Services from the cloud are all the rage. Quite a few companies are transferring comprehensive business processes, software tools and even complete application landscapes such as ERP systems to the cloud. Image processing modules from the production plant are now also being connected to cloud systems across the board. This means that material can be reordered directly via the logistics ERP module, for example. The results from quality assurance can also be collected in the cloud and evaluated transparently. Due to the strong demand, well-known cloud service providers have already responded and are making corresponding offers available specifically for the automation industry. Microsoft Azure and Amazon Web Services (AWS), for example, support the Open Platform Communications Unified Architecture (OPC UA) standard, which enables end-to-end data transfer between different system platforms.

The required resources can be scaled flexibly and additional storage capacity or greater computing power can be added at any time if required. The cloud procurement model is also cost-efficient, as you only have to pay for the services you actually use. There is also the security aspect: cloud service providers generally operate clusters of several mirrored data centers, so they can guarantee highly available and fail-safe systems. Last but not least, the use of the cloud can considerably simplify development processes and the distribution of software.

Reluctance to use the cloud

Due to the many advantages, it makes sense to also offer solutions for industrial image processing (machine vision) in the cloud. However, some industrial users are still reluctant to use such services because they fear for the security of their data. The image information from machine vision applications often contains a lot of internal know-how that needs to be protected. Many users therefore have reservations. They are reluctant to outsource image data to external servers as they do not want to lose control of their systems and want to retain an overview of data access.

These concerns can be allayed if the most suitable cloud usage model is selected: For example, users do not necessarily have to use the public cloud, i.e. server resources from well-known providers such as Microsoft, Amazon or Google. The numerous advantages can also be realized with a private cloud. For example, the data can be stored in an externally shielded data center in your own factory building or in a local partner company. Here too, software, for example, can be distributed much more easily than in traditional on-premise infrastructures. A private cloud also makes it easier to ensure the reliability and scalability of systems centrally within a company or factory. However, if a public cloud is preferred, users should pay attention to the location of the servers. If the cloud provider operates its data centers in Germany, they are subject to the strict requirements of the European General Data Protection Regulation, which ensures a high level of security and protection against unauthorized access.

"Container as a service" model

The common cloud providers today are also flexible enough to offer demanding customers individual contracts, exclusive servers and private network connections.

MVTec Software is also bringing its image processing solutions to the cloud to enable users to benefit from the many advantages. For example, the "Container as a Service" (CaaS) model is being used as part of a pilot program.

The library of the standard machine vision software Halcon runs in a Docker software container in a cloud instance and is activated by a license server in the cloud. The number of containers and their hardware resources can be freely scaled. With this model, the customer benefits from maximum flexibility when hosting in a public cloud such as AWS, Azure or Google Cloud, but also in a self-operated private cloud. The only requirements are compatibility with Docker and a connection to the licensing server. In comparison: In the "Software as a Service" (SaaS) model, the cloud provider itself offers individual services and hosts them in its own data centers.