An Autonomous Adaptive Scheduling Agent for Period Searching

TL;DR

This paper presents an autonomous software agent that adaptively schedules observations of periodic phenomena using robotic telescopes, optimizing data collection in uncertain environments.

Contribution

It introduces a novel adaptive scheduling algorithm that combines geometric sampling with proactive strategies for robotic telescope networks.

Findings

Effective coverage of period ranges achieved

Proactive scheduling improves data quality

Algorithm adapts to changing environments

Abstract

We describe the design and implementation of an autonomous adaptive software agent that addresses the practical problem of observing undersampled, periodic, time-varying phenomena using a network of HTN-compliant robotic telescopes. The algorithm governing the behaviour of the agent uses an optimal geometric sampling technique to cover the period range of interest, but additionally implements proactive behaviour that maximises the optimality of the dataset in the face of an uncertain and changing operating environment.