A process of rethinking is gradually taking place in the automation environment: The daily practice of the IT world is also finding its way into the software programming of automation devices. In order to secure, optimize and make developments more flexible, the focus is shifting to version management. Data backup and data management are often addressed at the same time.

The sticking point is changes to control devices such as PLCs, which can quickly lead to system downtime or faulty production if programming errors occur. The need to be able to monitor, track, compare and secure changes to PLC programs (EN 61131) is therefore very real. However, production managers and employees are often critical of the introduction of a version management or data management system. The opinion persists that manual administration without software support is sufficient. At first glance, this works well for years, with data being backed up and managed using various tools.

This ignores the fact that the number of users and devices, and therefore data and sources of error, increases steadily over time. However, with the first major incident, which causes unnecessary additional costs and a considerable amount of time, it should become clear that purely manual version management is too short-sighted. But what are the limits of manual version management and what does a version management and data management system actually look like given the current state of the art? What needs to be considered and what inhibitions need to be overcome when it comes to implementing such a system, or are the concerns justified at all? Automation specialists are often still unsure about such questions. The seven biggest inhibitions ('myths') regarding version and data management are addressed and explained below.

Myth 1: A version management system is unnecessary - our system runs without one!

Without modern version management and scheduled or defined upload, download and comparison processes, production managers can never be sure that the software version that controls the system corresponds to the version that was last released.

This is because without a comparison of the online (system) and offline (server) status and a detailed (graphical) comparison of different versions, the system is operated virtually blind.

Current version management systems therefore offer a watertight backup strategy. And even across plants. This is because backed-up data from distributed systems or locations can also be synchronized via the central memory and version differences can be detected. The basis for this is a server-client architecture in which the data is stored centrally on the server and can be accessed, managed and changed via any number of installed clients. In this way, several locations can be coordinated centrally by one main person responsible and program changes can be tracked.

Myth 2: Implementing a version management system is risky and costly!

The days of huge servers and lengthy software implementations are over. With today's software technology, version management can be implemented with extremely little effort - the system could even be operated directly from a USB stick. All that is needed is a central server and any number of installed clients.

Thanks to this server-client architecture, users can also work in offline mode and check in new versions at a later date. Another advantage is the intelligent user administration, which generates all user data through automatic synchronization with the Active Directory. This eliminates the effort of setting up user accounts initially, and existing accounts are regularly updated automatically or deactivated if necessary. The unique user assignment prevents unauthorized access and automatically documents the respective user.

Companies that focus on audit trails - i.e. the complete logging of all users and their activities - can even kill two birds with one stone: in addition to the development documentation, they receive reliable documentation of who changed what, when, where and why.

Myth 3: A version control system causes the rationalization of workplaces

Version differences can be easily identified and checked using a graphical comparison.

© Auvesy

Even in highly automated systems, qualified employees remain an important and irreplaceable resource. Supporting software systems can only ever be as intelligent as their programmers and users. Especially in the area of data management, correct and conscientious data maintenance is a decisive factor. The aim is to automate time-consuming but low-skilled tasks - such as manual data backups, manual version synchronization and time-consuming searches for data carriers and storage locations for data backups - as far as possible. This frees up personnel resources and, above all, their expertise for demanding, value-adding and forward-looking projects. Despite the dual control principle and multiple checks, human error can occur at any time. Maintenance staff are under great pressure, especially in times of high production capacity utilization. If changes have to be made, the certainty that the last valid version is always available and can be restored minimizes this pressure immensely. This is because incorrect programming is immediately reversible and there are no high follow-up costs.

Myth 4: Our ongoing manual version management strategy is sufficient!

The comparison of file size and date alone does not enable any meaningful version management, let alone a detailed comparison of the content of the (version) control programs synchronized on the server. Not to mention a clear identification and labeling of the valid and last released version.

Inhomogeneous automation systems require different project planning tools and editors, which are maintained and programmed by an ever-growing production and maintenance team in shift operation. Only a software-based solution can bring light into the darkness here. Leading systems integrate the familiar editors that the user works with on a daily basis and support the user with menu-guided documentation and automated data backup. Thanks to the familiar environment and intuitive operability, training costs can be reduced to a minimum.

Myth 5: A homogeneous automation environment is a basic requirement!

Individual control system manufacturers also offer version management solutions. However, these only selectively support the manufacturer's own devices and are therefore only really effective in homogeneous systems. But where can homogeneous automation systems still be found today?

The automation market and the number of manufacturers and suppliers are constantly growing, which also increases the complexity of the systems. For this reason, today's systems are a colorful mixture of different robot types, field devices, control programs, drive systems, programming languages and file formats.

The unique selling point of a future-proof version management system therefore lies in its manufacturer independence and openness. It is important that not only the most common automation systems are supported, but also that the latest device versions are regularly adapted in order to provide the required comparisons.

Myth 6: Version management only works if there are no external suppliers involved!

Terms such as lean production and lean maintenance have become an integral part of today's working environment. With a focus on increasing efficiency and productivity, it is therefore rather the exception that manufacturing companies do not commission external service providers and suppliers.

For this reason, a version management system must also be able to track, monitor, compare and check changes made to ECUs by system integrators and OEMs. In addition, the question of "why" plays a major role when working with external service providers. Only the documentation of reasons for changes ensures complete validation and traceability, which is still meaningful over time and with regard to the history of changes. This means that such a system offers great added value, especially for companies that regularly come into contact with external suppliers. Projects can be handed over to specific companies, changes can be clearly tracked and parallel processing by in-house employees can be prevented.

Myth 7: Backup and version management are two separate pairs of shoes!

The data management cycle: The central server serves as a data store on which all versions are stored. To edit, the last version is checked out to any client, changed and the new valid version is checked in.

© Avesy

Important: Version management is not a substitute for backups, even less is a backup a substitute for version management. These are two different tools, but together they provide the greatest synergy effects and ensure high availability of relevant data.

This is because central data backup or version control considered separately do not provide 100% security for consistent data. Whether the centrally stored projects actually match the productive programs (offline and online versions) can only be ensured by regularly (automatically) comparing the software versions. Unnoticed changes can also be detected and analyzed in this way. On the other hand, it makes no sense to version backup data automatically. Solutions that cover and combine both aspects as a complete package are therefore also referred to as data management systems with a view to the future.

At the end of the day, not all backups are the same. In order to protect production from longer downtimes, disaster recovery, i.e. system recovery in an emergency, must be feasible in the event of data loss due to overwriting, deletion or device failure. A restoreable backup of the last valid version is required for such a disaster recovery. This means that symbols and comments must also be removed. When selecting an automated data management system, production managers should therefore always consider the type and quality of the data backup performed. After all, a restore-capable backup is the basis for achieving maximum data and system availability.

Author:
Silke Glasstetter is Marketing Manager at Auvesy.