FAQ: 5G and the Factory
By Marco Contento
January 15, 2021
Factories are already experiencing 5G’s faster speeds and low latency and are implementing 5G-leveraging solutions, such as augmented and virtual reality, automation, and wearables, on the manufacturing floor. The coronavirus pandemic has brought new challenges to manufacturing, like remaining agile and efficient with limited personnel. 5G can help factories pivot to a more secure and remote manufacturing approach and drive new business opportunities in greater numbers.
During our IoT Solutions World Congress webinar, “5G in the Factory: How 5G IoT at the Edge Can Elevate Your Manufacturing Business Strategy,” we received many excellent questions. Here are answers from Telit’s Marco Contento to the top questions we received.
5G Release (Rel) 16 aims to support LAN-type services over 5G radio links, focusing on Time-Sensitive Networking (TSN).
The role of MNOs and MVNOs in smart factories can vary, depending on the architecture they offer and the position they take. It will be an OPEX-based approach for factories whenever they get service from an MNO or MVNO. On the other hand, the MNO or MVNO will have the opportunity to deliver a service and be paid for it since managing the wireless infrastructure requires knowledge that enterprises don’t have in-house or don’t want to build.
5G is a new technology. It’s complicated and, in this phase, expensive. From the device perspective, we all use wireless technology from the mobile phone market, which isn’t optimized in specs or cost for manufacturing. We will need a bit more time to see more affordable and suitable components and devices for the manufacturing industry. From an infrastructure perspective, private network solutions are expected to become more available from traditional large system integrators and several disruptors and innovators. Telit is working with all of them to enable the full solution, with device-side hardware, device management and data orchestration crucial to private 5G solution architecture.
The most common challenge is usually interoperability with infrastructure implementation.
It depends on the spectrum allocated in the country considered. For instance, in Germany, they have allocated 100 MHz spectrum in 3.5 GHz/n78 band, so it’s single band. On the other hand, Japan will deploy local 5G in NSA (non-standalone), using LTE as the anchor on B41 plus mmWave n257.
Sub-6 spectrum (also known as FR1 spectrum in 3GPP language) are all the bands below 6 GHz. Therefore, they include legacy 3G/LTE bands and additional new bands still below 6 GHz.
mmWave consists of new bands above 24 GHz allocated for 5G.
It depends on the requirements. If low latency is required, Wi-Fi 6 isn’t suitable. Also, 5G is usually deployed in licensed bands, so there’s better quality of service (QoS). Contrary to 5G, Wi-Fi works in unlicensed bands, and the QoS is “best effort.”
Yes. The ambition is to use 5G to replace wired real-time communications. 3GPP Rel 16 aims to do that.
One option is to implement ML on a host board attached to the module or run ML on the module itself if there will be enough computational power. It’s difficult to judge whether TinyML can fit since it depends on the requirements and microcontroller available on the module, which can vary.
There’ll be either specific converters exposing Ethernet or, when 5G becomes more affordable, the 5G modem will be integrated into the PLC when and if required.