RAD@home: Citizen-Science Collaboratory

Updated 5 October 2025

RAD@home is a decentralized research platform that engages science-educated citizens in analyzing multi-wavelength astronomical data to identify AGN feedback and exotic radio sources.
The collaboratory employs distributed training, multi-wavelength fusion, and human pattern recognition to complement automated detection methods in expansive datasets.
By democratizing astronomical research, RAD@home offers accessible career growth, real-time professional mentorship, and robust data governance in resource-limited settings.

The RAD@home Citizen-Science Collaboratory is a decentralized, zero-infrastructure research platform established to engage science-educated citizens—primarily from India but with global relevance—in the direct analysis of multi-wavelength astronomical data. Its primary scientific mission is to identify and characterize relics and signatures of AGN (active galactic nucleus) feedback, as well as to discover exotic and rare extragalactic radio sources by leveraging trained human pattern recognition, particularly where automated algorithms underperform. Launched on 15 April 2013, RAD@home is notable for democratizing astronomical research in resource-poor regions, publishing discoveries using Indian telescopes, and training the next generation of citizen scientists for the era of large radio surveys and facilities such as the uGMRT and Square Kilometre Array Observatory (SKAO) (Hota et al., 2014, Hota et al., 2016, Hota et al., 14 Oct 2024).

1. Founding Principles and Organizational Model

RAD@home was founded as India’s first citizen science research (CSR) platform in astronomy with an explicit commitment to zero funding and zero physical infrastructure (Hota et al., 2014, Hota et al., 2016). Its participants are typically graduates with a BSc or BE, and entry is provided entirely remotely, allowing engagement from underdeveloped and geographically isolated regions (Hota et al., 2016, Hota et al., 14 Oct 2024). Professional astronomers act as facilitators and validators, particularly for follow-up observations and publication of discoveries. The collaboratory employs widely accessible online tools for training and collaboration: Facebook for discussions, Google services for file-sharing and communication, and NASA Skyview and SAO ds9 for data visualization and image analysis. Key features of the organizational workflow include:

Distributed Training: Structured online instructions in multi-wavelength image making (RGB overlays) and archival data mining (e.g., combining UV, optical, IR, and radio data).
Discovery Camps: In-person or hybrid week-long intensive workshops where select “RGB-qualified” participants receive advanced training in radio astronomy, galaxy evolution, and software tools.
Crowdsourced Reporting: Candidates for new discoveries are posted and discussed in blind (coordinate-hidden) sessions, then submitted using standardized forms for professional vetting and telescope proposal integration.

This organizational approach is tightly aligned with best practices in CSCW (Computer-Supported Cooperative Work), emphasizing modular task structure, loosely coupled collaboration, and robust data governance (Haines, 2015, Yadav et al., 2016).

2. Scientific Methodologies and Workflows

RAD@home exploits the human visual system’s superior capacity for recognizing faint, diffuse, or structurally complex phenomena in large astronomical datasets. The principal workflow involves:

Multi-Wavelength Fusion: Overlaying radio, optical, IR, and UV images to highlight peculiarities in morphology. Key tools include custom web-based RGB-makers and Python-based scripts for batch processing.
Morphological Typing: Participants are trained to distinguish classical radio galaxy types (FRI, FRII, DDRG, WAT, etc.) from “relic,” episodic, or otherwise non-canonical structures. Relics of AGN feedback typically appear as faint, misaligned, and steep-spectrum lobes (where the flux density scales as $S(\nu) \propto \nu^\alpha$ with $\alpha < -1$ ).
Quantitative Diagnostics: Flux density measurements across frequencies (e.g., TGSS at 150 MHz, NVSS at 1.4 GHz) are used to compute spectral indices via:

$\alpha = \frac{\log(S_1/S_2)}{\log(\nu_1/\nu_2)}$

This enables aging estimations for relic lobes, sometimes indicating plasma $\gtrsim$ 100 Myr old.

Environmental Assessment: Environmental effects (e.g., ram pressure, jet–galaxy interaction, or group/cluster context) are inferred by combining morphological radio features with ancillary multi-wavelength data (e.g., H $\alpha$ maps, H I imaging, or UV star-formation tracers).
Collaborative Data Vetting: Candidate discoveries undergo collective scrutiny, then detailed investigation (including high-resolution follow-ups with GMRT or MeerKAT under approved programs).

