Black Hole Finder: public classification connected to active astronomy

The research challenge

Astronomical surveys and alert streams produce candidates that require interpretation. Black Hole Finder creates a public role inside that process: participants inspect image sequences, learn the visual task and contribute classifications that can support an active research workflow.

What Pocket Science built

The platform includes mobile and web participation, account and experience logic, classification workflows, candidate imagery, project administration and backend services. It connects an approachable public task to a domain in which timing, provenance and expert interpretation matter.

The project is maintained in an active relationship with Radboud University Astrophysics. Experienced-user agreement can contribute to follow-up logic, while scientific decisions and interpretation remain with the astronomy team.

Participation design

Classification is not made credible by showing users two buttons. Participants must understand what changes between frames, distinguish relevant visual patterns from artefacts and develop experience. The system records classifications and supports thresholds without pretending that public consensus replaces expert science.

Independent attention

Why we cared

A candidate can begin as a few unassuming frames on a screen and end with a robotic telescope looking back at that part of the universe. We love that connection. Black Hole Finder gives people a place inside a living research process and lets software, machine learning, experienced participants and astronomers each do the part they are good at.

The transferable capability

Black Hole Finder demonstrates custom scientific classification, participant learning, consensus and escalation, production backend operation and collaboration with an ongoing research team. The same architecture can support image review, environmental classification and other domains where human judgement complements automated analysis.

What this demonstrates

This project is part of the evidence behind Pocket Science’s custom development work. If your research requires a participant-facing method, scientific sensing, classification or production data infrastructure, tell us what you are trying to make possible.