00 FloLabs Group / TARRL

CosmoBrain

AI Search for Universal-Scale Information Networks

An open, remote research and product initiative using AI to study cosmic-scale structure, signals, anomalies, and information-flow patterns across public space data.

Open to the world No experience necessary Research + product

SCROLL TO EXPLORE

01 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

The big question

Can AI help us detect hidden structure, signal behavior, and information-like patterns in the observable Universe?

We are not claiming the Universe is conscious. We are building a disciplined system that tests whether public cosmic data contains patterns too large, subtle, or multidimensional for humans to inspect manually.

The inspiring idea: galaxies, filaments, light, spectra, plasma, gravity, and time-varying signals may form a vast network. The scientific task is to measure what is real and reject what is illusion.

STANDARDOur standard: every candidate pattern must survive data provenance checks, known-physics controls, statistical testing, and independent review.

02 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

What we are building

One initiative, three connected outputs.

01

Cosmic research engine

Ingests public astronomy data, learns normal astrophysics, creates cosmic graphs, and ranks anomalies for review.

02

AI Space View App

A public-facing app for exploring space images, maps, signals, and AI-discovered anomalies through an educational visual interface.

03

FloBrain Space Buddy

A conversational AI guide inside FloBrain that helps users learn astronomy, review datasets, organize research tasks, and explain discoveries.

The research team and product team work in parallel: science creates credibility, the app creates public engagement, and FloBrain Space Buddy turns the work into a scalable AI learning experience.

03 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

Why this is possible now

Open sky archives plus modern AI create a new way to study the Universe.

01

Open space archives

NASA, ESA, SDSS, DESI, ALMA and others expose enormous public astronomy datasets.

02

Foundation models

Self-supervised AI can learn structure from images, spectra, light curves, catalogs, and graphs.

03

Graph intelligence

The cosmic web can be modeled as nodes, edges, clusters, paths, voids, and recurring motifs.

04

Signal search maturity

SETI and radio astronomy methods provide useful controls for noise, interference, and false positives.

Reference archives and methods: NASA MAST, NASA HEASARC, ESA Gaia, ESA Euclid, SDSS, DESI Legacy Surveys, Berkeley SETI / Breakthrough Listen.

04 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

Scientific anchor: networks can rhyme across scales

The cosmic-web / brain analogy is a mathematical starting point, not a conclusion.

Published research has compared structural, morphological, network, and information-capacity properties of neuronal networks and the cosmic web. That does not prove cosmic consciousness. It does justify careful, quantitative network comparison.

Our method begins with measurable network structure, time-varying signals, anomaly detection, and evidence standards.

Network topology, clustering, centrality, motifs

Signal light curves, spectra, radio pulses, bursts

Evidence replication, prediction, falsification

Scientific basis includes Vazza & Feletti, “The Quantitative Comparison Between the Neuronal Network and the Cosmic Web,” Frontiers in Physics.

05 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

The data universe: not just space pictures

Beautiful images inspire people. Raw and calibrated data trains the AI.

01

Images

FITS, sky surveys, morphology

02

Spectra

chemical signatures, redshift, lines

03

Light curves

brightness over time, pulses, flares

04

Radio

narrowband, broadband, drift patterns

05

High energy

EUV, X-ray, gamma-ray, CMB

06

Graphs

cosmic web, clusters, filaments

Core data sources to integrate: NASA MAST, HEASARC, Exoplanet Archive, ESA Gaia, ESA Euclid, SDSS, DESI Legacy Surveys, ALMA, and open SETI datasets.

06 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

Research pipeline: learn normal before searching weird

Most “patterns” are ordinary physics or instrument effects. The model must learn that first.

01 Normal sky stars, galaxies, quasars, nebulae, survey artifacts

02 Normal variability pulsars, variable stars, transits, flares, supernovae

03 Normal instruments noise, saturation, radio interference, compression, calibration effects

04 Normal physics gravity, lensing, plasma jets, redshift, dust, known spectra

Only after learning “normal” can the system rank what remains anomalous.

07 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

What counts as evidence?

The project succeeds by rejecting weak patterns quickly.

1 Visible in raw data

2 Replicated across instruments

3 Survives known astrophysics

4 Statistically unlikely under controls

5 Information-rich or compressible

6 Predicts future observations

No prediction = interesting pattern.

Prediction + replication = research candidate.

This is how we protect the work from hype and protect contributors from weak science.

08 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

AI architecture: the cosmic foundation model stack

A multimodal AI stack that converts open archives into testable candidates.

1 Open science data lake

2 Multimodal embeddings

3 Cosmic graph

4 Anomaly + information engine

5 Validation pipeline

6 Candidate dashboard

Vision model

galaxy morphology and survey images

Spectral model

emission/absorption lines and redshift

Time-series model

light curves, bursts, pulses

Graph model

cosmic web connectivity and motifs

Information engine

entropy, compression, symbolic structure

Control filters

instrument artifacts, radio interference, false positives

09 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

Two tracks move in parallel

The initiative is both a research program and a product launch path.

01

Research Track

Build data pipelines, train models, map the cosmic graph, document methods, rank anomalies, and create review workflows.

02

Product Track

Turn the work into the AI Space View App and FloBrain Space Buddy: a visual, conversational, educational, and commercial space-intelligence experience.

