Four key ways manufacturers should scale cloud deployments

Manufacturers have faced unprecedented demands to adapt rapidly over the past couple years—from pandemic-driven spikes and drops in market demand to the unpredictable supply chains and shortages of raw materials and components that continue today. In response, even more traditional hold-outs among manufacturing firms have embraced Industry 4.0, with its focus on automation and data-driven insights, to gain the agility required to not just survive but also capitalize on new opportunities as market conditions change.

As they incorporate Industry 4.0 best practices, these manufacturing companies are doubling down on their use of Internet of Things (IoT) technologies—buying new smart machinery and retrofitting old machines with sensors—while increasing their use of artificial intelligence (AI), machine learning (ML), and other advanced software technologies. They are also increasing their integrations internally and externally with suppliers and customers to streamline processes, primarily through the use of APIs. More recently, they have accelerated their adoption of cloud computing.

The drive toward greater cloud adoption

As a whole, the manufacturing industry has been slower than other sectors to migrate to the cloud, but that is changing as more businesses shift all or part of their operations to the cloud. For packaged applications, such as enterprise resource planning (ERP), manufacturers increasingly have options for moving to a software as a service (SaaS) or managed cloud solution. Meanwhile, in-house legacy applications are being “lifted and shifted” to infrastructure as a service (IaaS).

By utilizing cloud services, manufacturers are able to realize one of its most lucrative benefits: aggregated manufacturing data that combines multiple data sources for a single point of analysis. This cloud-based, aggregated information can auto-scale based on the demand to create high availability of data, processes and applications. Organizations can create valuable insights without manually comparing manufacturing sources because data stored in the cloud and built-in data-processing services have already done the work.

Moreover, data collection becomes more nuanced as the cloud brings together systems of record for transactional, reference, and observability data. By using the latest hybrid tools to aggregate the analysis of data from local plants to central data lakes residing in the cloud, manufacturers can obtain the insights they need to keep growth consistent.

The robust security that comes with cloud-based solutions has also become a critical benefit following the rapid growth in cyber-attacks against manufacturers, which became highly visible in 2021 and are only expected to increase going forward. Many manufacturing firms are realizing that they do not have the technical talent in-house to keep pace with security demands, and they are turning to cloud solutions providers who can deliver this protection for them.

Collectively, these benefits are incentivizing manufacturers to adopt cloud-native best practices for all their internal IT and external customer and supply chain needs to stay ahead of the curve. Now, let’s look at how manufacturing firms can further maximize their potential for effective digital transformation and their financial performance by exploring how cloud-based computing supports four key areas: zero-trust environments, IoT device expansion, data policy standardization, and globalized AI and ML heuristics.

Transformation with protection: Establishing zero-trust environments for operations and data

Manufacturers need to secure their data, users and all dependent systems using identity and access management (IAM) principles. This means that before harvesting data from the frontlines on a micro-scale, manufacturers must tighten their security to protect all data movement across, into, and out of the business.

Application-level and infrastructure-level security provide the foundation for a zero-trust environment. Cloud platforms enable these two levels of security, minimizing demands on the manufacturer’s IT team. To extend the core security capabilities of the cloud, the organization can leverage identity-as-a-service solutions (iDaaS) to leverage single-sign-on (SSO), multi-factor authentication (MFA), identity federation, adaptive authentication, and many more fundamental security standards.

APIs have become the glue for sharing information within manufacturing organizations as well as with their suppliers, distribution and retail partners, and customers. Not only do APIs secure the communication channels; they also help to standardize interactions. In this way, APIs secure the edge and enrich the backend security. Moreover, cloud-native technology is API-centric by default. Therefore, cloud infrastructures provide a significant level of functionality for implementing an API practice across different business entities in the manufacturing industry.

Finally, for those manufacturers with in-house application development teams that seek to automate the entire lifecycle of IT delivery with continuous integration and delivery (CI/CD), utilizing cloud infrastructures, such as GitHub and processes such as GitOps, will secure and govern the deployment pipelines in the cloud.

