A process-guided uncertainty-aware deep learning framework for reliable and interpretable industrial fault diagnosis
Babar Hayat, Shabeer Ahmad, Muhammad Asfandyar Shahid, Adil Khan, Md Rajibul Islam, Md Shohel Sayeed, Yasir Ullah · 2026
Read A process-guided uncertainty-aware deep learning framework for reliable and interpretable industrial fault diagnosis by Babar Hayat, Shabeer Ahmad, Muhammad Asfandyar Shahid, Adil Khan, Md Rajibul Islam, Md Shohel Sayeed, Yasir Ullah (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0349385.txt.
Description
Babar Hayat and co-authors built a deep learning framework in 2026 for diagnosing faults in industrial plants. Missed faults threaten safety and quality, yet many systems ignore how a plant is wired and which sensors can be trusted.
Their design reads plant data with layered networks, then aims attention using process structure and a running estimate of each sensor's reliability. A sampling trick attaches honest confidence to each call, and twin outputs handle detection and diagnosis together. On a standard chemical plant benchmark the framework reported strong, interpretable results. Knowing when a diagnosis is unsure is as valuable as the diagnosis itself.
Catalog record
| CREATOR | Babar Hayat, Shabeer Ahmad, Muhammad Asfandyar Shahid, Adil Khan, Md Rajibul Islam, Md Shohel Sayeed, Yasir Ullah |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56044 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |