Open source is widely used in IT and makes it possible to react quickly and agilely to changing market requirements.
Traditional automation and control technology only offers this openness to a very limited extent, as it is generally based on proprietary technology. In view of ever shorter production cycles, integration in and networking between companies and the ability to update, these isolated solutions are reaching their limits.

Classic versus open systems

Traditional control technology requires experts for proprietary systems. Modern automation and control technology based on open source is accessible to more developers and enables reusability; existing software can be used in drives, embedded controllers and IPCs. This results in lower R&D costs and faster responsiveness.

The integration of information technology (IT) makes control and automation technology more flexible.
In the course of the Industrial Internet of Things, OT (Operational Technology) and IT are increasingly merging. Data is being made available to and analyzed by a wide variety of systems from the field level to the cloud. IT is finding its way into automation and device control: it enables signals and inputs from various sources to be aggregated and fed into automation algorithms. Remote control of production via IT technology helps to react more quickly to malfunctions and downtimes, thus reducing downtimes.

With classic control technology, the control logic is closely linked to the underlying device.
underlying device. Apps, on the other hand, decouple functionality and hardware, thus separating the control logic from the devices. The advantage of this is that the control logic can be easily transferred to other devices. In addition, control applications can be reused and improved more easily and optimized more quickly thanks to the wider range of applications. With the applications, the operating technology becomes more software-defined, while the hardware becomes a commodity.

New data processing functions

Edge and cloud computing bring new data processing and storage functions to control technology. For example, there are control applications that make use of deep learning. Such applications process large amounts of sensor data in real time for predictive maintenance or quality assurance. One application of edge computing is computer vision. Intelligent computer vision devices are used at the edge to detect anomalies on production lines.

These devices feed real-time metadata into industrial control loops that make automated decisions if a fault is detected.
The main advantage of control technologies that use the cloud is data analysis. The data generated on the machine can be processed in the cloud and fed into ERP and MES applications. This creates more business intelligence and increases operational efficiency.

Programming open control technology

A loosely coupled operating system architecture based on snap containers.

© Canonical

The core of the open automation and control solution is the Linux distribution Ubuntu Core, which brings cloud-native technologies to machine control. Cloud native refers to an approach that aims to develop software more agilely and quickly and to build it more robustly. The application is broken down into small, largely isolated components and designed in such a way that all of these components can be developed independently of each other. As a rule, the software no longer runs on-premise, but is designed to run directly from a cloud platform. Companies can work more flexibly overall and provide their users with updates, corrections and new functions more quickly.

With Ubuntu Core, so-called snap containers form these components. Containerization divides control software into individual modules, making it cheaper to develop, easy to reuse and easy to maintain in the long term.

The system also contains a variety of security mechanisms to support mission-critical control devices in control technology. It works with strict application restrictions: applications work as self-contained systems in their snaps. The software contained in snaps is unchangeable, which reduces the risk of safety hazards. Machine manufacturers using these applications automatically receive updates for security fixes and patches for the life of the device without the need for engineers to intervene.

System daemon ensures security

Galem Kayo is a product manager at Canonical and is responsible for the product strategy for Snaps, Ubuntu Core and Anbox Cloud.

© Canonical

The core of Ubuntu Core is a system daemon called Snapd. Snapd is the
central nervous system in the controller, which ensures the exchange of real-time and non-real-time data for all apps. Snap containers are the smallest units of systems. Each individual component of the operating system is packaged in a separate snap container (kernel, boot loader, root file system and applications), which in turn has narrowly defined tasks. Snapd manages access permissions to the system to prevent unauthorized actions and takes care of communication between containers.

The operating system provides a REST API via which systems can communicate with the cloud, for example for software updates or device management. Snapd automatically fetches software updates to perform the smallest and planned system updates. With Ubuntu Core and Snaps, software updates can be rolled back to the last known working version to increase uptime and reliability in the factory.

Applications packaged in Snaps are fully constrained with all their dependencies as everything they need is in a container. This makes it easy to run any application on an Ubuntu core system - regardless of the programming language. Developers have programming languages of their choice, including C, C++, Python, Javascript or Go. Traditionally, in an industrial environment, they were limited to specialized programming languages such as IEC 61131 or G-code.

Software-defined future

The future of industrial control is software-defined. Hardware will become more generic and software will play a more important role. Due to latency requirements and the distributed nature of industrial operations, edge computing will be more practical than cloud computing. Industrial companies will gain productivity from the IoT, which will reduce uncertainty through more real-time visibility and greater insight into operations.

Artificial intelligence will increasingly play a decisive role. The keywords here are 'predictive maintenance' and 'predictive quality', i.e. predictive maintenance and quality assurance. These goals, which are much discussed in the industry and have already been implemented in some cases, are to be achieved by linking IT with automation and control technology: sensors provide data on the status of machines and devices, which is processed in the cloud or at the edge. From this, software developers and engineers can in turn develop AI-based models that help to make production lines more efficient in the long term.

Ubuntu Core, developed for embedded devices, and Snaps, the universal Linux application containers, form an open source platform that removes the traditional boundaries between machine control, IT and the Internet of Things. Thanks to the open architecture, machine manufacturers are no longer tied to PLC specialists and proprietary systems, as the software is decoupled from the hardware. Developers, in turn, can use a modern CI/CD and DevSecOps approach to deploy applications on edge devices in a traditional control engineering environment.

Ubuntu Core unites the worlds of automation and the Internet of Things (IoT) in an open, modular and secure way for future-proof and innovative automation in mechanical and plant engineering. This creates the basic prerequisite for successful Industry 4.0 or Factory of the Future projects.