The Business Value of Unified Namespace for Industry 4.0

Welcome to Part 1 of the blog series, An Advanced Guide to Building UNS for IIoT: Beyond the Basics. It is no curious coincidence that digital-native companies like Amazon and Tesla are leading the way today. They have embedded data-driven experimentation into every aspect of their business, embraced continuous change based on rapid learning, and leveraged machine intelligence to transition from intuition-based decisions to actions driven by observations and data predictions. The principles of successful digital transformation are universal; if manufacturers are to succeed with digital transformation, they also need to make data the most essential commodity in their businesses.

At the heart of successful manufacturing digital transformation lie three key factors: the speed at which value can be derived from data, the cost-effectiveness of this process, and the flexibility to scale data use cases across multiple factories. In essence, the quicker, more cost-effective, and seamless your data-driven ideas can be brought to market—both internally and externally—validated, iterated, and improved, the more competitive and high-performing your business can become.

However, successful data-centric business operations are not a matter of chance. Manufacturers need a well-thought-out data strategy, architecture, and operating model that empowers them to scale and accelerate experimentation, learning, and adaptation using data. The Unified Namespace (UNS) is an emerging architecture that brings this strategy to life, as evidenced by its adoption by companies across the manufacturing spectrum, from Fortune 500 corporations to SMEs.

In this article, I’ll explore how UNS enables manufacturers to generate business value from data at scale. We’ll start by discussing the tenets that are crucial for manufacturers to generate business value from data at scale, explore why traditional approaches to industrial data integration fail to meet these demands, and then discuss how Unified Namespace significantly increases the ratio of value from data to investment.

Key Tenets Generating Business Value from IIoT Data

Data Integration

I recently watched an interview with Bill McDermott, CEO of ServiceNow, who stated that 85% of digital transformation projects fail to deliver a return on investment due to one main issue: integration. This is especially true in manufacturing, where decades of siloed solutions have led to fragmentation that needs addressing to unlock the value of data. As discussed in the next section, traditional data integration methods have failed because companies' data aspirations quickly outpace their execution capabilities, causing their data investment returns to stagnate.

To realize business value from digital transformation, manufacturers need a data integration model designed to support dynamic and evolving data flows. Organizations operate within complex supply chains and ever-changing production environments, and their systems must handle this complexity. The integration system should scale vertically to manage large information models and horizontally to accommodate numerous data points and connections. It should prevent the formation of data silos while acknowledging the complex topologies within which it operates. Therefore, the ability to transparently access required data locally or remotely is crucial.

Data Accessibility in Industry 4.0

A decade ago, manufacturers primarily focused their data aspirations on business intelligence (BI). They sought to generate reports and dashboards to manage operational risks, comply with regulations, and make informed business decisions, albeit at a slower pace. 

Despite substantial data and analytics infrastructure investments, results plateaued because today's data needs have evolved. Data aspirations now extend beyond BI to encompass every aspect of an organization, such as using machine learning in product design, enhancing customer experiences through personalization, and optimizing real-time logistics. Moreover, to help organizations derive value from data more cost-effectively, there is an expectation to democratize data to a broader population of generalist technologists to become information workers. 

Meeting these expectations requires a new approach to data management that seamlessly supports diverse data usage. This diversity demands various data access modes, from simple structured views for reporting to continuously reshaping semi-structured data for machine learning and from real-time, fine-grained event access to batch aggregations.

The Importance of Data Quality in IIoT

Improving data quality is one of the most crucial factors for reducing delivery times and total costs in Advanced Analytics. According to a recent MIT Technology Review survey, 57% of manufacturing executives said that AI use-cases are hampered by inadequate data quality. Only about one in five manufacturers surveyed have production assets with data ready for use in existing AI models.

Data quality is essential for successful AI initiatives and can significantly increase revenue. High-quality, curated datasets not only make more AI projects feasible by reducing delivery times and costs, but they can also be monetized multiple times. Every dollar invested in data quality generates compounding returns across various datasets and use cases.

