Interview about Cloud

In 2013 / 2014, we at M&M Software have already identified Cloud Computing as one of the central technology drivers for the future digitalization in the industry. Therefore, at that time, we set ourselves the goal of positioning ourselves as the leading provider of cloud-based solutions and services in the industrial and automation environment and realigned our strategy accordingly.

When we entered the topic of Cloud Computing, most users were still very uncertain about the benefits and data security. Therefore, in the early days, it was mainly necessary to explain the basic principles and advantages of Cloud Computing and the associated security concepts. With workshops and self-developed prototypes, we worked on convincing potential users of the advantages of cloud-based solutions and their feasibility.

Today, there is no need to convince people of the Cloud and its advantages. Most people can classify the term Cloud Computing and know the difference between Public and Private Cloud. They have realized that the big Cloud portals offer a great level of data security. It is difficult, if not partly impossible, for your own professional data centers to replicate this level of security and availability.

Cloud Computing has established itself in many areas of our lives. Social networks like Facebook, Twitter or WeChat are operated on Cloud infrastructures. Entertainment programs are increasingly streamed from the Internet with services such as Netflix and Spotify. More and more purchases are being made in the large online shops, such as Amazon or Alibaba, instead of in the shop next door. Personal data is stored in unlimited available Cloud Storage like DropBox or Onedrive.

Innovative cloud-based cooperation platforms such as Teams or Slack promote agile working. Larger companies are increasingly moving their mission-critical business applications to the Cloud, be it SAP's S4Hana or simply Microsoft Office365. Major car manufacturers such as VW and BMW are announcing strategic partnerships for the production systems or mobility platforms of the future with leading Cloud providers such as AWS or Microsoft Azure. The goals of Industry 4.0 and especially the pragmatic implementation in the Industrial Internet of Things (IIoT) have proven to be additional drivers for the implementation of Cloud Computing. Topics such as OPC UA, MQTT or even Digital Twins are being taken up by the IoT platforms of the major cloud platforms and implemented on a large scale through corresponding services.

Cloud computing has thus arrived and been accepted in both private and professional life.

Because size plays an important role for platform providers. Only so-called hyper scalers such as Amazon (AWS), Microsoft Azure and Google Cloud Platform can offer increasingly sophisticated cloud services competitively and on a large scale. The three companies mentioned above cover around ¾ of public cloud offerings. Even large companies, such as Siemens or SAP, are now entering into partnerships with them to host cloud-based solutions such as Mindsphere or S4HANA.

While large companies often work with several public cloud providers in order not to become too dependent, we have deliberately specialized in services around the Azure platform from Microsoft from an early stage. There are historical reasons for this, because as a long-standing Microsoft partner, our relationship with the Redmonters has traditionally always been very close. On the other hand, the variety of services offered by the large public Clouds is growing so rapidly that for technical and economic reasons it makes little sense to master all platforms equally well and up to date. For us, our decision was and is the right one. Today, we can offer our customers tailor-made cloud solutions based on Microsoft Azure, we can act as an Azure Cloud Service Provider, procuring the required Cloud resources and, if desired, also provide long-term, secure operation.

The situation is a bit different in the Chinese market, which is addressed by our subsidiary in Suzhou. Although there are also offers from Western Cloud providers here, Chinese companies prefer the offers of large Chinese players such as Alibaba. Therefore we have a slightly more open technology strategy in the Chinese market, which is important for us.

Today, we offer our Cloud solutions and services mainly for our traditional target industries of factory and process automation, mechanical and plant engineering, energy, building automation, but also related areas such as oil and gas or maritime technology. In the meantime, Cloud Computing is successfully used everywhere as a major technology driver for digitalization projects. Cloud computing can play out its strengths particularly when it comes to managing widely distributed technical assets in a central location and merging data for further analysis. This makes extended service offerings or completely new business models possible in the first place. Be it in the case of a mechanical engineering company, which thus enables better predictive maintenance and thus a more efficient service business with its customers. Or in the case of a municipal utility company, which can use it to better monitor its local distributors and thus better manage energy consumption peaks. Or, as in the case of a provider of ship automation technology, which enables its customers to manage their fleets worldwide. All these providers save on their own data centers, rely on the public cloud and can thus focus on their industry-specific solutions, i.e. the software.

The development was rapid, and we had to expand our horizon quickly, not only in the technical area. The pace of development of cloud services was breathtakingly fast, as some of them sprang up like mushrooms. But we rose to this challenge and mastered it with a lot of commitment and newly acquired knowledge.

In particular, the first stage of strategy implementation was characterized by a somewhat naïve approach on our part. We were indeed ahead of the times with our first IoT platform called dXpert, but then we were virtually overwhelmed by the massive development of the large Cloud platform providers like AWS and Microsoft. When they recognized the great role that IoT platforms would play in the digitalization environment in the industry, they invested massively in building their own IoT platforms as integrated basic services in their Public Clouds. In the face of so much concentrated market power and resources, our own platform approaches were not far-reaching enough and not promising in the long term. For this reason, we changed our strategy in such a way that we wanted to build our own solutions consistently on the platform of a major Cloud provider such as Microsoft Azure in future.

