How Edge Computing and IoT Can Reinvent Manufacturing
By Chris Pemberton, Aug 09, 2022
This article is the third in a series exploring use cases and innovative approaches to edge computing and IoT across a range of industries where these technologies are expected to deliver outsized business value.
The edge-to-cloud opportunity
The manufacturing sector tends to gain real-time access to much of their data at the edge versus in the cloud. Many reasons for this include lower latency, increased cybersecurity, manageable data analytics, expanded interoperability, and reduced storage costs. These all may be valid reasons for staying near the edge, but by doing so, manufacturers are missing an opportunity to modernize their operations, increase revenues, and accelerate ROI.
Cloud computing provides an untapped vein for conducting big-data insights at scale. By expanding the capabilities of data-driven analytics in the cloud, manufacturers can identify and solve issues across a global big-data footprint.
Edge computing is a first step toward digital transformation, and some manufacturers are beginning to take subsequent steps by moving their data to the cloud for greater business impact. Here are three use cases of edge computing and Industrial IoT (IIoT) — the use of IoT in industrial sectors and applications — that are having an impact.
IIot takes networked sensors and intelligent devices and puts those technologies to use directly on the manufacturing floor, collecting data to drive artificial intelligence, predictive analytics and preventative maintenance, and machine learning.
1. Inventory tracking and plant monitoring
Tracking inventory and monitoring plant operations are becoming common uses for IIoT to improve operational efficiencies and productivity. To track the lifespan of its inventory, Microsoft’s plant in Suzhou, China took advantage of machine learning (ML) at the edge. Within hours of launching ML, the plant was able to identify inventory that was close to being obsolete. As stated by Darren Coil, director of business strategy, in a McKinsey report, “The data was always there, but we weren’t seeing it until IoT highlighted it for us.” The report also notes that a five-person team discovered and addressed this finding, saving the plant nearly $5 million in one year and cutting inventory costs by $200 million.
The next step for Microsoft would be to apply cloud-based inventory management to monitor and maintain not just the health of its inventory, but also inventory levels across global teams, locations, and applications. Data derived from tracking inventory at the edge could then be disseminated at scale, resulting in benefits such as new innovations that provide value throughout the enterprise.
Texmark Chemicals wanted to monitor its plants more effectively as part of its vision to build a “refinery of the future.” In 2018, the Texas-based petrochemical company completed an IIoT project designed to make chemicals safely and more cost-effectively, as well as to obtain better process analytics and increase up-time, productivity, and worker safety. Texmark worked with a number of partners, including Deloitte, Hewlett-Packard Enterprise, and Aruba on a three-phase build-out that included both a digital foundation of edge-to-core connectivity and location-based services that used wireless sensors, both high-speed data and video capture analytics, and a turnkey IoT solution featuring predictive maintenance and asset monitoring.
The plant is automated by sensored devices and advanced analytics, which generate insights and reduce any risks for human errors. For example, sensors were placed on the plant’s pumps to monitor performance. Before that, the pumps were monitored through a hardwired connection to a control center. This approach cost between $2,000 and $7,000. By placing $300 worth of wireless sensors on the pumps, Texmark has derived considerable savings. In addition, the company has seen a 50 percent reduction in planned maintenance costs, with the potential to significantly reduce the 35,000 hours that employees spend monitoring the plant each year. Imagine the possibilities if Texmark could run data and analytics in the cloud across multiple plants and its supply chain.
2. After-sales support
Superior after-sales support can make or break a sale. Once a product leaves the factory, manufacturers can gain a competitive edge by monitoring the product’s operation to provide the right support at the right time. This was the case at Johnson Controls-Hitachi Air Conditioning (JCHAC), a manufacturer of industrial air conditioning equipment in 150 countries. The company wanted to provide superior customer service and generate new revenue, yet it needed a cost-effective way to proactively monitor and conduct maintenance and other services on its air conditioning installations to improve product performance.
During manufacturing, the destination of an air conditioning unit is usually unknown, making it impossible to fit the right local SIM in the factory. Marubeni Network Solutions worked with JCHAC to provide it with a global connectivity service for an initial 120,000 units annually using a global communications management end-to-end IoT connectivity platform from Nokia. The technology enables JCHAC to factory-fit its air conditioning units with a standard SIM to provide connectivity in its target markets, avoiding the costs and complexity of manufacturing variants for every country.
JCHAC can also manage its installed air conditioners remotely, regardless of location, for an accurate view of their performance. Rather than implement maintenance in fixed intervals, the company can identify problems before they escalate, saving money on repair costs and call-out fees and extending product lifespans. The next step for JCHAC would be to aggregate all of its sensor data residing at the edge and run large-scale analytics and ML in the cloud. This would provide data teams with insights into not just regional edge environments, but also global performance.
3. Improving brownfield sites
Contrary to what some data leaders may believe, not all IIoT applications require the development of new greenfield facilities. In fact, updating older plants and legacy equipment can be enough in certain situations.
At its Czech Republic-based Rakona plant, which dates back to 1875, Procter & Gamble (P&G) produces four million cases of dishwashing soaps and fabric enhancers daily. After ultimately shifting to liquid products, the plant had to ramp up capacity, which required digitization and automation. To address the shortcomings of manual sampling and subsequent delays in product releases, the company eventually rolled out an in-process quality-control system in its legacy systems. Sensors now monitor product characteristics, allowing operators to obtain data that helps determine batch quality for release. Using sensors also lets workers stop the line if a deviation occurs. The results of using sensors are a 50 percent reduction in reworking and complaints, less scrap, fewer quality inspections, and a reduction in throughput time of 24 hours.
By expanding the capabilities of data-driven analytics in the cloud, manufacturers can identify and solve issues across a global big-data footprint.
The Rakona plant is just one example of P&G’s efforts to digitize manufacturing. Accessing data at scale is the next step in P&G’s digital transformation. In collaboration with Microsoft, P&G will leverage the Microsoft Azure cloud platform to digitize and integrate data from more than 100 manufacturing sites globally and will also enhance its AI, ML, and edge computing services for real-time visibility that gives P&G employees the ability to both analyze production data and use AI to make quick decisions that drive improvement and exponential impact.
With the enormous amounts of IIoT data at their disposal, manufacturers have the potential to access and act on that data at scale by moving edge datasets to the cloud for large-scale analytics and insights. The edge-to-cloud opportunity provides a number of benefits: reduction in plant operating costs; increased agility, such as the ability to keep up with new developments and technologies; and the ability to scale AI, ML, and data analytics globally in order to make smarter business decisions.
Chris Pemberton, VP of Marketing, WANdisco
Chris has deep expertise in modern B2B technology marketing and connecting brands with buyers through powerful messaging, content, demand gen, and digital experiences. He has led content strategy with Gartner and Charles Schwab as well as digital marketing and demand-gen programs for Persado AI, SANS Institute, and NextHealth Technologies. He delivered double-digit traffic, lead, and revenue growth for GU Energy Labs and Ghirardelli Chocolate Company. Chris spent years advising C-level technology executives while at the Corporate Executive Board (now Gartner) and holds an MBA from the Middlebury Institute of International Studies as well as certificates in advanced digital marketing analytics.