Moreover, data quality is not solely an AI concern. Poor data quality affects reporting, operational insights, dashboards, descriptive analytics, diagnostic analytics, daily decision-making, and all other forms of analytics. 

Challenges with the Traditional Approach to Digital Transformation

A recent McKinsey report found that 74% of digital transformation projects in manufacturing are stuck in the pilot phase. The reason is apparent: many manufacturers treat digital transformation as a series of isolated, siloed projects. This approach might work during the pilot phase, where integration with other systems is less critical and new and existing methods can run in parallel. However, to achieve success at scale, a comprehensive digitalization strategy is needed to seamlessly integrate with the organization’s broader IT and operational technology (OT) infrastructure. This is where conventional use-case-based approaches fail.

In the traditional approach to digital transformation, each use case operates with its own connectivity, data ingestion, storage, and contextualization processes. Because so many different projects fall under the digital transformation umbrella, manufacturers often piece together numerous solutions and applications—such as machine learning solutions, digital twin technologies, and quality monitoring systems—without a cohesive data management strategy. 

This approach to digital transformation in manufacturing creates tightly coupled point-to-point integrations, leading to significant data infrastructure replication. This redundancy in data ingestion and collection makes sustaining such a digital infrastructure prohibitively expensive. Each new use case requires substantial investment in configuring systems to collect and store data, often duplicating efforts for the same data. This severely limits the agility of manufacturing companies, resulting in numerous pilot applications, questionable ROI, and an inability to scale effectively.

Moreover, the market is flooded with innovative IoT, robotics, AI, and analytics startups offering niche solutions for complex operational challenges. This increases factory and supply chain complexity, making data more challenging to access and use and stifling innovation. Vendors tend to drive value from complexity and closed infrastructures, not simplicity and operational value, leading to the proliferation of new cloud platform technologies that trap data and worsen this complexity. These complexities also heighten security and operational risks, further slowing implementation and scaling.

The lack of consolidation complicates projects and increases costs. As manufacturing organizations become more digitally integrated, they create numerous custom point-to-point interfaces between applications, leading to duplicated efforts and escalating budgets. A project that initially needed only a few connectors and applications can rapidly escalate to needing many more, causing a significant increase in costs.

Establishing cost-effective and seamless data access at scale is challenging, but deriving value from this data is even more so. With traditional approaches, data from across the manufacturing business is typically dumped, without context, into data lakes and warehouses. Only specialists with data science expertise can use this data, often with long lead times from data to insights. This bottleneck hinders scaling data use across diverse use cases, delaying time-to-market for innovation and preventing the democratization of data use across the organization, thereby reducing the ROI on data investments.

The core issue is that a 30-year history of complex, fragmented solutions already burdens manufacturing systems, and conventional digital transformation approaches only exacerbate these data integration complexities. As manufacturers increase their spending on digital transformation and AI in the coming years, failing to address these issues will likely limit the returns on those investments. Manufacturers need robust data foundations to support their data and AI ambitions effectively.

The Value of Unifying Data and Information for OT/IT Interoperability 

As highlighted in the first section, manufacturers increasingly recognize that they need a modernized data architecture to maximize the value of their data investments at scale. This architecture must effectively unify data across both OT and IT domains within their organizations, providing intuitive and seamless access to high-quality data for everyone who needs it to make informed and timely decisions.

Such an approach is essential for maintaining agility, especially as data-generating sources grow and use cases expand from basic business reports and trend analyses to advanced analytics and intelligent decision-making across all business functions. This realization has led many companies to adopt the Unified Namespace (UNS). 

A new industrial IT-OT architecture is emerging to improve data accessibility and reduce asset vulnerability. 

