5G networks will change numerous industries, including manufacturing. 5G in manufacturing won’t just bring speed and efficiency but new capabilities through data-driven processes.
There has been a lot of talk about the impact of 5G in manufacturing.
It’s part of a larger conversation about the Fourth Industrial Revolution—a new era ushered in by a collection of technologies, including ubiquitous mobile broadband through 5G networks, Internet of Things (IoT)- devices, artificial intelligence, robotics and cloud, edge and quantum computing.
The Fourth Industrial Revolution, and 5G networks in particular, brings data insights and operational efficiency to the factory floor and the supply chain. With 5G networks being able to transmit data 20 times faster than 4G, the factory floor will not only be more operationally efficient and automated but also more data driven, using contextual information to complete tasks and make decisions.
Consider that factory managers, for example, will know what the demand for a certain product is before production commences. That will help them determine in real time how much to manufacture. That’s just one example of how a convergence of these technologies will bring speed and real-time data to manufacturing for real results.
To run manufacturing machinery efficiently, machines need to operate under optimal conditions. No manufacturing facility can become 100% efficient. Machines run at less-than-optimal rates. The wear on parts can degrade functionality. Preventable energy loss may be ignored or missed. To maximize the efficiency of a manufacturing process, one must measure key metrics from machines, such as temperature, vibration, throughput and other device-specific characteristics. With a steady stream of metrics, however, monitoring applications can detect problems early and alert operators to issues that need attention.
Monitoring requires sensors—many of them. It also requires the ability to transmit large volumes of data from sensors to analysis devices. This is where 5G networks enable a level of data collection that was not possible with previous technologies. This data-collection capacity is used in a variety of ways to improve efficiency. An Industrial Internet of Things (IIoT) analytics firm, for example, has developed software that identifies conditions that degrade battery life in IoT devices, enabling manufacturers to extend the useful life of sensor batteries. A common challenge in manufacturing is that production lines can stop for many reasons, including missing raw material or component failure. With sensors deployed throughout the factory, it is possible to identify a full range of stoppages, from major ones that can be quantified using traditional management techniques to minor ones that are difficult to analyze.
Advances in 5G will also enable greater use of robotics in manufacturing. The 5G release 16 standard includes the ability to synchronize devices to a master clock. This is important for distributed systems that need to operate on a common time. For example, multiple robots on a production line may need to coordinate activities in ways that require a synchronized time and the ability to share information among the devices, such as current throughput. With precise and synchronized time and real-time data exchange, manufacturers can deploy more complicated processes, including those that require some adaptability to current conditions on the part of robotic components. This is just one way in which 5G in manufacturing will make robotics more efficient and pervasive.
Some manufacturing costs, like labor and supplies, are obvious and easy to measure. Others can be more difficult to quantify, such as the cost of a down production line or poor-quality outputs. Both of these costs can benefit from data collection and analytics enabled by 5G networking.
5G networks will bring data insights and operational efficiency to the factory floor and the supply chain.
Manufacturing processes depend on myriad devices and machines. When one breaks down, it can stop production and disrupt operations. Detecting potential failures before they happen not only keeps production lines operating, it also saves money. Preventive maintenance is 12% to 18% less expensive compared with the cost of reactive maintenance. This approach is well known, with more than 45% of manufacturers implementing preventive maintenance programs.
Poor quality is costly. One quality control vendor estimates that poor quality in the manufacturing industry costs 15% to 20% of total operations costs. One challenge with collecting data on quality control is that the process requires manual labor. This limits the scalability of some quality-control practices. Continuous monitoring of manufacturing processes can allow manufacturers to detect poor quality, anomalous events or other changes that can adversely affect quality. But this information is most beneficial when it’s delivered in near-real-time. This network responsiveness is precisely what 5G enables.
Worker safety is a primary concern in manufacturing. The industry invests heavily in personal protective equipment, safety training and connected worker devices among other risk mitigations. Still, injuries occur. The International Labor Organziation (ILO) estimates that there are approximately 340 million occupational accidents every year. Hazardous substances cause more than 650,000 deaths per year. In addition, older workers are more likely to be injured in work-related accidents. With an aging workforce, businesses are facing potentially higher rates of injuries in the future if safety practices for older workers are not improved.
5G networking enables new types of safety measures while enabling advances in existing safety technologies.
One of those areas is automated safety analysis of videos of operations on factory floors. Video monitoring of manufacturing facilities can help identify operations or areas that need additional attention to safety. Video monitoring can capture the context surrounding an accident, as, for example, when a human is injured working alongside a robot. Monitoring also helps identify insufficient safety practices, such as an area where workers may not always wear all required safety equipment. 5G provides the networking bandwidth needed to capture video from across a manufacturing facility, and advances in video-processing AI will be used to analyze images and detect problematic safety practices.
5G in manufacturing also enables advances in wearable safety technology. Workers in manufacturing facilities can wear sensors that monitor the movements of a worker as well as the area around them. This provides the foundation for more intelligent safety monitoring and alerting. For example, connected worker devices can detect autonomous vehicles in the area that are transporting materials and help avoid injuries. Connected worker devices is an area ripe for innovation. As noted by Bill Pennington, a senior analyst at Verdantix, an environmental health and safety (EHS) consulting firm, the “mission statement for EHS is rapidly moving from a compliance focus to an innovation focus to add business value and support growth.”
A confluence of technologies is fueling innovation in the manufacturing sector. The pace and types of changes are significant enough that industry observers describe it as a fundamental shift to a more digitized manufacturing environment dubbed Industry 4.0. McKinsey and Co. defines it as the next phase in the digitization of the manufacturing sector, driven by critical disruptions: increases in connectivity, data, and computational power; advances in analytics and business-intelligence; innovations in human-machine interaction; and developments in robotics and 3-D printing.
Expect to see manufacturers taking advantage of 5G networking as they deploy more sensors, adopt machine intelligence for predictive maintenance and process optimization and use change processes to leverage the benefits of joint human-robot operations.
Dan Sullivan is a software architect specializing in streaming analytics, machine learning and cloud computing. Sullivan is the author of NoSQL for Mere Mortals and several LinkedIn Learning courses on databases, data science and machine learning.