Hewlett Packard Enterprise (HPE), Intel and Microsoft are launching the Distributed Industrial Cloud, an Industry 4.0 solution designed to enable the end-to-end exchange of data and services in heterogeneous and distributed manufacturing environments. The solution is now available from HPE in Germany, Austria and Switzerland.

The centerpiece: an Industrial Service Bus

The three companies have now integrated their Industry 4.0 technologies into an edge-to-cloud solution and are offering it under the name Distributed Industrial Cloud, including services. At the heart of the architecture is an Industrial Service Bus (ISB), which was developed as open-source software together with production companies. The ISB is designed to avoid data silos, point-to-point connections and the resulting spaghetti architectures. Instead, it offers an end-to-end communication architecture that connects machines, systems and applications across locations and clouds.

HPE, Intel and Microsoft have integrated their Industry 4.0 technologies into an edge-to-cloud solution.

© HPE

"Standardization is often described as a prerequisite for interoperability in Industry 4.0 - but in reality, manufacturing environments are still largely heterogeneous," says Florian Dörr, Head of Edge Practice in the DACH region at Hewlett Packard Enterprise. "Our approach helps companies to achieve their Industry 4.0 goals even under these conditions. The openness of the architecture means that future standards can be integrated."

Technology-neutral any-to-any communication

The Industrial Service Bus (ISB) is an open-source message-oriented middleware that works with a publish-subscribe protocol. Each machine and each application can publish data services via the ISB, which are subscribed to by other machines and applications. This could be sensor data, deviations from target values or the results of processing such data, for example from MES or IoT applications.

The Industrial Service Bus: It is designed to avoid data silos, point-to-point connections and the resulting spaghetti architectures.

© HPE

The ISB can be used for both local and global data exchange. The ISB is based on a microservices architecture and is implemented with the Distributed Application Runtime (Dapr), an open-source event-driven runtime environment. Machines and applications communicate with Dapr - Dapr in turn regulates communication with the technology used for data exchange, such as nats.io, RabbitMQ or Kafka. This means that the ISB is technology-independent, fits into existing messaging, streaming and protocol environments, and these can be exchanged with relatively little effort.

The hardware and software components

In addition to the ISB, the Distributed Industrial Cloud includes the following technologies and platforms, which are designed to create a complete edge-to-cloud environment for Industry 4.0:

• HPE Edgeline EL300 Converged Edge System: an edge-optimized system that forms the physical OT-IT interface. It can be used both for real-time data analysis on site and for data transfer to the cloud. The system is capable of bi-directional control and communication with industrial systems, for example via high-speed digital input/output, CAN bus, Modbus or Profinet.
• HPE Edgeline OT Link platform: The platform has a workflow engine that can be used to orchestrate and automate the interaction of industrial networks, controllers and data with drivers, middleware and IT applications via a graphical user interface. At the same time, OT Link enables the operation of container applications directly at the edge. Central administration of the distributed infrastructure is carried out with the Edgeline Workload Orchestrator.
• Intel OpenVINO: A free toolkit for the rapid development of applications for machine vision with deep learning. It enables deep learning on the basis of hardware accelerators and on various Intel platforms. It includes the Intel Deep Learning Deployment Toolkit with a Model Optimizer and an Inference Engine as well as optimized computer vision libraries and functions for OpenCV and OpenVX. OpenVino thus supports the entire spectrum of image recognition solutions.

The scalable architecture of the Distributed Industrial Cloud.

© HPE

• Microsoft Azure and Microsoft Azure Stack: A collection of services, for example PaaS (Platform as a Service), IaaS (Infrastructure as a Service) and managed database services, which can be operated in the public cloud (Azure), hybrid or private cloud (Azure Stack). This allows production companies to operate their services where their technical, regulatory or organizational criteria are best met.
• Hardware and software-based security functions from HPE, Intel and Microsoft: such as hardware-based cryptography, tamper-resistant data, zero-trust network security and user and entity behavior analytics.

The fast-start package

HPE offers consulting, implementation and operational services for the Distributed Industrial Cloud. This includes a fast-start package costing around 20,000 euros, which comprises a workshop and all the necessary software, hardware and services for the rapid implementation of an initial use case. The architecture then enables smooth scaling from the first test run to a global edge-to-cloud production infrastructure. Via HPE GreenLake, the Distributed Industrial Cloud can also be procured in an as-a-service model - for example, by pricing per month and machine for all hardware, software and service components provided. This allows production companies to reduce their initial investment and scale their costs in line with capacity utilization. HPE takes over the end-to-end management of the environment, relieving the customer of the burden of managing the edge-to-cloud infrastructure.