Tyler Bikaun University of Western Australia

Tyler Bikaun

PhD Student

Theme 1

Tyler is a mechanical engineer turned PhD candidate at the University of Western Australia and a scholar with the Mineral Research Institute of Western Australia. His research focuses on the intersection of computer science and industrial engineering, particularly applying computational techniques to technical texts in the context of industrial maintenance. Tyler's research interests lie in using deep learning to extract knowledge from technical texts to improve maintenance decision support and strategy optimisation through automated natural language understanding. However, this area is challenging due to poor quality, complex, and scarcely available datasets characteristic of the mining and mineral resources industries.

 To date, Tyler has spent a considerable amount of time engaging with industrial partners of the ARC CTMTDS and academia, where he has:

  • demonstrated the value of technical language processing through semi-automated reliability measure estimation,
  • developed numerous software tools for acquiring high-quality datasets to support deep learning research, and
  • collaborated on software to translate his academic research into usable software for the asset-intensive industries.

 Tyler's research is supervised by Associate Professor Wei Liu, Dr Tim French, Dr Michael Stewart, and Professor Melinda Hodkiewicz.

PHD Research - Technical Language Processing for Industrial Maintenance Records

Tyler's research focuses on deep learning-based knowledge extraction from technical texts authored by humans containing considerable n0ii$3 (noise). Noisy technical texts are ubiquitous in industrial maintenance, capturing critical information that drives decision-making, process optimisation and asset performance understanding. Consequently, the noise in these texts poses a formidable challenge to the current state-of-the-art systems, limiting the ability to extract knowledge to support data-driven decision-making.


Highlights of Tyler's research so far have seen him:

  • Design two peer-reviewed software systems that enable rapid cleaning and knowledge extraction from technical texts,
  • Develop a process for semi-automated reliability measure estimation from maintenance work order records, and

Conceive a procedure for extracting meaning from technical texts to support maintenance strategy optimisation and asset performance understanding.

LinkedIn




There is no content with the specified labels


  • Page:
    Periscope (...) —              

    2020-01-31

    Data-driven adaptive reliability estimation to identify asset faults before you run aground

    Theme 1 Theme 5