Introducing augmented reality can increase the efficiency of business processes and enable new services and business models. Yet, real-time data analysis and graphical processing places increasing demands on IT. The pump manufacturer Flowserve managed to solve this through a decentralised IT architecture bringing computing closer to the edge, and this is expected to be a prerequisite for future digital production processes.
By Wong Soon Yee, R&D Manager of Global IoT Innovation Lab – APAC, Hewlett Packard Enterprise Singapore.
When looking to digitise manufacturing processes, companies today are focusing increasingly on human-machine interactions. This includes the use of augmented reality to enhance physical and sensory experiences through digital graphics and computer-generated simulations.
Forming the basis for augmented reality applications is the ability to create a “digital twin” of the physical product or machine. This is the virtual image created during the product ideation and development stages which will remain connected to its physical counterpart throughout its entire lifecycle.
The Role Of Digital Twins In Manufacturing
Digital twins serve as a visualisation of a product’s current state or the graphical simulation of possible states in augmented reality form. When applied in engineering, augmented reality offers the possibility to display different shapes, designs, colours and other variable properties, without having to create an entirely new model. By digitally visualising ongoing product developments, digital twins provide engineers with a tool that allows for necessary changes to be made quickly and easily.
Another use of industrial augmented reality applications is for quality assurance. A major challenge with assuring quality in product manufacturing is in finding the right combination of accuracy and rationalisation to keep costs at a minimum. Augmented reality systems allow users to interact with quality assurance systems by graphically displaying critical measurements or instructions directly over the actual object. This direct and visual interaction enables active fault management and allows employees to more effectively discover existing errors and avoid mistakes. This in turn also enables efficient recognition, acquisition and analysis of quality-related data.
With augmented reality, users can also project production instructions as a 3D representation or animation on machines and products. Additionally, the production line is able to engage virtually with experts remotely and receive any feedback in real-time. This ultimately accelerates workflows with reduced error and complements hands-on training.
Leveraging Edge Computing To Drive Business Efficiencies And New Service Models
The potential of augmented reality is not limited solely to product development and production, but can also help maintenance technicians in the field of troubleshooting and maintenance work.
A company that has successfully introduced augmented reality is Flowserve, one of the world’s largest manufacturers of industrial pumps, valves, seals and services for power generators and the oil and gas industry. To date, Flowserve has sold approximately three million pumps, but found that service and maintenance of those pumps were often carried out by third parties who may not be utilising the most efficient maintenance processes. Flowserve realised it needed to improve aftermarket services for its customers and keep it cost-efficient, while avoiding a direct price competition. In addition to product sales, the pump will be offered as a service (flow as service) in the future.
The company then embarked on offering their product offering as-a-Service, and together with National Instruments (sensors), PTC (software), and Hewlett Packard Enterprise (IT infrastructure and services), Flowserve developed an augmented reality solution to collect and evaluate all relevant data for billing, monitoring, predictive maintenance and support.
Through a digital dashboard, the production manager can now view the state of all pumps, their current efficiency, and time to the next maintenance cycle using predefined criteria. Any changes in these values – which are recorded and evaluated in real-time – are reflected directly as an adjustment in the dashboard, so that the production manager can extend the maintenance period by reducing the load or revising the change in time directly into the production schedule until the next maintenance.
The same information is made available to the onsite maintenance technician, and further supplemented by detailed sensor data that is presented in the form of an augmented reality solution with maintenance instructions. For specific maintenance tasks, the technician can view individual work steps using a 3D animation. And if extensive support from experts is required, the technician can access this remotely.
As each Flowserve product individually customized and produced, all information and visualisations are uniquely based on the digital twin of the respective pump. As a result, this solution must be installed directly at the pump and will need to run on a powerful IT platform.
The reason for this is simple: the calculations required for predictive maintenance, the augmented reality visual representations and the dashboard are highly sophisticated, and need to process very large amounts of data. Transferring this data to be processed in a remote data centre would not only cause too much latency, but would also be expensive and prone to transmission errors. Instead, only the status information is transmitted for processing in the central data centre, and all other sensor data is processed near the site with the help of decentralised converged edge systems that provide data centre-level computing power and control of devices. This architecture ensures that the pumps can be monitored autonomously, allowing onsite technicians to access instant insights and take immediate action – and without dependency on an Internet connection.
Future Outlook On Manufacturing: Edge Computing As A Prerequisite For Digital Production Processes
For such strategies to be implemented more broadly in manufacturing processes, intelligent systems that enable convergence between the digital and physical worlds will be increasingly required. The resulting data volume and the computing power required for data analysis at the edge will place high demands on these systems. Market analysts predict that within the next few years, half of the collected data will be analysed at the data source right at the edge, and the findings will be translated directly into production instructions.
This means that these intelligent systems must not only be suitable for use in an industrial environment, but a new class of “edge computing” systems and solutions offering sufficient performance, scalability and availability will be required. Such a decentralised IT environment will be the prerequisite in the digitisation of manufacturing production processes, and edge computing will be involved in multi-tiered and hybrid IT architectures, including enterprise data centres and cloud platforms.
READ MORE ISSUES AND INSIGHTS
WANT MORE INSIDER NEWS? SUBSCRIBE TO IAA NOW!
CHECK OUT IAA’S CURRENT AND PAST ISSUES: DIGITAL MAGAZINE