This methodology directly complements automated machine learning pipelines, particularly in cases where source morphologies depart from training set standards or require nuanced physical interpretation (Hota et al., 2 Oct 2025).

3. Key Discoveries and Scientific Contributions

RAD@home has produced more than a decade’s worth of discoveries across rare and previously unrecognized classes of extragalactic radio phenomena (Hota et al., 14 Oct 2024). Notable results include:

Discovery Class	Example(s)	Distinctive Features
Episodic radio galaxies	Speca, RAD-giant DDRG	Multiple misaligned or overlapping lobes, steep spectrum, evidence for repeated AGN outbursts (Hota et al., 2014, Hota et al., 14 Oct 2024)
Offset relic lobes	RAD-Rabbit	Host galaxy spatially offset from diffuse radio emission, consistent with galaxy in motion or prior AGN activity (Hota et al., 14 Oct 2024)
Jet–galaxy interactions	RAD12 (see below)	One-sided jets abruptly halting at a companion galaxy with lateral bubble inflation, not seen in standard FRII systems (Hota et al., 2022, Hota et al., 2023)
Ram pressure–shaped lobes	NGC 3898, RAD-Thumbs up galaxy	Extended radio lobes asymmetrically displaced; morphologies quantified via $P_{\rm ram} = \rho_{\text{ambient}} v^2$ (Apoorva et al., 13 Aug 2025)
Extragalactic radio rings/ORCs	RAD J131346.9+500320, RAD J122622.6+640622, RAD J142004.0+621715	Limb-brightened radio rings, intersections, or arc-like structures, often in group/cluster environments, spectral indices $\alpha\sim1.2$ (Hota et al., 2 Oct 2025)
Collimated synchrotron threads	RAD-jet-burl	Filamentary radio bridges possibly indicating re-energized magnetic channels or jet-ICM coupling (Hota et al., 14 Oct 2024)

Among these, the RAD12 system is notable as the first observed case where a major companion galaxy acts as a barrier to a one-sided radio jet, resulting in a $\sim$ 137 kpc radio bubble (“mushroom” shape) that is neither a mere loss of collimation nor moderate mean deviation as in Minkowski’s Object or 3C321 but a truly abrupt cessation and lateral inflation (Hota et al., 2022, Hota et al., 2023).

4. Platform Design, Motivation, and Socio-Technical Impact

RAD@home is deliberately constructed as a user-centric, collaborative socio-technical system, aligning with guidelines for successful citizen science platforms (Yadav et al., 2016, Haines, 2015). Defining features include:

Cost-Effective and Accessible: Open-source and free services ensure participation is not limited by institutional affiliation or resources (Hota et al., 2014).
Strong Feedback Loop: Learning and creativity are enhanced via real-time feedback, peer review, and direct mentorship by professional astronomers (Yadav et al., 2016). As participation increases, new discoveries motivate deeper engagement and further learning.
Diversity and Knowledge Sharing: The platform supports cross-category projects and encourages diverse participation through dedicated communication channels, online seminars, and documentation, fostering emergent community expertise.
Data Management and Governance: Attribution, peer-feedback, and transparent moderation systems ensure data quality without stifling volunteer contributions.
Scalability and Security via Middleware: Integration with platforms like CitizenGrid provides flexible deployment, secure credential management, and modular project scaling—enabling team-based computation and easy onboarding for new projects (Yadav et al., 2017).

The RAD@home model has proven particularly effective for career and academic growth: participants have progressed to co-investigator status on telescope proposals, and have secured internships or higher research degrees (Hota et al., 2016).

