Nothing has pushed the digitization more into the middle of our daily lives than radical changes to our comfort zones, as it has happened for example in the pandemic of Covid-19. All of a sudden, tools and methods that were partially in use in professional business matters such as video conferencing, online meetings and trainings became a part of our new normal. While this sudden increased demand in digital eco systems and services could be called a revolution, the digitization of the manufacturing industry took a much longer time to ramp-up. It has been in progress after several years of governmental activities, demonstrating the benefits in proof-of-concepts and moving standards forward. However, the desire of having a seamless sensor-to-cloud communication, as it is required for business models such as predictive maintenance and quality or smart robotics, still faces some gaps in the network infrastructure.
Single-pair Ethernet, as a new initiative to enable IP based networks to each sensor, adds value through thinner cabling, small footprint plugs and connectors and less space requirements. For that reason, it is suitable to replace numerous existing fieldbuses in the sensor and peripheral area by integrating sensors directly to IT- and Cloudsystems.
Hilscher expects an adoption in the process industry somewhere in 2021. Discrete industries are expected to follow later towards 2025, as the Ethernet infrastructure and specifically IO-Link solves already numerous challenges in digitization today.
From Hilscher perspective, there is no doubt, that SPE will be adopted and take market share in the longer run. Our netX family of network processors offer an opportunity to adopt to SPE infrastructure easily and build gateways from existing network infrastructure to SPE in order to enable a migration path.
This Blogseries discusses the status and progress of single-pair Ethernet from Hilscher’s point of view.
Digitization progress and benefits
Despite the fact that it is called the 4th industrial revolution, the manufacturing industry started their journey into the digitization age around 2011 already. The initiatives behind carried titles like “Plattform Industrie 4.0”, “Industrial Internet Consortium” or “Made in China 2025” and targeted to position and fund programs to use digital methods and techniques to overall improve the performance of production facilities. Various global consulting companies analyzed and called this a “game changer, seizing a trillion dollar opportunity” (Accenture, 2015).
A “new industrial paradigm” became visible with a “need for increased intelligence in embedded systems and value creation though smart services” and key success factors to be advanced analytics for predictive production actions and process transparency” (Capgemini, 2015). A McKinsey analysis also in 2015 based on 100’s of interviews with industrial companies concluded similarly that “disruptive technologies will enable the digitization of manufacturing sector” and lists among others cloud technologies, advanced analytics, touch and next-level GUI, virtual and augmented reality, advanced robotics and additive manufacturing. Based on a number of research institutions, they concluded to value drivers as in the following chart:
Therefore, the expectations appear to be quite high and the potential indeed tremendous, but despite numerous proof-of-concepts driven for example through the Industrial Internet Consortium, the challenge remained how to enable vertical networking of smart production systems from factory floor into IT systems.
Today, after several years of research and search, we face now a much more focused and commonly shared understanding of the needs and benefits of digitization in the manufacturing sector. Technically, OPC UA has become the most common standard to help solving the vertical semantic challenge between IT and OT world. On the business side, several organizations drive initiatives forward that enable value add for production systems in the various industries similar to those predicted back in 2015. We at Hilscher engage and drive programs and business models in the Open Industry 4.0 Alliance.
We support an open shared eco-system to commonly deliver customer value in collaboration with a large partner network with our netFIELD products and services. However, connecting cloud services to the sensor level is still a challenge and one opportunity to resolve the access is coming up with single-pair Ethernet.
Starting in early 2000 and further in parallel to these digitization initiatives, the industry adopted Industrial Ethernet as the dominating communication standard in all industries gradually replacing the former fieldbus systems such as PROFIBUS, InterBus, CC-Link, Sercos or DeviceNet. Several variants of Ethernet were standardized to support the high demand for higher bandwidth and deterministic real-time communication that is specifically required in the factory automation related industries, such as automotive, packaging or food and beverages.
Different standards of Ethernet are driven from different automation companies and have a different level of modification demand of OSI layers on MAC and Link-layer to support real-time demand. Some protocols are based on the classical TCP/IP based Ethernet system, others modify the layers 3 and 4 and some require modified hardware to the Data Link Layer. The latest Real-Time Ethernet system Time-Sensitive Networking (TSN) and is under final release by the IEEE to standardize the real-time functions at layer 1, 2 and 3 for a common hardware base. All Ethernet variants share the ability to network the factory from the office side to the shop floor and enhance overall production performance, diagnosis and availability. However, the complexity of Ethernet and the network topology as daisy chain or switched networks limited its use in peripherals such as sensors and actuators. This gap in a seamless Ethernet-based IP network infrastructure is now addressed by standardization efforts for single-pair Ethernet (SPE).
The start to standardize Ethernet based on a single twisted pair originated from the automotive industry. In their in-car network, the existing standards such as CAN, MOST or FlexRay were too costly in cabling and software. Hence, a common standard to gradually replace these was found with Ethernet. However, the cabling effort for standard Ethernet was comparably high. Broadcom made a first pitch with BroadR Reach to show, that a simple twisted pair cable is sufficient to deliver high-speed data over shorter distances. The IEEE has picked up the standardization initiative under the well-known 802.3 Ethernet Standard and widened the scope to industrial and building area as well in order to address similar challenges there. In fact, the fieldbuses mentioned above in industrial space as well as LON, BACnet or Modbus in Building Automation are in focus, when targeting a seamless IP based network to each sensor.
As the vertical integration from sensors to cloud is a main basis for digital business models, the benefits to enhance IP networks to the sensor are attractive:
- Enhanced visibility, diagnosis and control
- Access to all automation equipment through one semantic with OPC UA
- One common vendor-independent tooling
- Enhanced robustness and availability
- Enablement of predictive quality and maintenance
Single-pair Ethernet now adds value through the thinner cabling, small footprint plugs and connectors and the less space requirements. For that reason, it is suitable to replace numerous existing fieldbuses in the sensor and peripheral area. However, a key question is how a deployment into the different areas in automation will happen.
Thank you for reading our first part. Next week up: Deployment and characteristics of SPE
Read part 2 here: Part 2 - Single-pair Ethernet: Deployment and characteristics
Accenture (2015), Purdy, Davarzani, “The Growth Game-Changer: How the Industrial Internet of Things can drive progress and prosperity”
CapGemini (2015), Bechtold, Kern, Lauenstein, Bernhofer, “Industry 4.0 - The Capgemini Consulting View - Sharpening the Picture beyond the Hype”