We learned from this development that as a service provider, it is not enough for us to develop basic infrastructure components ourselves. Instead, we must be able to assemble effective industry solutions based on the diverse services of the major cloud platform providers. That is challenge enough, because the speed of development of PAAS components in the area of cloud computing in general and IoT in particular is breathtaking. We and our customers accept the associated dependence on the platform provider in order to benefit from the greater speed and efficiency in implementation.

At the beginning, we underestimated the complexity of the contractual and legal aspects of services that are supposed to be open to many end customers - often across borders. Here we had to learn to reconcile what was technically feasible with what was legally permissible and contractually robust. The implementation of service-based business models is not possible without the involvement of lawyers in order to make the necessary terms of use and service level agreements (SLA) legally secure. In the meantime, we have been able to develop a solid basis of solution samples and contract templates, which hopefully can be transferred to new customers and projects with less effort in the future.

We have also learned early on that cloud-based solutions in the industry are not limited to just implementing services in the Cloud. The realization of cloud connectivity is also important on the side of the controllers and field devices, which together with the Cloud unite to form the Industrial Internet of Things (IIoT). This was one of the reasons for setting up the business unit System Solutions at M&M, which should specialize in the realization of such vertical IoT solutions in industry.

Today, the business unit System Solutions can implement integrated solutions from the sensor to the Cloud. To do this, a wide range of different technologies must be mastered. In addition to expertise in cloud-based Azure services and the development of modern, web-based applications and mobile apps, this also includes skills in the embedded area, in the development of PLC applications, in industrial 4.0 communication based on OPC UA and MQTT and in the integration of modular edge services based on docker containers. Our Data Scientists use the Cloud to train AI algorithms with the collected data and then integrate them for execution either in software architectures in the Cloud or directly on-site on edge devices or powerful controllers.

DevOps is a process improvement approach in which software development and IT operations are combined into an integrated process based on powerful, often cloud-based, tool chains. New features are developed, checked into the version control system, built centrally, tested automatically, then subjected to further system testing in dedicated test environments and, after successful test completion, rolled out automatically to the final runtime environments. DevOps works in principle for any software solution, but especially well for cloud-based solutions. Thanks to DevOps' approach, we are not only able to provide our customers with new, improved releases in high frequency, but also - thanks to the appropriate instrumentation - permanently monitor them for correct function and thus meet guaranteed availability requirements. Over the next few years, we will gradually apply the DevOps approach to all our software projects and expect to be able to further increase the efficiency of our software development.

A major issue is data sovereignty. Who owns the device diagnostic data collected from the field devices of a sensor manufacturer in an operator's plant? The sensor manufacturer or the plant operator? Both certainly have a legitimate interest in such data. Therefore, technical and contractual prerequisites must be created that meet the interests of both parties. Cloud-based services from a wide range of providers will also be increasingly interlinked in the future, but this will require standardized interfaces. Initiatives, such as the Open Industry 4.0 Alliance, also attempt to do justice to such questions by underpinning corresponding reference architectures with concrete implementations and thereby bringing together manufacturers from a wide range of areas. This is a significant advance compared to the highly theoretical approaches of the German government's Industry 4.0 platform. A reason for us to join the Open Industry 4.0 Alliance.

In the environment of Cloud and IIoT, you can also observe today again the perpetual battle between two fundamental forces in information technology. On the one hand, the attempt is being made to run services in the Cloud as centrally as possible in order to benefit from simple management and massive scalability. On the other hand, real-time, availability and data protection requirements are pushing decentralized concepts, such as fog or edge computing, with the aim of bringing the execution of functions as close as possible to the production environment.

At the moment, the trend towards decentralization is somewhat stronger again, driven in particular by technologies such as edge computing and docker containers. In production, edge devices are positioned logically between controllers and field devices on the one hand and cloud applications on the other. The mostly Linux-based edge devices thus support local intermediate processing, but also form the bridge to centralized and cloud-based asset management. Docker containers have become a widely used, lightweight virtualization technology that makes it easier to distribute and manage individual applications and services, even on edge devices.

However, the past has also shown that the trend towards centralization and the trend towards decentralization are in waves and that technical innovations can often trigger a trend reversal. With the availability of Time Sensitive network (TSN), control functions can be executed in virtualized environments within central data centers in the future. With the broad availability of 5G, many applications that are currently still executed locally in a decentralized manner will be able to migrate to the cloud in the future.

So, it remains exciting and we are curious to see what the automation world in industry will look like in another five to ten years. It will certainly be significantly different again from today.