AIoT‑Based Predictive System for Pharmaceutical Manufacturing: The SafeBatch Framework

Yasmina SMIDI; Achraf BENBA; Salma BATAHI; Ouafaa IKHZMANE; Yahia CHERRAH; Faïza MEIOUET; Latifa DOUDACH; Latifa Doudach

Article Details

AIoT‑Based Predictive System for Pharmaceutical Manufacturing: The SafeBatch Framework

Authors:

Yasmina SMIDI | Achraf BENBA | Salma BATAHI | Ouafaa IKHZMANE | Yahia CHERRAH | Faïza MEIOUET | Latifa DOUDACH | Latifa Doudach

Publisher:

The National Center for Scientific and Technical Research, Morocco

Journal:

Journal of Innovation and Digital Health Vol.2 Issue 1, pp.40-46

Document Type:

Article

Publication Date:

10/2025

Document Language:

English

Document ID (Org No.):

2001847361

pISSN:

3009-5417

Keywords:

Abstract:

 Ensuring product quality and compliance in pharmaceutical manufacturing requires more than descriptive monitoring; it demands predictive and proactive approaches. This paper presents the development of SafeBatch, an Artificial Intelligence of Things (AIoT) predictive monitoring framework designed to enhance Manufacturing Execution Systems (MES). Batch record data were cleaned, normalized and engineered, then analysed with Decision Trees, Naïve Bayes, Random Forest, Neural Networks and XGBoost. XGBoost achieved the highest predictive accuracy (96%), with strong precision, recall and F1‑score. The methodology integrates IoT‑enabled MES data acquisition, an ETL/feature store, and ensemble learning validated via k‑fold cross‑validation. Results confirm the superiority of ensemble methods for structured industrial data, and illustrate how AIoT can detect process deviations in real time to support proactive intervention. Perspectives include Explainable AI (e.g., SHAP), stronger cybersecurity (21 CFR Part 11, GDPR) and multimodal extensions. SafeBatch shows a pragmatic path from descriptive MES to predictive, human‑centred ecosystems aligned with Industry 4.0 and digital health.

Article Details

AIoT‑Based Predictive System for Pharmaceutical Manufacturing: The SafeBatch Framework

Authors:

Yasmina SMIDI | Achraf BENBA | Salma BATAHI | Ouafaa IKHZMANE | Yahia CHERRAH | Faïza MEIOUET | Latifa DOUDACH | Latifa Doudach

Publisher:

The National Center for Scientific and Technical Research, Morocco

Journal:

Journal of Innovation and Digital Health Vol.2 Issue 1, pp.40-46

Document Type:

Article

Publication Date:

10/2025

Document Language:

English

Document ID (Org No.):

2001847361

pISSN:

3009-5417

AIoT (Artificial Intelligence of Things) Pharmaceutical

Manufacturing Predictive Analytics

Manufacturing Execution Systems (MES)

XGBoost

Process Deviation Detection

Issues

Article Details

AIoT‑Based Predictive System for Pharmaceutical Manufacturing: The SafeBatch Framework

Authors: Yasmina SMIDI | Achraf BENBA | Salma BATAHI | Ouafaa IKHZMANE | Yahia CHERRAH | Faïza MEIOUET | Latifa DOUDACH | Latifa Doudach

Publisher: The National Center for Scientific and Technical Research, Morocco

Journal: Journal of Innovation and Digital Health Vol.2 Issue 1, pp.40-46

Document Type: Article

Publication Date: 10/2025

Document Language: English

Document ID (Org No.): 2001847361

pISSN: 3009-5417

AIoT (Artificial Intelligence of Things) Pharmaceutical

Manufacturing Predictive Analytics

Manufacturing Execution Systems (MES)

XGBoost

Process Deviation Detection

Issues

Authors:

Yasmina SMIDI | Achraf BENBA | Salma BATAHI | Ouafaa IKHZMANE | Yahia CHERRAH | Faïza MEIOUET | Latifa DOUDACH | Latifa Doudach

Publisher:

The National Center for Scientific and Technical Research, Morocco

Journal:

Journal of Innovation and Digital Health Vol.2 Issue 1, pp.40-46

Document Type:

Article

Publication Date:

10/2025

Document Language:

English

Document ID (Org No.):

2001847361

pISSN:

3009-5417