IoT device expansion—connecting the cloud to the machine

Many manufacturers were using sensors well before the term “IoT” became popular. However, the data collected was primarily used for historical analysis to support statistical process control (SPC) and understand overall equipment effectiveness (OEE). Increasingly, this data—whether coming from sensors or smart machines—is being collected via real-time process and product monitoring software and then analyzed immediately to alert employees on the shop floor and in the back office about potential issues.

The sheer volume of data being captured has grown exponentially. More manufacturers are adding IoT devices and replacing older machinery with smart, energy efficient machines that enable monitoring and control of processes by automatically capturing information about their own health, status, and environment. At the same time, many of these companies are seeking to maximize their productivity by running lights-out or near-lights out shifts—or monitor multiple facilities from a single control center—by leveraging process and product monitoring to alert managers or supervisors when human intervention is needed to take preventive or corrective action.

The cloud provides a central location with the scale needed to store data from IoT devices, smart machines and sensors. Cloud-based data lakes can tap into the local manufacturing plant and collect IoT data on a granular level to understand if, for example, temperatures or component dimensions are within acceptable ranges—or if they are starting to trend toward moving outside of acceptable parameters. The cloud-based data can also be aggregated across production runs and different plants to understand and compare operations across a manufacturer’s different facilities. At the same time, the low latency of cloud platforms supports the ability to gain the real-time or near-real-time insights needed to keep operations on track.

Standardizing data-policy compliance

Manufacturers face industry and government regulations around product quality and safety in practically every sector. Now, as more companies conduct business globally, it has become critical to comply with regional government mandates on how to protect data and ensure the privacy of data from consumers and other entities.

Perhaps, the best example is the European General Data Protection Regulation (GDPR) mandating how online data is used. GDPR regulations are among the strictest measures to address, and enforcement of these mandates comes at a cost to those who don't comply—starting with fees and various penalties that quickly stack up. Many businesses with international customers have chosen a simple fix: global standardization of policies across all offices.

The International Association of Privacy Professionals (IAPP) reported that nearly half of respondents in their October 2021 survey use a single global privacy strategy, aiming to simplify the compliance process across offices. For manufacturers, uniform policies can help generate proactive solutions for facilities worldwide that operate under one organizational umbrella. The benefits go beyond regulatory compliance—they ensure that applications can communicate and change seamlessly, without worrying that data from an unregulated application will make a regulated region's data noncompliant.

The cloud enables unified governance and policy enforcement across the data and applications deployed. While a cloud data plane allows the movement of data and interactions, a control plane permits organizations to administer and apply policies, a fundamental feature of a cloud-native architecture.

Deploying AI/ML heuristics in data analysis

As noted earlier, it is helpful for teams to parse through aggregated data. However, the mountain of aggregated data being collected—not only from IoT devices but also ERP and other manufacturing systems—makes it difficult for manufacturers to analyze and understand.

Successful organizations are cutting the red tape by using AI/ML heuristics to generate a machine-developed analysis for more robust and faster data reviews. With these heuristics, management and analytics teams can move quicker through data analysis and provide critical insights for the manufacturing business.

The models help guide AI programs to progressively learn how to compute and recognize the data it collects and analyze it for teams to dissect later on quickly. Through these resources, the programs can know enough to begin building predictive models, anticipating the future performance of local machinery, workforce, and overall business directions.

The auto-scaling and high compute power available in cloud infrastructure can run the large data models used in AI and ML processes. Some of these processes can be long-running and require a variable resource plan to fulfill some data-processing needs. At the same time, there is a rich set of AI/ML functionality provided by cloud service providers, which data scientists in the manufacturing industry can utilize.

Conclusion

In today’s era of continued globalization and rapid change, manufacturers are accelerating their digital transformation and adoption of Industry 4.0 best practices. Increasingly, the cloud is providing a central pillar around which manufacturing firms can optimize their operations, security, policy enforcement, and decision-making. In doing so, they are gaining the agility to adapt quickly to new market conditions while driving growth and profitability.

By Asanka Abeysinghe, CTE, WSO2