Manufacturing must have total clarity about cloud and the edge

by , ,

The manufacturing sector in Europe and the UK faces a set of problems few could have foreseen prior to the pandemic.

Russia’s invasion of Ukraine has inflicted serious disruption to global trade and caused the EU to slash its four per cent growth forecast for 2022, issued only a month before the tanks rolled in.

Even before the war, manufacturing was plagued by increasing inflation, skills shortages and seemingly intractable supply chain difficulties. In response to this onslaught, more dynamic manufacturers advanced digital transformation initiatives, aware they should become smarter and more efficient through greater adoption of cloud infrastructure. They realised the future for manufacturing lies in the implementation of cloud-based applications, mostly delivered “as-a-service”.

Read more: The growth of digitalisation with Industry 4.0

A poll by one of the hyperscale cloud providers last year showed how Covid spurred manufacturers towards the implementation of AI projects. The findings indicate plans to use AI and machine learning on a day-to-to-day basis to vastly improve supply chain and process optimisation, and boost quality control. Prescriptive analytics is set to transform maintenance. Smart manufacturing systems are also moving off the shop floor and into the cloud, avoiding the costly requirements of on-site control systems in cleanroom environments.

This hunger for the transformational advantages of cloud-based applications is why market intelligence company IDC predicts the discrete side of manufacturing will invest $19.7bn in cloud services over 12 months, part of the rapid growth of public cloud adoption which Gartner predicts will be worth $482bn in 2022.

More far-reaching Industry 4.0 initiatives and the industrial internet of things (IIoT), however, can only be fully implementable with the adoption of edge computing, which brings processing power closer to where manufacturing companies generate data. Edge is where the digital and the physical worlds intersect.

Read more: Pets versus cattle: cloud computing in manufacturing

For example, sensors that generate thousands of metrics from devices and machinery have become cheaper and easier to install. Yet, for a manufacturer to analyse and make use of the masses of data these sensors generate in real-time requires ultra-low latency. Companies can only achieve that through the adoption of edge computing, processing data at a data centre on the edge of the cloud with high-bandwidth 5G connectivity and on-ramps to the major public cloud vendors.

The increasingly obvious potential of edge computing is why IDC estimates that discrete and process manufacturing companies will invest $33.billion in this technology in 2022, led by eagerness for the IIoT, asset-optimisation solutions and advanced automation.

Optimising cloud and edge infrastructure

Yet as manufacturers move into cloud and edge environments, they will inevitably hit management and cost problems, which they can only overcome using smarter tools built for this next-generation infrastructure.

Rising cloud costs can be a significant headache, given the difficulty of predicting requirements and the complexity of most cloud vendors’ charging structures. Over-provisioning is common, as are unforeseen charges for transferring data. It is easy to overlook the cost of shipping high volumes of data to large data centres in the big-name cloud providers’ availability zones. These costs come with cloud deployments but get out of hand.

The Flexera 2021 State of the Cloud Report found, for example, that respondents waste approximately 30 per cent of their cloud expenditure. When companies misalign their traditional workloads in the cloud using a “lift-and-shift” method, the outputs can be expensive and ineffective.

Planning and discovery

Manufacturers moving into the cloud need to follow a planning and discovery methodology to identify the best use cases first as part of a transformation framework. A transformation framework will deliver continuous improvements and the ability to use new technologies.

Controlling costs boils down to strategy. Certain services are just not ready for the cloud, so it makes sense to prioritise efforts where the greatest value can be gained.

To get the most bang for their buck, manufacturers need to run their cloud deployments as close to their configured performance limit as possible without holding back applications. Yet organisations’ current cloud environments often lack meaningful visibility and reporting, which causes difficulties in designing services and forecasting their cost.

Hybrid cloud management

Without access to an effective, purpose-built management platform these optimisation complications multiply. This is all too common when a company uses more than one cloud service provider or mixes public and on-premises or private cloud environments in a hybrid approach.

Properly supervised and managed, however, a hybrid strategy will deliver many important gains in organisational agility, innovation and of course, cost. A hybrid approach allows businesses to prioritise workloads and choose a location that works best for their needs while avoiding dependence on a single cloud vendor. There is less likelihood of being held to ransom or being compelled to adapt to vendor-mandated changes in technology or operational practices.

A hybrid cloud strategy allows businesses to take full advantage of advanced AI or machine learning applications while maintaining older, more traditional workloads in a cost-effective and easy-to-use manner.

Embracing next-generation cloud management

To help manage this complexity of deployments, the major cloud providers provide their own tools, but their involvement in edge computing is limited. On the face of it, this seems a significant problem.

Fortunately, the edge providers are increasingly offering next-generation hybrid cloud management solutions. These tools act as a “single pane of glass” across hybrid and edge infrastructure. Enabling businesses to optimise the cost and use of assets wherever they are.

The benefits are significant. Businesses obtain a consistent, holistic view of their deployments and the details of their cost, enabling them to shift and adapt workloads for maximum operational efficiency and cost-effectiveness. Each application and edge “instance” is visible, as is real-time usage of cloud resources.

A manufacturer with a hybrid infrastructure can collaborate with its edge provider to assess all its workloads and provide comprehensive asset discovery, usage reporting and dependence mapping. This enables engineers to provide cost optimisation reports, identifying the most suitable environment for services.

Read more: What is Industry 5.0? Talking with Visual Components

From a networking perspective, organisations using a next-generation platform maximise the advantages of using edge data centres with a private network transport capability or network interconnect. This delivers lower data transit costs nationally and internationally, as required. IT managers can intelligently separate their traffic by placing the right workload in the right location, reducing the amount of data they transmit over long distances, shrinking costs and lowering latency. Businesses can see exactly where their monthly spending is going and with greater visibility, make better decisions based on accurate insights.

As all types of manufacturing rightly seek the immense transformational advantages of cloud migration, they must ensure that the complications of management and any lack of visibility do not undermine their ambitions. Once they have adopted the right planning and discovery strategy and can effectively manage hybrid cloud, edge deployments and networking, they will be far more confident about reaping the rewards of smarter manufacturing.

- The author, Simon Michie, is CTO, at Pulsant.

Back to Homepage

Back to Technology & Innovation