All-round connectivity

Continuous infrastructure © IBM

The most important aspect here is probably the all-encompassing integration within the company, between the companies themselves or with partners and increasingly also with customers. This flexible infrastructure is what makes changing business processes possible in the first place. In a digitalized factory, more and more machines will be connected to each other and to the internet. In addition, many IT and OT systems (ERP, PLM systems, SCADA, controllers and sensors) must be connected via many protocols, even if standardization (OPC-UA, MQTT) is being sought. Examples include the evaluation of sensor data in real time, direct access from MES systems to inventory information in SAP or the use of quality data in the development of new products. This makes efficient, flexible and controllable OT-IT integration based on a service orientation for adaptability, manufacturer independence and exercising data sovereignty for Internet connections all the more important. The Plant Service Bus (PSB) makes this possible.

The PSB supports a range of standard protocols from the IT and OT environment - from OPC-UA, MQTT and REST to databases and adapters to databases as well as standard applications, ERP, MES, EDIFACT and mainframes. For programmable logic controllers (PLCs) that are not yet supported, it is possible to develop a new adapter or use a gateway that converts this protocol, usually fieldbus, into a standard protocol. If required, IBM Streams, a stream computing platform with limited service bus functions, can be used to transport large data streams. Thanks to the use of modern virtualization methods (containers) and comprehensive automation, it is also possible to implement this solution in several plants with limited effort and a uniform appearance.

In particular, increasing communication with external systems in the cloud or with business partners is not only a technical challenge, but also a risk that can be reduced by using the Plant Service Bus.

The right decisions at the right time

A structured approach is necessary for successful digitalization. © IBM

The amount of sensor data and its heterogeneity (structured and unstructured) cause considerable effort in finding the causes of error patterns, anomalies or failures. Advanced analytics and AI techniques are increasingly replacing traditional business intelligence (BI) - not least because of the sheer volume of data, whereby a wide variety of data from previously separate sources has to be evaluated.

Ideally, this data should first be transformed into a form suitable for analysis and, if necessary, pre-processed and cleansed. Depending on the scope of the data or the required response time, the data is not only evaluated in the central company IT system, but also in an edge device or gateway on the machine. In addition to increasing system availability, quality and throughput, cross-company collaboration is another trend that manufacturers are using to find out how their products can be used and optimized. Thanks to data evaluation, costs can be reduced and resources better deployed.

An agile and iterative approach is recommended for the successful implementation of analysis projects, as well as a continuous exchange between the specialist department and analysts and a gradual increase in analytical complexity. Thanks to comparatively simple visualizations - especially on mobile devices - production and system statuses can be called up on the basis of integrated data. Building on the project team's experience, complex correlations, predictions, optimizations and recommendations for action can be developed in further phases. Pre-trained AI solutions can cover common IoT issues and accelerate projects in your own company. These include, for example, predictive maintenance or production optimization. Watson Studio is a tool that supports data scientists in analyzing a wide range of data. Based on open source standards such as Jupyter Notebooks and Spark, it can be used to carry out a wide range of analyses of large volumes of data in teams locally or in the cloud.

Applying and marketing findings

The insights gained can then be used in applications such as IoT and smart data apps, which are often quickly created in the specialist departments as part of a pilot project. The data enables new business models and commercial exploitation. Development for industrial applications often takes place step by step based on user requirements. Not least because other company divisions proceed in a similar way. Later, of course, the question of integration with each other and with corporate IT arises. The challenge is then to remain dynamic and to operate the solutions economically, independently of providers. In many cases, companies want to operate the applications in-house for security reasons, for example, and still benefit from the advantages of the cloud.

Platforms such as IBM SpliceX help to digitize production sustainably. © IBM

Platforms such as IBM SpliceX address precisely this requirement by already including frequently required infrastructure functions. Thanks to the lightweight platform, companies can focus on the actual application and significantly accelerate the implementation of applications. It is important that the project is driven from a business perspective in order to achieve early results with minimal effort and to be able to control development flexibly. This gives users the opportunity to select the microservices that are important to them in a very granular way and independently of platform providers.

A robust architecture is important

Consistent IT infrastructures and standardized interfaces are key elements for the success of Industry 4.0 projects. By networking heterogeneous systems from different vendors through common industry standards, a robust and open architecture can be developed for multiple use cases, including the Digital Twin, that integrates into the existing environment and can adapt to frequently changing requirements. SmartFactoryKL, the world's first multi-vendor smart factory, will be demonstrating these concepts in its sixth edition at Hannover Messe 2019.

Peter Schleinitz, Executive Architect / The Open Group Distinguished IT Architect at IBM / ag

If you would like to find out more about the connection between operation technology and IT in modern production, register for our "OT meets IT" congress.