Five edge computing use cases for the manufacturing industry
The manufacturing industry has been touted as one of the early adopters of edge computing with many potential use cases. Arguably, edge computing has existed in the sector for a number of years: manufacturing plants have significant processing power on-premises, whether it be in programmable logic controllers (PLCs), the machines themselves, or an on-premise data centre.
The need to be more flexible and cost-effective in how manufacturers run their plants, plus pressure on their core business due to global competition, means that there is considerable effort around digital transformation, aka Industry 4.0.
Edge computing fits into this wider context by allowing manufacturers to use more flexible, standard hardware and software to be able to access and share data relevant to their manufacturing processes.
Below we outline five use cases that will drive the use of edge in the manufacturing industry.
1. Condition-based monitoring
Manufacturers face a challenge in simply trying to access data from their machines, processes and systems. Plants have historically been built using many proprietary systems, which do not talk to one another. Operational technology is still quite traditional and does not yet use the same standards as IT – hence the need for IT/OT convergence. One of the challenges to extracting data from all these machines in the factory is that it results in huge amounts of raw data which would overload a central server. Edge computing allows manufacturers to filter data to reduce the amount sent to a central server, either on site or in a cloud.
The ability to monitor the condition of their assets remotely helps manufacturers generate new revenue streams. Rather than selling machines one-off to their end-customers, they can provide services. For example, maintenance services based on the actual condition of the asset, or even dramatically changing their business models so the customer pays a managed service for uptime.
2. Predictive maintenance
Predictive maintenance refers to being able to pre-emptively detect when a machine will fail – through data analytics – and mitigate this by conducting maintenance in advance of potential breakdown.
Although the term predictive maintenance has been in the industry for some time, the reality is that manufacturers have found it very difficult to implement. Partly this is because there have been challenges integrating the insights from operational technology into IT systems (e.g. ERP systems). Other issues stem from the inability to predict outcomes effectively because there are not enough variables being measured and the machine learning platforms are not mature enough to produce real insights.
Like condition-based monitoring, edge computing helps to process data closer to the end-device, avoiding the cost of transporting data to a remote cloud as well as ensuring data is accessed reliably. Predictive maintenance requires even more data in order to be implemented well; a problem can only be predicted if there are parameters considered.
One of the manufacturing sector’s challenges is being able to leverage the benefits of economies of scale from highly standardised, automated processes. Another is having a manufacturing process that is flexible enough to meet changing demands from customers.
Manufacturing can be made more flexible and mobile by reducing the time it takes to set a site up as well as creating more sharing models where multiple parties can use the same facility. Both these things have a clear edge cloud use case. First, systems need to be available no matter where the site is (cloud-like) and still meet stringent latency requirements, given that they are mission-critical for running the manufacturing process. Second, processing data at the edge overcomes manufacturers’ concerns regarding data security concerns.
Another aspect of this is being able to “pop up” the network infrastructure required for a potentially temporary site. This could be enabled via virtual private LTE or 5G networks, running on edge compute infrastructure. AWS sees this as a key use case for its Outposts offering.
4. AR/VR in the manufacturing plant
There are many ways in which manufacturers could use augmented/mixed/virtual reality in the plant, whether it be to train employees on how to use equipment or new processes; for health and safety (to guide a worker through a hazardous environment); assisting a maintenance and repair worker with remote expertise; or detecting product faults during quality inspections.
The challenge with using VR headsets is that they are heavy and/or unable to process significant amounts of data, which make them impractical for the scenarios highlighted above. However, taking processing off the device and into the cloud results in too much latency and can sometimes make the wearer feel nauseous despite the lag being under 100 milliseconds.
Processing the data and rendering the stream from an edge compute node – either on-site or on a network edge – eliminates this problem and makes the headsets lighter and therefore more user friendly.
5. Precision monitoring and control
A key Industry 4.0 goal is to be able to use data from multiple machines, processes and systems to adapt the manufacturing process in real-time. This precision monitoring and control of manufacturing assets and processes uses huge amounts of data and needs machine learning (ML) to determine the best action as a result of the insight from the data.
Edge computing is not only relevant in collecting, aggregating and filtering the data in order to send outcomes to a central server, but it will be critical for AI/ML. In some cases, the edge will be used to train a ML algorithm, as well as executing it. Given the amount of processing required for AI/ML, a manufacturer may choose to distribute processing across multiple processors, i.e. edges, rather than do this in the cloud.
More research on the roles of 5G, IoT and edge computing for the manufacturing industry can be found on our Enterprise Stream. We discuss this in detail in two of our recent reports:
About Dalia Adib
Edge computing practice lead
Dalia is the Edge Computing Practice Lead at STL Partners and has led major consulting projects with Tier-1 operators in Europe and Asia Pacific on edge computing strategies, use cases and commercial models. She co-authored the research report “Edge Computing: Five Viable Business Models” and been an active speaker at events including Edge Europe and Data Cloud Congress. Outside of edge computing, she supports clients in areas such as 5G, blockchain, digital transformation and IoT.
Read more about Edge Compute
About edge compute and edge cloud
An overview of edge computing and edge cloud to highlight the key questions being asked by the wider ecosystem and telecoms operators who are exploring the opportunity
Turning vision into practice
Our Telco edge computing: Turning vision into practice research gives an overview of the telco opportunity and seeks to address the key challenges for operators pursuing edge
Edge business models and how to execute them
A joint webinar with MobiledgeX and STL Partners exploring edge cloud business models and the value proposition for application developers in augmented reality