EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Leakage-Controlled Evaluation of CNN-LSTM Models for Process Temperature Forecasting and Soft Sensing in Machining: Evidence from the AI4I 2020 Benchmark

Ayoub Alsarhan, Mamoon Obiedat, Malek Barhoush, Tamam Alsarhan, Sami Aziz Alshammari, Nayef H. Alshammari, Khalid Hamad Alnafisah, Mohammed Aleinzi

Machines · 18.11.2026 · 14 (11) · 1375

DOI: 10.3390/machines14111375 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Machines, online erschienen am 18.11.2026. Verfasst von Ayoub Alsarhan und 7 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Northern Border University.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Convolutional recurrent networks are widely reported as being able to forecast process temperature in machining accurately. Many such evaluations, however, split overlapping sliding windows at random, which lets near-duplicate samples cross the train–test boundary. We re-examine process temperature forecasting and soft sensing on the AI4I 2020 benchmark under a leakage-controlled protocol: a chronological split with a purge gap, training-only scaling, five seeds, block-bootstrap confidence intervals and Diebold–Mariano tests with an autocorrelation-robust variance. LSTM, GRU, CNN-LSTM and an attention-augmented CNN-LSTM are compared with persistence, linear, random forest and multilayer-perceptron baselines that receive the same ten-record window. Under a random split the attention model reaches R2=0.899±0.006; under the chronological split it falls to 0.305±0.089. Rolling-origin cross-validation, a purge-gap sweep, a factorial ablation and sensitivity analyses support this picture. In soft sensing, the error is dominated by a drifting process–air temperature offset that the machine signals in this benchmark do not observe, and refitting on recent data matters more than architecture. When lagged process temperature is available, persistence (MAE 0.054 K) is not beaten at horizons of 1–30 records, and convolutional and attention stages add no benefit while costing up to 2.4 times the training time of a plain LSTM. On this benchmark, the evaluation protocol changed the conclusions more than did the choice of architecture. Claimed gains of deep sequence models for machine monitoring should therefore be reported against persistence and information-matched baselines under a time-ordered split.

Förderung

  • Northern Border University

Zitieren

APA

Alsarhan, A., Obiedat, M., Barhoush, M., Alsarhan, T., Alshammari, S. A., Alshammari, N. H., Alnafisah, K. H., & Aleinzi, M. (2026). Leakage-Controlled Evaluation of CNN-LSTM Models for Process Temperature Forecasting and Soft Sensing in Machining: Evidence from the AI4I 2020 Benchmark. Machines, 14(11), 1375. https://doi.org/10.3390/machines14111375

Vancouver

Alsarhan A, Obiedat M, Barhoush M, Alsarhan T, Alshammari SA, Alshammari NH, et al. Leakage-Controlled Evaluation of CNN-LSTM Models for Process Temperature Forecasting and Soft Sensing in Machining: Evidence from the AI4I 2020 Benchmark. Machines. 2026;14(11):1375. doi:10.3390/machines14111375

BibTeX
@article{Alsarhan2026,
  author = {Alsarhan, Ayoub and Obiedat, Mamoon and Barhoush, Malek and Alsarhan, Tamam and Alshammari, Sami Aziz and Alshammari, Nayef H. and Alnafisah, Khalid Hamad and Aleinzi, Mohammed},
  title = {Leakage-Controlled Evaluation of CNN-LSTM Models for Process Temperature Forecasting and Soft Sensing in Machining: Evidence from the AI4I 2020 Benchmark},
  journal = {Machines},
  year = {2026},
  volume = {14},
  number = {11},
  pages = {1375},
  doi = {10.3390/machines14111375}
}
Kostenlos starten →