Martin Danzer is Director Product Management at Congatec and studied electrical engineering at the Deggendorf Institute of Technology. He has over 20 years of experience in technical service, development management and product management for Computer-on-Modules, including his time at Kontron and JUMPtec.

The Industrial Internet of Things is generating more and more data that needs to be managed. This data complexity needs to be mastered. What role does edge computing play in this?

Danzer: Edge computing enables data to be processed directly at the point where it is generated, i.e. at the edge of the network, and is therefore an alternative to traditional cloud solutions. This data needs to be processed so that it can be used effectively for analyses and derivations. Previously, this was carried out in central data centers due to the required CPU performance. This was particularly necessary if the data was to be evaluated by AI, e.g. using deep learning algorithms. Modern CPU generations now have sufficient performance and built-in acceleration technologies to evaluate the data directly where it is collected. Edge computing conserves the network's resources and results can be reported back with little delay. The current trend is therefore towards a decentralized structure, as technological progress allows. This saves network bandwidth as no or very little data needs to be sent between the edge unit and central servers. Data processing is independent of the availability of high-speed connections and is not dependent on reliable Internet connections at all times. In addition, this allows for quick and easy scaling as additional edge points can be integrated into the networks at relatively low cost.

Scalable edge computing is therefore particularly suitable for applications where large volumes of data need to be processed and where only minimal latency times are permitted. This is playing an increasingly important role in a wide range of applications, such as collaborative computing, autonomous driving, factory and building automation and industrial IoT in general.

At the edge, sensor data is pre-selected, software applications are processed or even AI calculations are carried out. What should users bear in mind with regard to the edge software landscape? Which technologies should/must be taken into account?

Danzer: In order to run the software as effectively as possible, the applications should be adapted to the features of the CPU and use the right frameworks so that the software can also exploit the full potential of the new CPU generations. It is also often the case that edge devices have several tasks, e.g. collecting sensor data and at the same time processing, displaying and transmitting the data obtained via protected data connections. To minimize interference between these two subsystems, it makes sense to install Workload Consolidation with the help of a hypervisor. The subsystems are completely isolated from each other and the system's resources are clearly separated and distributed. Instead of an operating system that cannot cover all sub-areas equally well, a multi-core platform and the real-time hypervisor make it possible to select the best operating system in each case. The hypervisor distributes the system's resources, then boots the operating systems with the allocated resources and is then no longer active. There are no restrictions - even real-time operating systems run with unaffected latency.

Congatec offers embedded and edge computing products and services.

© Congatec

In addition, the ability to perform over-the-air (OTA) updates is attracting more and more attention, which of course requires a high level of security. Security is not only a prerequisite for the base system itself, but also for the secure connection to sensors and actuators, so that local attack vectors are eliminated here too.

The demands on the hardware grow with the tasks. What should users look out for when selecting the right edge computing hardware?

Danzer: The performance requirements for the hardware must be carefully analyzed during the concept and development phase. Users need to get a clear picture of the specific operating conditions, the necessary computing performance, the required longevity and long-term availability of the hardware at an early stage. The modular concept enables simple and effective scaling as requirements increase. Therefore, care should be taken during the selection process to ensure that all required interfaces are supported and that sufficient scaling potential is available. The environmental conditions to which the system will later be exposed must also be taken into account. A commercial platform with a temperature range of 0 to 60 °C may not be sufficient if the edge system is to be used outdoors. The maximum physical size of the planned system is also crucial. With COM-Express and COM-HPC, a wide range of sizes can be realized with various gradations. Priorities may have to be set here in order to find the ideal CPU platform on the ideal form factor.

With computer modules, SMARC, COM Express or COM-HPC depending on the line requirements, it is possible to realize high operational reliability in demanding environments, scalability, upgrade and expansion options using compact form factors. This means that customers are also ideally equipped to meet increasing requirements in the future.