Poster Abstract

P3.2 Emmanuel Joliet (Caltech)

Theme: Data science challenges: tools from statistics to machine learning

Control any telescope with Firefly visualization tool and machine learning

During Caltech Astroinformatics 2019 Hackathon, i worked on a project proposed by Alberto Krone-Martins and came up with the idea to extend the project using Firefly as a feature-rich visualization tool to enhance the observing experience. Firefly software details can be found here: https://github.com/Caltech-IPAC/firefly
The original hackathon project idea was about improving and making more accurate the autonomous telescope & satellite control pointing using machine learning from a training set of few images to eliminate the 'noise' coming from the control system.
Once the control loop is complete and working, the observer could use Firefly to select the target and activate the telescope control system to accurately point to the sky and make the observation.
The poster will explain the concepts and walk through each of the building block required to put together such system in place using docker and python. The result software stack can be found here: https://github.com/ejoliet/indi-firefly
Tags: #firefly #python #autonomous #telescope #control #pointing #machine-learning #indi