This is the second of two blogs about how digital twins and IIoT platforms can be used to create models of shop-floor operations that boost productivity, improve maintenance and speed time-to-market. Part 1 focuses on digital twins and Altran’s involvement in the European Commissions’ Virtual IoT Maintenance System (VIMS) initiative. Part 2 focuses on the selection process for the optimal IIoT platform.
The promise of Industry 4.0 to improve factory performance, improve maintenance, and lower costs is all about connecting the physical and digital worlds enabled by Industrial Internet of Things (IIoT) platforms and digital twins of the manufacturing environment.
IIoT platforms use various sensors, artificial intelligence (AI) and machine learning (ML) to collect and analyze data from industrial equipment. Digital twins are virtual representations of systems, such as product design or business processes, a piece of manufacturing equipment, a production line or an entire factory. Together, they create systems that improve factory performance, optimize maintenance and repairs and create new revenue opportunities from innovative digital services.
To remain competitive in the coming years, manufacturers must invest in IIoT sensors and platforms and digital-twin technology that will enable them to acquire and analyze the massive volume of data to better monitor, maintain and innovate factory operations.
Altran is in the vanguard of this effort, spearheading the Virtual IoT Maintenance System (VIMS) project, a European research program funded by the European Commission’s Community Research and Development Information Service. The VIMS project integrates IIoT platforms with digital twins to develop new digital ecosystems for industrial and manufacturing maintenance systems. Altran is working with VIMS industry participants Airbus Operations SL and F. Hoffman-La Roche AG to develop IIoT-digital twin solutions for their operations.
Six criteria for selecting an IIoT platform
Choosing the right IIoT platform is critical for maximizing the value of the factory’s digital twins. There are a number of platforms to choose from; some are more comprehensive and mature than others or more appropriate for specific use cases than others. (See Figure 1.) Here are six criteria that will help you select the most suitable platform for the application:
- Connectivity: Key features include the platform’s versatility and how well it integrates assets and data sources. The more pre-built connectors (i.e., software communication interfaces with external agents) the platform has, the easier it is to implement new connectors, and the more likely they will be integrated seamlessly. Supporting open and standard protocols and application programming interfaces (APIs) such as object linking and embedding for the OPC unified architecture (OPC UA), message queuing telemetry transport (MQTT), RESTful and WebSockets is a base requirement. Also important is how well the IIoT platform can integrate third-party services with the ones provided by other manufacturers.
- Analytics and Visualization: Data is not much use if it doesn’t provide insight. Key analytics features include the ability to develop ML algorithms and implement AI techniques, such as convolutional neural networks and other data-analysis methods. It is also essential to display all the data and insights generated in a human-readable format.
- Security: Production information is highly valuable and must be protected to prevent theft and ensure the flow of data is uninterrupted. Critical security issues to consider include automatic secure edge device provisioning and monitoring, end-to-end data-flow encryption, protected access to sensitive information and role privileges, to name a few.
- High Availability and Hosting: The VIMS platform needs to be reliable and highly available, which requires back-up servers to keep the application always running even if a failure occurs in the primary servers. Other options to ensure high availability include multi-zone hosting with servers physically located in different regions of the world or auto-load balancing to divide traffic demands among servers.
- Support: Direct support from equipment and tool manufacturers, as well as documentation resources, training and certification, are all critical requirements.
- Scalability: A solution built on an IIoT platform needs to scale quickly, which requires investment in additional storage and compute resources as the applications of the solution grow. Also, the platform should be able to ingest data at variable rates, depending on the operational status, from a few samples per minute to thousands per second.
Figure 1: Leading IIoT platforms for digital twins
Source: Compiled by Altran
Selecting the Altran VIMS IIoT platform
VIMS is an ambitious project that will bring digital twins closer to our corporate partners’ production lines. It opens possibilities that are only just now starting to emerge. Digital twin technology is poised to be a must-have asset for businesses that are on the Industry 4.0 journey and want to digitalize, optimize and manage their factories in a smarter, more efficient way.
In our research to select the best IIoT platform fit for the Altran VIMS digital twins’ specific requirements for Airbus and Roche, we evaluated all seven platform vendors based on how well they satisfied each of the six criteria listed above. After much deliberation, we concluded that the best option for the VIMS project was Microsoft Azure.
Azure offers a modular IIoT platform that delivers the essential services and functionalities without having to buy features that are not needed. The Azure modules provide developers with options spanning every part of the data treatment pipeline: data ingestion, storage, analysis, communication and representation. In many cases, the development of solutions is as easy as enabling components to connect functionalities, which minimizes the need to write code for common tasks.
Azure allows the connection and management of millions of assets thanks to modules such as the Azure IoT Hub. It provides many storage options thanks to a wide range of available databases, such as Azure Cosmos DB, Data Lake and Blob Storage. Azure also has deep analytics capabilities with modules such as Azure Machine Learning and provides on-premise functionalities with Azure IoT Edge. Visualization of information is possible with Power BI, and security features are integrated into every module to ensure communications are safe and data is protected. Of particular importance for VIMS is the recent release of Azure Digital Twins, an open, standards-based platform for representing connected processes within IIoT solutions.
Altran is looking forward to expanding VIMS to include more use cases, which starts with a core of technologies that can adapt to the client’s specifications and business needs. As VIMS evolves, Azure offers advantages. For example, Altran only needs to invest in the modules that VIMS needs. Also, Azure can be deployed as Infrastructure as a Service, Platform as a Service, Hardware as a Service and even Solution as a Service, providing a flexible menu of opportunities.
As Altran develops more complex digital-twin systems, there will be challenges that need to be resolved. For that, Azure has a wide range of documentation, examples and support features to help developers and users find the right solutions.