5. Interplay with Automation, Machine Learning, and Professional Pipelines

A major theme in RAD@home’s development is that human classification and discovery are not merely placeholders for future automation; they represent a qualitatively distinct, complementary form of cognition (Christian et al., 2012, Hota et al., 2 Oct 2025). Key contextual points include:

Pattern Recognition Limits of ML: Rare phenomena (e.g., exotic ORCs, intersecting rings, or “burl-like” jet features) are structurally underrepresented in current ML training sets, which are biased toward canonical double-lobe morphologies. Direct visual inspection remains essential for uncovering unexpected cases.
Human-in-the-Loop Model: Citizen science projects provide rich, structured annotations and large labeled datasets essential for developing new ML pipelines; in turn, hybrid human–algorithm loops can optimize classification throughput while preserving sensitivity to serendipitous sources (Christian et al., 2012, Marshall et al., 2014).
Integration with Observatory Data Pipelines: The future-facing model anticipates feeding citizen-created classifications and discoveries directly back into observatory archives, ensuring that rare or subtle structures are preserved in the scientific record (Christian et al., 2012).

The platform thus operationalizes a model in which distributed human sensemaking is structurally integrated with automated processing in scalable data analysis pipelines.

6. Scientific Significance and Directions for Future Research

RAD@home’s discoveries have broad implications for several domains of extragalactic astronomy:

AGN Feedback: The detection of relic radio lobes, sometimes with ages of 100–500 Myr, offers empirical constraints on the timescale and impact of AGN-driven feedback relative to star-formation history—a major theme in current galaxy evolution theory (Hota et al., 14 Oct 2024).
Environmental Effects: Statistical samples of ram-pressure–altered lobe morphologies and ringlike radio structures enable tests of how group- and cluster-scale environments mediate jet propagation, backflow, and radio plasma confinement (Apoorva et al., 13 Aug 2025, Hota et al., 2 Oct 2025).
Extragalactic Radio Morphology Taxonomy: The growing “zoo” of observed morphologies is forcing revision of classification systems historically based on the FRI/FRII dichotomy, supporting new physics-driven models of synchrotron structure formation.
Preparatory Training for Next-Generation Surveys: With SKAO and other facilities poised to deliver orders-of-magnitude more sensitive images, RAD@home’s methodology in training volunteers for faint-feature identification positions it as an essential human resource pipeline (Hota et al., 14 Oct 2024, Hota et al., 2016).
Synergy with Machine Intelligence: The association of human-detected outlier and rare-source discoveries with ML workflows suggests future research will focus on optimal hybrid strategies that combine expert intuition, citizen science, and automated algorithms for maximal scientific yield.

7. Technical and Mathematical Foundations

Quantitative analysis in RAD@home relies on established radio astronomy formulas and survey comparisons, for example:

Spectral Index Calculation:

$S(\nu) \propto \nu^\alpha$

$\alpha = \frac{\log(S_1/S_2)}{\log(\nu_1/\nu_2)}$

Synchrotron Aging and Morphology: Relic structures’ steep spectrum indices ( $\alpha \lesssim -1.4$ ) and surface brightness ratios across surveys (e.g., $S_{\text{TGSS}} \sim 10\,S_{\text{NVSS}}$ for certain relics) provide estimates of plasma lifetime.
Jet Power Estimates (for AGN bubbles):

$P_{\rm jet} \propto L_{\rm radio}^{6/7} V^{-2/7}$

where $L_{\rm radio}$ is the integrated radio luminosity, and $V$ is the emitting region’s volume (Hota et al., 2022).

Ram Pressure on Radio Plasma:

$P_{\rm ram} = \rho_{\text{ambient}} v^2$

where $\rho_{\text{ambient}}$ is the environmental gas density and $v$ the relative velocity of the host (Apoorva et al., 13 Aug 2025).

These technical metrics, combined with multi-wavelength qualitative assessments, constitute the core analytical toolkit of the collaboratory.

The RAD@home Citizen-Science Collaboratory represents a robust, multi-dimensional model for distributed research, leveraging citizen involvement, open data resources, and systematic training to generate scientifically significant discoveries in galaxy evolution and extragalactic radio astronomy. Its sustained output—encompassing rare morphological source discoveries, quantitative analysis, and integration with professional observatories—demonstrates both the necessity and efficacy of structured human participation in the astronomical big-data era (Hota et al., 2014, Hota et al., 14 Oct 2024, Hota et al., 2 Oct 2025, Apoorva et al., 13 Aug 2025, Hota et al., 2022, Hota et al., 2023, Hota et al., 2016).