The commercial project matters because it creates a public interface for learning, discovery review, recruitment, and long-term sustainability.

10 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

AI Space View App

A commercial product that lets anyone explore the Universe with AI.

01

Explore

Search and browse open space images, sky maps, spectra, light curves, and cosmic regions.

02

Explain

Ask AI what a galaxy, nebula, signal, light curve, or anomaly might mean in plain English.

03

Compare

View cosmic structures alongside brain-network, graph, and information-theory visualizations.

04

Contribute

Help classify patterns, annotate data, review candidates, and participate in public science tasks.

Product vision: the easiest way for students, researchers, lifelong learners, and space enthusiasts to see what AI is discovering in open cosmic data.

11 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

FloBrain Space Buddy

A conversational AI research and learning companion inside FloBrain.

01

Learn

Guided explanations of astronomy, AI, network science, SETI methods, and evidence standards.

02

Research

Dataset summaries, literature review support, task tracking, meeting notes, and project memory.

03

Analyze

Prompt-driven review of candidate anomalies, plots, metadata, and control-test status.

04

Build

Support for interns and research associates working on app features, datasets, workflows, and demos.

FloBrain Space Buddy turns a complex research initiative into a guided, persistent, AI-powered learning and productivity environment.

12 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

Open to anyone, anywhere

Remote. Global. Beginner-friendly. Serious about contribution.

No experience necessary to begin. Curiosity, reliability, persistence, and willingness to learn matter more than credentials at the entry point.

Join as an intern if you want structured learning and hands-on contribution. Join as a research associate if you already have stronger skills and want to help lead methods, review, or product work.

01

AI / ML

models, embeddings, anomaly detection

02

Astronomy

data realism, sky catalogs, known phenomena

03

Data Engineering

pipelines, metadata, reproducibility

04

Design / Product

AI Space View App, UX, visualizations

05

Writing / Education

explainers, curriculum, research pages

06

Project Ops

tasks, documentation, coordination

13 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

Role pathways

Start where you are. Grow into deeper responsibility.

01

Intern

Entry path for learners. Help with research summaries, data labeling, app research, content, testing, documentation, and guided technical tasks.

02

Research Associate

Contribution path for stronger builders and specialists. Help with model development, graph methods, signal analysis, review systems, and product architecture.

03

Team Lead

Coordination path for reliable contributors. Help organize tasks, mentor others, maintain standards, and connect research work to product output.

Everyone contributes to a real initiative: research outputs, AI systems, product features, education materials, and public demonstrations.

14 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

Build path: from open data to public product

No fixed timeline. The team advances through milestones.

01

Map

Identify public archives, datasets, sky regions, formats, metadata, and research questions.

02

Ingest

Create repeatable pipelines for images, catalogs, spectra, light curves, and graph-ready structures.

03

Model

Train baseline models to understand normal data and rank anomalies for human review.

04

Validate

Apply evidence standards, control tests, review workflows, and reproducibility checks.

05

Productize

Turn research workflows into AI Space View App features and FloBrain Space Buddy experiences.

Milestone discipline matters more than deadlines: document progress, ship small outputs, improve the model, and keep connecting research to product.

15 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

What contributors will produce

Visible outputs keep the project real, useful, and accountable.

01

Research briefs

methods, literature summaries, benchmark datasets, anomaly findings, and negative results

02

Open data maps

curated archive lists, metadata guides, sample regions, and reproducible notebooks

03

Candidate dashboard

ranked anomalies, evidence-ladder status, raw-data links, and review notes

04

App features

AI Space View exploration tools, visual explainers, search, annotation, and guided learning

05

FloBrain workflows

Space Buddy memory, task systems, dataset summaries, and research operations

06

Public education

lectures, explainers, website pages, short videos, and student-friendly guides

Even if no extraordinary signal is found, the work still advances AI astronomy, anomaly detection, science education, and the Space Buddy product ecosystem.

Join the search.

Help build the AI layer that studies the Universe as a dynamic network of data, signals, structure, and possibility.

Open to interns and research associates around the world.

Remote participation. No experience necessary to begin.

Contribute to both open research and the AI Space View App / FloBrain Space Buddy commercial product track.

BEGIN EXPLORATION

FloLabs Group / TARRL

16 FloLabs Group | Texas Advanced Robotics Research Lab (TARRL)

Selected sources and evidence base

The initiative is grounded in open public archives, peer-reviewed research, and reproducible methods.

01 Vazza & Feletti, Frontiers in Physics Quantitative comparison of neuronal networks and the cosmic web, including methodological caveats.

02 NASA MAST / STScI Public archive for optical, ultraviolet, and near-infrared astronomy data from major missions.

03 NASA HEASARC Public archive for high-energy astrophysics including EUV, X-ray, gamma-ray, and related data.

04 ESA Gaia Archive Large-scale public star-position, motion, and photometric data for mapping the Milky Way.

05 ESA Euclid Large-scale mapping of cosmic structure, dark matter, and galaxy evolution through public science releases.

06 SDSS + DESI Legacy Surveys Large public sky imaging, spectroscopy, and extragalactic survey catalogs.

07 Berkeley SETI / Breakthrough Listen Machine-learning and signal-processing methods relevant to technosignature search and false-positive control.