There are already various companies that have started to use cloud-based portals to centrally connect any network-compatible sensor technology and act as data collectors in this way. While the basic approach is heading in the right direction, in many cases it is difficult to determine the added value for customers, as the portal manufacturers and operators do not have the specific domain knowledge of the customers in question and therefore have functional deficits. MES manufacturers are aware of the very different requirements of individual industry segments. From a technical perspective, traditional MES manufacturers have already assumed precisely this role of portals in the past in the locally defined environment of a production plant by consolidating diverse, heterogeneous data sources such as machines and sensors in a central location, thus creating real added value for the user.

So what could be more obvious than extending the scope of an MES to the cloud, not only covering the localized environment of immediate production, but also supporting the use of products by customers in the field? This is done by collecting the data generated there via a portal system and, in conjunction with the data from production, creating real added value for both the customer and the producer.

Various public clouds possible

A prerequisite for successful implementation by MES manufacturers is that they take the cloud approach into account at the basic level when designing their systems. The iTAC.MES.Suite has already been increasingly used by customers in private and hybrid cloud systems for several years. With the iTAC.IoT.Service, Itac has now created a solution that can also be easily used in public cloud systems such as Amazon AWS or Microsoft Azure, thus taking the potential of classic MES to a new dimension by combining the portal concept for any data source.

The technological basis must of course also take into account all previous aspects of a production environment - such as the longer life cycle of the equipment, heterogeneous interfaces, simple software updates and the protection of data against unauthorized access.

Representation scheme/principle of cloud platform independence using 'Docker' technology. Docker is a software that allows applications to be isolated (clustered) in containers with the help of operating system virtualization, so that each application works in a resource-optimized manner.

© Itac Software

The solution uses 'Docker containers', which allow a very lean type of virtualization at process level. The containers enable the encapsulation of certain services and can usually be managed quite easily in the standard Docker-based container management systems of the various cloud providers. Alternatively, certain Docker containers - such as databases - can also be used from the cloud providers' platform services. Here too, the focus is on simplified system operation and increased availability.

The interfaces between the IoT services should be based on standards. Protocols such as OPC UA are used in the industrial environment: it is characterized by platform independence, flexible modelling, efficient protocols and security through encryption and authentication.

Interfaces between cloud and devices

The OPC Foundation is currently working on an extended model to support publish-subscribe scenarios in combination with the Ethernet extension Time Sensitive Network (TSN). While the real-time requirements of TSN are aimed more at shop-floor IT infrastructure, the pub/sub scenario is ideal for IoT applications as it does not require a permanently open point-to-point network connection. The use of standardized protocols such as AMQP enables connection to any IoT cloud platform.

Specialized embedded systems on the store floor allow existing machines and systems to be retrofitted with the appropriate connectivity for the IoT age. Many systems that have already been connected for MES application scenarios in the past can also make their data available to the MES on the IoT platform via the existing interfaces.

In addition to the cloud-based architectural approach, a key differentiator between a classic MES and an MES in the IoT age is the expansion of services, such as machine learning.

Goal: Extended service

The embedded PC based on Linux iTAC.smart.MESDevice enables the simple integration of system-related devices such as barcode readers, RFID scanners and transport systems.

© Itac Software

All process-relevant data that has a direct or indirect influence on the quality of the products is required for the optimal operation of the machine learning services. The aim is to reduce costs, but users can also consider the efficient use of available resources. This can be achieved, for example, by increasing the availability of machines and systems through predictive maintenance or by supporting maintenance personnel in troubleshooting, for example through pattern recognition and tools such as augmented reality. En passant, the reliability of planning is increased, which leads to higher customer satisfaction in addition to the increased quality level of the products delivered.

The possibilities of IoT platforms also include networking with information sources from the customer's field. For complex products in particular, it makes sense to collect data over the entire life cycle of a product, i.e. beyond production. This data can provide valuable information and, in combination with production data, can lead to further product improvements through design changes or changes in the manufacturing process.

The requirements

All of these scenarios require the use of smart sensors in many areas - from the production process to the supply chain to the customer. In addition to the challenge of integrating and managing these swarms of sensors in an IoT platform, it must be possible to process these enormous volumes of data in near real time. Here too, it must be possible to expand the IoT platforms using suitable technologies, such as cluster-capable, scalable databases and analysis platforms. The cloud-based approach offers the flexibility required in this field in particular to adapt computing capacities to requirements at short notice.

The large number of networked devices naturally also requires a closer look at the security of the systems. Various aspects need to be taken into account here, from encrypting data transmission to securing individual devices against attackers. As soon as the sensor technology has a direct or indirect influence on the production processes via machine learning, protection against external interference becomes a necessity. This can mean that, in addition to the software, the hardware must also have suitable mechanisms to protect the devices.

Another elementary component of the IoT platform to support the sensor swarms is device management. It must have functionalities for managing the devices - this includes managing firmware versions and over-the-air updates, i.e. wirelessly from a distance.

To summarize, the IoT platforms can be classified as the central hub of the sensor networks, but only the combination of distinctive technical services of an MES and extended analytical possibilities creates real added value in Industry 4.0.

Author:
Volker Burch is Vice President Advanced Technology at Itac Software.