Automation technology discovered Docker containers some time ago as a useful functional extension for certain products. However, this virtualization technology can be used to tackle even greater technical challenges, such as end-to-end data integration and more efficient application development.

By using a data integration docker, various tasks can be solved in the direct environment of machines and systems. On the one hand, all existing control systems that have suitable interfaces - Modbus, CAN, Profinet, OPC UA, IO-Link - can be coupled with such a docker. Furthermore, additional sensors can also be used to connect assemblies and components without corresponding communication capabilities to a Docker. This data integration means that continuous signals are available in the Docker, which can be permanently evaluated with the help of machine learning and corresponding Docker software. This would allow errors to be detected before they occur and, for example, before a production line breaks down completely. A typical application example would be an AI-based condition monitoring sensor system that recognizes the different machine states using a machine learning model inference and makes them available as categorical variables via an API. This enables higher-level applications, which could also be implemented as Docker, to implement needs-based maintenance for production systems and entire sites.

Relatively platform-independent

Industrial IoT applications are usually distributed across four layers. Docker-based concepts are a good choice for achieving end-to-end data integration in terms of software. Docker containers can be used in a network and can communicate with each other via application-related APIs (Application Programming Interfaces).

© SSV

Docker technology allows platform-independent development of application software under certain conditions. A simple single-board computer from the maker world can now run the same Docker container or Kubernetes cluster that was originally developed for high-end cloud use. This means that data-intensive object recognition with high-resolution image data, for example, which should actually run in the cloud, can be embedded directly in the camera hardware on site. One of the biggest acute problems in automation technology, the lack of software developers, could thus be at least partially solved: an experienced developer from the IT environment can now, for example, create applications that run directly on an edge gateway on a machine.

Security upgrade

Author Klaus-Dieter Walter is a member of the management team at SSV.

© SSV Software Systems

Data integration dockers can also significantly improve IT security in networked automation applications. Many security functions that have long been state of the art in the IT world are relatively rare in the machine and system environment, and in many cases do not even exist at all. These include, for example, end-to-end connections secured by a public key infrastructure (PKI) with digital signatures, two-factor authentication (2FA) and AI-based monitoring systems using complex neural networks that automatically detect abnormal communication behavior. A Docker, as shown in the image, which is operated between a machine controller and the higher-level network structures, can contain both the required cryptographic elementary functions as well as 2FA and a PKI with certificates. This would also enable legally secure pay-per-use applications for billing machine and tool usage fees, for example, which use existing control data as the basis for calculation.