Maintenance Matters Focus on: Smart Maintenance 16 | Plant & Works Engineering www.pwemag.co.uk February/March 2025 Driving AI accessibility for predictive maintenance The transformative impact of artificial intelligence (AI) on predictive maintenance in industrial automation is becoming increasingly evident. The integration of AI with existing maintenance systems presents both challenges and opportunities for manufacturers seeking to enhance efficiency and reduce downtime. PWE reports. The predictive maintenance market is projected to experience significant growth, increasing from USD 10.6 billion (£7.4 billion) in 2024 to USD 47.8 billion (£33.4 billion) by 2029, at a compound annual growth rate (CAGR) of 35.1%. This expansion is driven by various factors, including the increased adoption of machine learning (ML) and AI, as well as the ongoing need to minimise maintenance costs, equipment failures, and unplanned downtime. Detecting anomalies early The advantages of predictive maintenance are well established, offering benefits such as reduced maintenance costs, improved equipment reliability, increased production, optimised scheduling, and enhanced quality assurance. However, AI-driven predictive maintenance, underpinned by real-time data mining, is enabling manufacturers to realise even greater benefits. By leveraging AI, anomalies in process data can be easily combined with machinery data, allowing for evaluation using analytical models and cloud-based solutions. This facilitates the detection of deviations from normal operational states at an earlier stage than traditional condition monitoring methods, enabling rapid alerts that help to reduce unplanned downtime and improve overall efficiency. Additionally, overall equipment effectiveness is enhanced by minimising unexpected production stoppages. For manufacturers of automation machinery, delivering this new vision for predictive maintenance is a complex challenge, as customers require substantial support to integrate these advanced technologies. Machinery manufacturers must now offer digital analytical solutions alongside traditional mechatronics expertise, necessitating the development of new skillsets. The success of an AI-driven predictive maintenance project relies on three fundamental elements: 1. Accessing consistent and machine-readable data of sufficient quality 2. Data pre-processing 3. Implementing an appropriate machine learning model for the specific application Increasing productivity To address these challenges, Steve Sands, Technical Consultant at Festo GB, explains that Festo has developed a solution that seamlessly integrates predictive maintenance into machinery: the Festo Automation Experience (Festo AX). This user-friendly software utilises AI and ML to extract value from machinegenerated data, enabling informed decisionmaking. The software analyses data in real-time, allowing for quick decision-making based on live information. According to Sands, Festo AX has been designed to enhance productivity, reduce energy costs, prevent quality losses, optimise shop floor operations, and support new business models by analysing and interpreting AI is unlocking the potential for predictive maintenance.
RkJQdWJsaXNoZXIy MjQ0NzM=