COVID-19 Has Disrupted 5G Development but Not Stopped It

While COVID-19 has disrupted the 3rd Generation Partnership Project (3GPP)’s Release 16 and 17 timelines, it has not stopped the industry’s overall march toward developing 4G and 5G for industrial applications. The crux of Industry 4.0 is a wireless and highly automated factory, envisioned to be able to reconfigure itself according to manufacturing demands while being able to carry on operations 24/7 with little to no manual contact.

While 5G is still in early development phases, and 4G is being developed to support industrial use cases, we have seen many promising industrial developments that are continually being developed, studied, and trialed. These include Nokia’s own 4G-powered Oulu base station factory, which has thus seen a 30% increase in productivity; ABB’s Milan factory using collaborative robot (cobot) YuMi, which uses 5G for data transfer and machine vision; Ocado’s private 4G network for over 3,000 cobots; and the partnership of Amazon Web Services (AWS) wavelength with Verizon, Vodafone, SK Telecom, and KDDI to develop 5G edge services.

The backbone of these Industry 4.0 applications is connectivity. Today, many industrial applications are wired, with majority of wireless applications using Wi-Fi.

ABI Research’s latest Internet of Things (IoT) Market Tracker: Manufacturing (MD-IOTMMAN-107), forecasts 4G IoT revenue for consumer, durable goods, and industrial manufacturing as well as automotive to grow from US$428 million in 2020 to US$1.6 billion by 2026, at a Compound Annual Growth Rate (CAGR) of 25.7%. While 4G’s market is larger, 5G’s is expected to show rapid growth in the next few years and could potentially overshadow the 4G market in the long run. The forecast for 5G should galvanize stakeholders to begin developing 5G ecosystems and solutions for industrial applications alongside 4G, as IoT revenue for 5G is forecast to reach US$707 million by 2026, at a six-year CAGR of 532%. 5G can enable many mission-critical use cases, as well as enable new business models such as network slicing, and utilize new technologies such as cloud-native, which can reduce Operational Expenditure (OPEX) and increase system resilience and flexibility.

4G now forms the foundation for cellular connectivity. With 50 millisecond (ms) latencies that can support Automated Guided Vehicles (AGVs), video surveillance, and connecting sensors for process automation, many applications can be connected via 4G first before 5G for industrial applications become standardized. Currently, the industry is waiting on 3GPP Release 16 to support the integration of 5G with IEEE 802.1, in order to support Time-Sensitive Networking (TSN). Due to COVID-19, Release 16 has been delayed to June 2020, and whether or not further delays are expected is still unknown. SN provides deterministic and isochronous communication, as well as bounded reliability and latencies to guarantee Quality of Service (QoS), that could eventually replace wired services provided over Ethernet networks such as PROFINET or MODBUS.

Looking further into the future, Release 17 is aimed at supporting Industrial IoT (IioT) with 5G with further support for 5G private networks. Beyond that, network slicing is also being developed, trialed, and tested heavily by Chinese network vendors. For example, China Mobile and Huawei deployed Mobile Edge Computing (MEC) at Haier’s Qingdao factory to enable 5G network slicing for machine vision quality inspection.

Network slicing refers to a network that has multiple individual logical networks, each with their own configurable Service-Level Agreement (SLA), and dimensions such as bandwidth and latency. While the 5G Standalone (SA) core is not yet ready, network slicing has already entered the market in some capacity:

• Network slicing is a cost-effective solution for small to medium-sized manufacturers to deploy Industry 4.0 applications, without having to deploy a full private network. It will transform Communication Service Providers (CSPs) from merely data traffic pipes to flexible service providers. Monthly subscriptions would enable users to order network slices on a store portal and have them automatically deployed to the facility. This includes support and SLAs, and an upfront cost for a MEC box to be deployed on the facility.
• Pricing can be based on each individual slice and the requirements and SLAs. For example, the exact requirements of each network slice can be customized by the user on a matrix, including bandwidth, latency, security, development support, etc.

Network slicing will allow users to customize on-demand needs for their factories while “leasing” resources from the CSPs. However, as the 5G ecosystem grows and expands, so will its functionalities, including edge computing, Network Functions Virtualization (NFV) and Management and Orchestration (MANO). Thus, we will see more intermediaries coming into the equation and providing some of these services that CSPs may not be able to capture with their current capabilities. Examples include AWS providing its cloud to enable the telco edge, and potentially the likes of Microsoft and other Mobile Virtual Network Operators (MVNOs) that are more equipped with cloud-capabilities, without the need to deal with legacy equipment that is not cloudified—a challenge that CSPs face. CSPs must define their working relationship with these intermediaries to determine a mutually beneficial pricing model and shift to a cloud-based network capable of network slicing.