Abstract
Understanding and quantify human performance is an essential component to guarantee and control the safety of critical installations where human intervention can represent the ultimate safety defence. Human reliability analysis is a time consuming and tedious task usually performed by a human factor expert and therefore subjected to error and variability. In addition, within human reliability analysis there are numerous opportunities to learn from data. However, how data are gathered, presented, shared, and used is an area of continuous development and discussion. In this work, we present a collection of artificial intelligence (AI) tools and methodologies developed to tackle different challenges within the field of human reliability. The aim is to automatise the process, learn from data and support the task of human reliability experts. The collection of tools includes: a tool to automatically classify human errors from accident reports and construct a Bayesian/Credal Networks. The developed works are freely available as part of the open source COSSAN software.
| Original language | English |
|---|---|
| Title of host publication | UNCECOMP 2023: 5th International Conference on Uncertainty Quantification in Computational Science and Engineering |
| Subtitle of host publication | Proceedings |
| Editors | M. Papadrakakis, V. Papadopoulos, G. Stefanou |
| Place of Publication | Athens |
| Pages | 437-452 |
| Number of pages | 16 |
| ISBN (Electronic) | 9786185827021 |
| Publication status | Published - 24 Oct 2023 |
| Event | 5th ECCOMAS Thematic Conference on Uncertainty Quantification in Computational Sciences and Engineering, UNCECOMP 2023 - Athens, Greece Duration: 12 Jun 2023 → 14 Jun 2023 |
Conference
| Conference | 5th ECCOMAS Thematic Conference on Uncertainty Quantification in Computational Sciences and Engineering, UNCECOMP 2023 |
|---|---|
| Country/Territory | Greece |
| City | Athens |
| Period | 12/06/23 → 14/06/23 |
Funding
This work was partially supported by the EPSRC grant EP/T517938/1.
Keywords
- Bayesian networks
- human error
- human reliability analysis (HRA)
- machine learning
- natural language processing
- software