fruitful path towards connection. And it was due to the help of Artem Dementyev (VHP co-creator) and Cathy Fang (who has shared her code to connect esp and vhp) that I finally managed to establish a connection!
I'm so happy with this that it is hard to express! Also because the ESP connecting to the VHP is the same one I used for the eCanaris experiment, so it already has the ICU sensor built with it. So from here onwards I'll have the first means to feel th ICU biodata in my body. I hope to be feeling the first pulses this month still (next week hopefuly?)
The Focus Experiment
The Focus Experiment encompasses my direct connection with a "peer being" for the period a full moon cycle.
As it is to be imagined, being able to connect the VHP and the sensor board is key to the Focus Experiment. And untill now, as I had not been able to proceed with this connection, I was really unsure about when it was going to happen. Again, the original May 05, 2023🌕Begin Focus Experiment schedule is past gone. But now I can begin to replan for it to happen. I'll avoid providing a specific date, because this has proven to fail.
Even my previous declaration of a chosen peer being has changed! I had chosen the Peace Lily due to its high sensitivity to environmental conditions and quick visual feedback. And I really meant that choice, as illustrate by the fact that I also weaved his roots to wires, so as to later on connect sensors. However, as the experiment design evolved, I returned to the initial idea of multiple plant sensors (as opposed to only the ICU as with the eCanaris), including the Galvanic Sensor by Sam (as well as soil sensor, and a light based one. These sensors would not fit well into the baby peace Lily. Besides that, the initial wires I had attached to his roots did not stood the test of time and set loose as I was moving the vase.
The good news is that I'm being called to experiment with another plant that lives in front of the lab.
Resembling a snake for its leaves growth, they have been with us for some months now and lately I have been caring more for them. They've gotten this brigtness recently and it has called my attention. When I moved them slightly I've noticed how their roots are going from the bottom of the vase to the earth itself, full of mycelium. I tried my best to add a corrosion resistant wires to 2 of the pending roots, and put them back in contact with bare soil. Hopefully that connection is still on and accessing activity close to the mycorizal interface! But the structure of this plant is actually what has called my attention. With them I'd be able to use all of the sensors I'm considering. From roots to steam to leaves, providing a very good scan of their activities. I hope to begin reading data (to the new database) very soon! Hopefully the very same data can be sent to the VHP!
Machine Interface
Surfing onthe LLM as avatar for non-human entities waves, I've been more seriously considering how ML and AI can be used in the context of this research. Initially, I mentioned the Nicla ML embedded capabities and while that's still possible, there's something on the way: ML friendly data (or annotated/labelled data). eCanaris' data for instance doesnt' seem to be suitable for novel prediction with ML (or is it?) on the state/activity of the plant. My recent exchanges with the Earth Species Project (where I'm the new community net-weaver YAY!) have been fruitful in this direction and I believe that it'll be very rich if my next sensing experiment with the new database is designed with those considerations in mind, so as to more effectively make use of patter finding algorithms. To that respect I antecipate here that I'm coming up with new experiment project focused on that. The focus of this Second Phase will be on the interface of biodata and ML alrogorithms, with the goal of creating some kind of Tech Avatar for the peer being, which is capable of making future predictions about the state of its source entity, and articulate such predictions via natural language*
*the way this tech avatar communitcate via natural language (the logic behind it) has been deeply inspired by the Aanissin AI concept by Prof. Jason Edward Lewis, at the fundational position paper from the Indigenous AI Working Group.
So imagine... Myself as a human avatar for the peer being (via the haptics interface fed with its biodata) taking challenge predictions against the tech avatar (using ML to interpret the sabe biodata, and communicating it via LLM). I believe this is how the focus experiment will be looking like. Soon ;)
0 comments