Home / News / The Road to Industry 4.0 Series: (Part 3) From Shop Floor to Products

The Road to Industry 4.0 Series: (Part 3) From Shop Floor to Products

Jun 05, 2023Jun 05, 2023

In Part 3 of this series, we explore the topic of ‘Digital Twin’ and it’s role in manufacturing. We explore how this revolutionary technology is transforming product development, shop floor optimisation, predictive maintenance, and supply chain efficiency. Building on the insights shared in the second article, we delve deeper into the application and integration of Digital Twin across the manufacturing value chain.

Manufacturing organisations have been utilising a host of digital tools and enablers to improve efficiency and effectiveness in plant operations to great effect. Today, they are turning this up a notch by leveraging Digital Twin technology in their manufacturing operations.

Many players have used core systems such as shop floor simulation, Advanced Production Scheduling (APS), Manufacturing Execution Systems (MESs) and Condition Monitoring Technologies (CMTs). When well implemented, these tools enable greater efficiency and are now essential in the sustainable, safe, and productive operation of a plant.

Now, to achieve the next level of performance in manufacturing, organisations are looking to build on this technology base to gain advantage by applying new and richer data sets. The integration of these core tools with a new wave of digital solutions including 3D modelling, Production Flow Analysis, and Machine Condition is enabling greater digital insight on the road to manufacturing efficiency.

This manufacturing digital capability is now enabling the adoption of the Digital Twin concept, which plays out in an industrial context on both the shop floor and in the final product. While commonly used to create a digital representation of a product, this chapter in the series will examine how it can be applied more broadly in the manufacturing arena.

Digital Twin is a virtual replica of a system, process or product that is used to simulate behaviour and performance under different conditions as well as monitor live events. In manufacturing this can be for specific production lines, or any real-world scenario within a production process.

The Digital Twin generally merges the visual with numerical results or verbal work instructions, and is typically made up of three main components:

With these three components in place, a Digital Twin can replicate, analyse and optimise the performance of its physical counterpart. This can be applied to streamline the production, maintenance, in-life design change, and stress testing processes, all taking place virtually. It allows for zero downtime impact, and for the fine tuning of changes to be tested and modelled before implementing smoothly and efficiently in the real environment.

Digital Twins can be applied to physical production, business process and decision-making activity across the manufacturing value chain. Two areas where current state applications and integration enable most predominately are:

When teams want to experiment with a change, lengthy trial and error process to test the manufacturing of a new or updated product often await.

With Digital Twins, manufacturers can test out updated production configurations virtually while minimising the risk of costly oversights. The digital simulation of various scenarios is far quicker and less resource-consuming than physical testing.

By altering the setpoints of the production process and predicting what the outcome may be, teams can identify potential problems and bottlenecks in the system. This generates less waste and can save a great deal of money and time, especially in factories with complex processes and machinery.

This is often done through 3D Emulation software integrated through an IoT platform with CAD, where routing and operational work instruction information can be used as the basis for commissioning teams to simulate scenarios for production line flow setup. This optimises line flow and helps set the visual design of the line for easier installation.

Digital Twin technology can also carry out the simulation of hazardous and dangerous scenarios and identify potential safety risks stemming from these, which enhances safety measures and can minimise the risk of accidents.

It is now fairly common practice for manufacturers to collect an array of machine condition information points about their production machinery, such as vibration, temperature and motion. In these applications, the data comes through in different forms and formats – vibration is typically in wave form while temperature might be captured in a thermographic image.

With Digital Twin, this information can be incorporated into a one-point solution which processes each input and automatically initiates maintenance planning work based on condition.

If machine learning (ML) is applied, the system can predict the potential failures and suggest timing horizons for any work to be carried out. This enables the maintenance and production planning teams to identify the issue before it halts production or develops into a hazard.

One application of this comes in the steel industry, where the use of thermal imaging to monitor the condition of refractory brick linings in torpedo ladle cars conveying molten iron to the steelmaking section. By applying computer vision technology to the thermal imaging, predictions can be made on the deterioration of brick linings; integrating this with the CMMS to enable automated launch of a maintenance notification then prompts the scheduling and planning of a proactive maintenance event.

As the Digital Twin continues to be used, any problems that repeat across one or more components produce detailed patterns, which are fed back into the Digital Twin for it to accurately predict when and where the next failure will occur.

It also can be used to optimise supply chain processes, such as tracking inventory levels and predicting demand. This stretches to monitoring production processes for quality control and detecting defects in real-time, allowing companies to quickly identify and correct issues.

The ability for a Digital Twin to predict and prevent system failures is high, which can lead to reduced downtime and increased availability. By monitoring performance in real-time, they detect anomalies and trigger maintenance before they become critical.

As energy prices and the drive to move away from fossil fuels sharpens the focus on renewable energy usage, wind power will play an increasingly prominent role. To drive efficiency in power generation while reducing costs, wind turbine manufacturers and operators are leveraging Digital Twin-based simulation in their both their product development and operation.

Simulation allows engineers to rapidly prototype and test designs in a virtual environment using real world data, which in turn reduces cost that would have been incurred on multiple physical design iterations and allows engineers to create radical designs that not only solve the issues with deploying turbines in the most challenging conditions but can also be manufacture more quickly and at a lower cost.

Taking the use of Digital Twin one stage further, the turbines themselves are designed to incorporate a multitude of sensors that allow the use of real time analytics to monitor multiple conditions, and companies operating the wind turbines can adjust in a further simulation to improve the output of the system before applying it to the turbine itself. These sensors together with ML-driven predictive analytics also allow for prediction of both failure and maintenance requirements, maximising the availability and output of the turbine.

When building a business case for Digital Twin, A&M knows the importance of considering the volumes and potential upside through digital techniques, and considering cost of machinery, asset utilisation and downtime. If this tallies, for a single farm of 100 turbines this could save around US$2-4 million in deployment cost and produce an additional US$1 million of electricity per year.

Sustainably applying Digital Twin solutions is not a trivial task and with its benefits comes the usual transformational groundwork in any performance improvement task. Some lessons learned from successful implementations are:

Have a compelling value case: The solution should give line of sight into an EDITDA impact material enough that it becomes an important improvement pillar for leadership. We normally see a high likelihood of scaling of digital investments with an EBITDA to investment ratio of two to five times.

Data quality and process performance are linked: By definition and design, Digital Twin is a highly integrated solution. Data quality is paramount for the integration and the quality must be consistent over time. Organisations with high levels of process performance and low variability have the highest likelihood of success with Digital Twins – if this area is a concern, use the opportunity to get things in shape.

Develop an MVP then scale with business value in mind: Establish a Minimum Viable Product (MVP) to run in one plant overseen by leadership that believe in the investment case and commit to scaling. Baseline performance and measure results, and should the MVP deliver on the vision in terms of features and performance improvement, the plant leadership will recommend the solution for the enterprise. This establishes a firm footing for scaling, which should be carried out with value being front of mind.

This concludes the third part of our series on digital techniques that help companies reduce costs and improve efficiency in industrial environments. To learn more about this topic, you can read Part 1: Tech and digital techniques in industrial settings and Part 2: Three strategies to help manufacturers acquire and exploit shop floor data. In our upcoming articles, we will explore intelligent automation and how to mitigate cyber risks. By effectively using these digital advancements, businesses can achieve substantial improvements and stay ahead in the ever-changing industrial landscape.

Digital definition of the physical environmentOperational data from the physical environmentInformation simulation and presentationHave a compelling value caseData quality and process performance are linkedDevelop an MVP then scale with business value in mind