It will be built upon an event-centric integration pattern that will use MQTT brokers, the Sparkplug B standard, and a Unified Namespace (UNS) design strategy.  – Reference Architecture for Integrating OT and Modern IT, Gartner

The Unified Namespace is a modern OT-IT data integration architecture tailored for scalable digital transformation use cases. It provides a centralized and common data infrastructure that contextualizes, normalizes, standardizes, and unifies data from all business units, accessible through a single interface using an MQTT platform.

The Unified Namespace (UNS) effectively removes data silos by providing a centralized platform where all data sources within a manufacturing ecosystem can be integrated, eliminating the complexity that results from the point-to-point use-case-based data integration approach, as described in the previous section. With UNS, the focus is on building a data management foundation on top of which use cases across design, engineering, production and supply chain can be addressed.

Data from various business units is integrated using MQTT and organized into a semantic data structure that mirrors the organization’s information model. This structure allows intuitive and repeated access to the data for diverse use cases. OT and IT data sources and their respective data objects and events are defined once within the unified platform. This eliminates the need for repetitive data integration efforts, enabling the defined data to be reused across multiple applications and systems. As a result, redundancy and inefficiencies associated with traditional data handling approaches are significantly reduced.

Once the UNS data infrastructure is in place, the same pool of information can be leveraged to move swiftly from one use case to another, enabling rapid deployment and scalability of solutions. A data use case idea that would typically take months from conception to implementation can be implemented within days to weeks. Moreover, organizations can seamlessly scale from small pilot projects to full-scale implementations. This efficiency saves time and resources and ensures consistency and accuracy in data usage.

The ability to reuse data is a fundamental advantage of implementing a Unified Namespace. It ensures that once data is collected and integrated, it can be leveraged across multiple applications and processes, maximizing its value and utility. The initial investment in data integration yields ongoing benefits, as the same data can support multiple use cases, driving continuous value from the original work.

Moreover, data objects defined and populated within the UNS for simple use cases like dashboard visualization and reporting become reusable assets; they can act as building blocks that can be transformed and combined to power more sophisticated and data-intensive applications such as predictive maintenance, quality control, and supply chain optimization.

Furthermore, by democratizing access to high-quality data, the UNS empowers citizen developers—employees who create applications and solutions without specialized data-handling skills. With a unified data platform, these individuals can leverage intuitive tools and pre-defined data models to develop custom applications that address specific business needs. This democratization accelerates innovation and problem-solving within the organization, as more employees can contribute to the development and optimization of processes, significantly increasing the ratio of value from data to investment.

In addition, UNS fosters a data-driven innovation culture by making data a central component of decision-making processes. When data is readily accessible and integrated, it encourages experimentation and the continuous improvement of processes. Teams can use data to test hypotheses, uncover insights, and implement innovative solutions that drive competitive advantage. This culture of innovation is essential for staying ahead in the modern manufacturing landscape.

Finally, the Unified Namespace enables the automation of business processes by providing a comprehensive, real-time view of all manufacturing operations. Based on current data, this holistic view allows for integrating advanced planning and scheduling algorithms to adjust production and shipping schedules dynamically, etc. This capability enhances responsiveness to changes in demand or production conditions, crucial for manufacturers who now view volatility and change as constant factors, especially in the wake of disruptions caused by the Covid-19 pandemic.

Conclusion

The journey towards successful digital transformation in manufacturing hinges on the ability to harness the full potential of data, and UNS provides the framework to achieve this vision, ensuring that manufacturers are well-equipped to navigate the complexities and opportunities of the modern industrial landscape. Embracing UNS and its principles positions manufacturers to thrive in the digital era, driving competitive advantage through enhanced responsiveness, efficiency, and innovation. 

Other Blogs from the Series 

Part 2: Why a Data Warehouse Can’t Be the Unified Namespace

Part 3: UNS Semantic Data Hierarchy with MQTT: Explained with an Example

Part 4: Data Modeling for The Unified Namespace: Best Practices

Part 5: Automating Manufacturing Business Processes with the Unified Namespace