Data-driven Technologies in Manufacturing Companies

In today’s post, Valerio Alessandroni, Zigurat’s professor of Global MBA in Digital Business and an expert in Industry 4.0 shares with us his vision of how data-driven manufacturing will improve the effectiveness of the processes. 

The way leading to Industry 4.0 always starts from a digitization step. This can be followed by a Business Intelligence step – concentrated on data visualization and analysis – and the exploitation of an Artificial Intelligence engine for predictive maintenance and process optimization. To correctly perform these steps, the manufacturing assets should be smart, networked, and integrated into a single cohesive system.

According to recent estimates, there are currently around 64 million machines used in manufacturing plants in the world, but only 8% of them are efficiently networked. This means that the Overall Equipment Effectiveness (OEE) at the factory level is under 60%, while the costs for unexpected outages and machine recovery are close to 56 M$/year worldwide.

There are currently around 64 million machines used in manufacturing plants in the world, but only 8% of them are efficiently networked.

By correctly managing and using the data collected from the field through acquisition and storage technologies, it is now possible to create a higher added value based on real data.

A ‘data-based’ approach to manufacturing offers several advantages, all connected to better visibility because data-driven manufacturing can ensure a better understanding of performances using the KPIs (Key Performance Indicators) set inside a company. Accurate data can provide information not only about the performance of individual assets but also about manufacturing operations as a whole. This helps decision-makers to focus on production improvements, solving bottlenecks like poorly performing shifts, recurring downtime, or missing materials.

Towards Manufacturing Analytics

Another advantage of the ‘data-driven manufacturing’ approach is possibility to use machine learning algorithms to solve complex problems. Combining manufacturing analytic technologies, artificial intelligence and machine learning, companies can implement advanced data-based decision-making processes such as predictive maintenance.

Also, automation technologies can benefit from a data-driven approach. First of all thanks to the automated collection of data, through the IIoT (Industrial Internet of Things) technologies, and their processing without manual interventions. Second, using the available data to manage automated decision-making processes.

Finally, should be emphasized the reduction in operating costs: in fact, data used in a lean manufacturing environment offer the possibility to simplify production processes and reduce waste at a minimum. Without real-time data, it would be difficult to accurately measure possible production improvements and be sure that the changes resulted in real cost savings.

An integral part of Industry 4.0, Manufacturing Analytics is a new class of software that combines IIoT, Big Data, Machine Learning, Cloud Computing, and Edge Computing technologies. This is helping to create a data-driven approach of a new generation, that allows reduce operating costs through better production planning, and to improve productivity by reducing waste and rework, and to optimize the Supply Chain as a whole.

To obtain the benefits described above, a platform is required for real-time data collection and analysis according to specific KPIs and industry standards, supporting the Manufacturing Analytics functions. The goal is to move from simple data collection and visualization to the ability to leverage real-time data to obtain insights that enable process and asset performance tracking, eliminate bottlenecks, increase returns, enable predictive maintenance, and make possible other vital operations for the company.

KPIs Under Control

In the past, it wasn’t possible to fully exploit all the end-to-end data collected from manufacturing processes, from the supply chain to production and delivery to the customer. In fact, only complex and expensive tools were available, that could collect data only from operators and machines. Days were sometimes required to pinpoint the reason for a production process downtime. But, in this highly competitive world, it is not acceptable to wait days or even weeks to get an answer. More than this, it’s more and more necessary to obtain complete and detailed visibility over the full manufacturing process, from the suppliers to the final customers, to get a 360° vision and really optimize the results. Manufacturing Analytics makes these targets possible, at the same time making more smart and scalable the manufacturing process.

It is clear that to keep under control functions like order management, demand forecasting, inventory optimization, product quality assessment, service, manufacturing efficiency and many others, with the related KPIs, data must be collected in a very distributed and capillary way. For this purpose, fundamental help is offered by IIoT technologies, with their ability to directly monitor machines and processes, providing fresh data that can be analyzed in a KPI perspective.

With the advent of Machine Learning and AI, IIoT allows writing algorithms that identify historical models and, based on these models, to make predictions about the future.

However, the simple acquisition of data through the IIoT is not enough. The real challenge is knowing how to acquire the right data, analyze it in the production process context, and map the information to the KPIs, achieving a higher level of “operational intelligence”.

At the same time, IT processing methods are merging with operational technology (OT) platforms, resulting in next-generation Edge Controllers that offer traditional control along with advanced data management capabilities in a single platform.

In fact, the collected data must be re-formatted in order to easily locate any problems and set up procedures to deal with them in real-time. For example, if too many defective products are leaving a production line, you can quickly identify the problem, perform diagnostics or analysis to pinpoint the problem, and then initiate the required actions to resume production on a regular basis.

Conclusion

Manufacturing Analytics can help the improvement of the quality of the end product through several processes such as data-driven product optimization, defect density levels management, and customer feedback and purchasing trends analysis.

Data-driven optimization can rely on IoT sensors and machine learning models. By analyzing product usage in detail, manufacturers can reduce or increase components that lead to higher usage rates.

Digital transformation incorporating the IIoT holds the promise to deliver extensive and significant data, leading to a better and deeper understanding of how systems are operating and how they can be improved.

Banner MDT

AUTHOR:

Valerio Alessandroni
Professor of the Global MBA in Digital Business
Industry 4.0 Consultant | Senior Consultant at EFCC

Blockchain for RPA systems2021 Digital Marketing Trends