How AI/ML Can Improve Manufacturing Operations

摘要： Stuart Gillen, Senior Manager at Kalypso, offers a few ways manufacturing organizations can leverage predictive maintenance to identify potential issues, reduce the occurrence and length of unplanned downtime, and get the most value from assets and budgets.

▲圖片標題(來源：USM Business Systems)

As manufacturers become increasingly connected, their systems, machines, sensors and other devices are generating a wealth of new data, and given the sheer volume of data generated, that isn’t easily analyzed. It is a challenge that traditional manufacturing systems are not designed for – and manufacturers are missing out on valuable insights as a result.

Machine learning (ML) and Artificial Intelligence (AI) technology can help, when implemented in support of an IoT strategy and validated through a strategic experiment that proves the potential value. Manufacturers should take a comprehensive approach to machine learning and analytics, integrating equipment, systems and people into a highly collaborative environment that rapidly adapts to changing operational requirements and operates on a scale much larger than simple IoT applications.

Here are a few ways manufacturing organizations can leverage predictive maintenance to identify potential issues, reduce the occurrence and length of unplanned downtime, and get the most value from assets and budgets.

Integrate with IIoT platforms to monitor machine health and performance

Enterprises can integrate predictive maintenance models into their manufacturing systems to actively monitor asset health and send alerts at optimal maintenance periods. For example, a worker installs sensors on machines and connects them to an IIoT platform. The sensors send machinery health data to the IIoT platform in real time and observe patterns of operation. The IIoT platform remotely monitors the health of the machinery – monitoring for anomalies or deviations. When conditions exceed machine learned thresholds, plant personnel are notified automatically through email/SMS. This allows organizations to react quickly to otherwise unknown events thus improving overall operations. And by understanding the health of the machines, asset owners can act on issues before they become critical.

Use ML to optimize production runs based on product, operator, and environmental conditions

Often referred to as “golden runs,” personnel can use ML techniques to evaluate hundreds or thousands of individual product runs to identify the optimal process parameter settings capable of producing the maximum throughput. This gives operators the ideal settings based on current conditions to maximize yield. Then going one step further, AI and model predictive control techniques can be implemented to automatically set the appropriate machine parameters allowing operators to focus on more pressing needs to keep a manufacturing line running optimally.

Unite additional plant systems to achieve an end-to-end solution

End-to-end automation provides an overall increase in labor productivity and helps plants operate at their optimal maintenance cost. For example, the predictive models integrated with Computerized Maintenance Management Systems (CMMS) can trigger automated work orders based on production schedules, resource availability and machine health conditions – a true end-to-end solution. Plant management derives value through production planning, asset lifecycle costing, improved throughput and resource allocation optimizations.

In summary, companies that implement ML capabilities into their digital transformation strategies can minimize downtime and production losses while improving the quality of goods. By automating important, yet labor intensive tasks like scheduling work orders, forecasting, and ordering new parts, manufacturers achieve greater efficiency and higher output by reducing human error.

轉貼自： Inside Big Data

若喜歡本文，請關注我們的臉書 Please Like our Facebook Page： Big Data In Finance