Dr. Mario Andrés Muñoz

Dr. Mario Andrés Muñoz

Senior Research Fellow at the School of Computing and Information Systems, The University of Melbourne. Black-box optimisation, Instance Space Analysis, Algorithm Selection.

Industry Collaborations

Published:

A key component of my research is translating methodology into real-world impact through industry partnerships. The following collaborations have been conducted through OPTIMA and the ARC Laureate program.

AGL Energy (2024–present)

Development of an Anomaly Detection Model for Predictive Maintenance on Wind Turbines. This collaboration was established through OPTIMA and has resulted in a fully funded ARC industry PhD scholarship (student: Lantao Zhang). A parallel Master of Data Science capstone project focused on predictive maintenance methods is also underway.

Boeing Australia (2021–2024)

Development of multi-fidelity surrogate modelling methodologies to improve the modelling of aircraft flight envelopes. This work combined bi-fidelity information to accurately characterise complex aerospace systems with reduced computational cost.

Related publications:

  • N. Andrés-Thio, M.A. Muñoz and K. Smith-Miles (2024) Characterising harmful data sources when constructing multi-fidelity surrogate models. Artificial Intelligence, 336:104207
  • N. Andrés-Thio, M.A. Muñoz and K. Smith-Miles (2022) Bi-fidelity Surrogate Modelling: Showcasing the need for new test instances. INFORMS Journal on Computing, 34(6)3007–3022 (Featured Article)

Future Fibre Technologies Ltd. (2016–2020)

Development of anomaly detection methods for streaming multi-dimensional data applied to fibre optic sensing systems used for perimeter security and intrusion detection.

Related publications:

  • S. Kandanaarachchi, M.A. Muñoz, R.J. Hyndman and K. Smith-Miles (2020) On normalization and algorithm selection for unsupervised outlier detection. Data Mining and Knowledge Discovery, 34:309–354
  • P.D. Talagala et al. (2020) Anomaly detection in streaming nonstationary temporal data. Journal of Computational and Graphical Statistics, 29(1)13–27

Powerline Bushfire Safety Program (2017)

Developed algorithms to identify plant species likely to cause faults when contacting 22 kV power lines, contributing to bushfire prevention.

Related publication:

  • S. Kandanaarachchi, N. Anantharama and M.A. Muñoz (2021) Early detection of vegetation ignition due to powerline faults. IEEE Transactions on Power Delivery, 36(3)1324–1334