AI-based Sensors for Predictive Maintenance : Pharmaceutical Industry
Laukkanen, Jenni (2025-05-21)
AI-based Sensors for Predictive Maintenance : Pharmaceutical Industry
(21.05.2025)
Julkaisu on tekijänoikeussäännösten alainen. Teosta voi lukea ja tulostaa henkilökohtaista käyttöä varten. Käyttö kaupallisiin tarkoituksiin on kielletty.
avoin
Julkaisun pysyvä osoite on:
https://urn.fi/URN:NBN:fi-fe2025052354070
https://urn.fi/URN:NBN:fi-fe2025052354070
Tiivistelmä
The pharmaceutical industry requires a high level of reliability, compliance and efficiency, which creates increasing demands on maintenance strategies. Predictive maintenance (PdM), which relies on real-time data and advanced analytics to predict equipment failures, offers an alternative to traditional maintenance approaches. This thesis examines how AI-based sensors support the implementation of PdM in the pharmaceutical industry, focusing on both technical performance and industry-specific requirements.
The study presents an overview of various sensor types, including vibration, temperature, ultrasonic and acoustic sensors, examines their working principles, strengths, limitations, and suitability for pharmaceutical applications. Particular attention is given to how sensor data is processed and integrated with machine learning models to enable early fault detection and optimise maintenance planning. Through a literature review, the thesis highlights the importance of data quality, regulatory considerations in deploying effective PdM systems.
The findings show that while no single sensor type is universally sufficient, combining multiple sensors and integrating them with AI-based analytics can significantly improve equipment reliability and reduce unplanned downtime. Moreover, the adoption of PdM supports compliance with Good Manufacturing Practices (GMP) requirements by enabling better traceability and reducing the need for reactive interventions.
The study presents an overview of various sensor types, including vibration, temperature, ultrasonic and acoustic sensors, examines their working principles, strengths, limitations, and suitability for pharmaceutical applications. Particular attention is given to how sensor data is processed and integrated with machine learning models to enable early fault detection and optimise maintenance planning. Through a literature review, the thesis highlights the importance of data quality, regulatory considerations in deploying effective PdM systems.
The findings show that while no single sensor type is universally sufficient, combining multiple sensors and integrating them with AI-based analytics can significantly improve equipment reliability and reduce unplanned downtime. Moreover, the adoption of PdM supports compliance with Good Manufacturing Practices (GMP) requirements by enabling better traceability and reducing the need for reactive interventions.