The intelligent manufacturing dream is crumbling — here’s why

Imagine a manufacturing ecosystem where every machine, robot and sensor is known, interconnected and working together in perfect harmony. Agentic AI orchestrates operations while digital twins simulate production environments and predictive analytics optimize inventory in real time.

This is the promise of intelligent manufacturing — but it’s slipping out of reach.

The reason? Most networks aren’t ready.

According to the Manufacturing Leadership Council, 68% of manufacturers believe AI will be foundational to their future competitiveness, with 61% planning to increase investments over the next two years. Meanwhile, global data creation is expected to surge to 394 zettabytes by 2028 — up from 149ZB in 2023 — and 75% of enterprise-generated data will be processed at the edge by 2025, compared with just 10% in 2024.

In this dynamic landscape, legacy networks — slow, fragmented and lacking visibility — have become liabilities. 

Consider the following statistics:

• $2.7 million: The average cost of a data breach in 2024
• 30–120 minutes: The typical duration of a network outage
• $5,600–$23,750 per minute: The typical cost of downtime

These aren’t hypothetical risks. They’re a threat right now.

• ALSO READ → Empower your workforce with AI in manufacturing operations

Industry 5.0: The new mandate for intelligent manufacturing networks  

Industry 5.0 marks a shift from automation alone to intelligent, human-centric collaboration. It’s about building resilient, adaptive systems that empower both people and machines to innovate together. This new era demands networks that are:

• Secure by design, not bolted on: Retrofitted security leaves gaps that adversaries will exploit.
• AI-native, not retrofitted: Intelligence must be embedded at every layer to enable predictive insights and autonomous responses.
• Edge-first, not centralized: Real-time decision-making must happen where operations happen, not miles away in a data center.
• Observability-rich, not opaque: You can’t protect what you can’t see.

Without these traits, manufacturers risk falling behind, unable to scale, secure or sustain operations in a world where agility and trust are non-negotiable. Industry 5.0 isn’t a vision of the future; it’s the new mandate.

The technology trifecta: Agentic AI, digital twins and predictive analytics

These three technologies form the backbone of Industry 5.0 — but only if your network can support them.

1. Agentic AI: The brain of the operation

Relying on little to no human interaction, agentic AI streamlines operations, takes specific actions and eliminates routine L1 tasks. It enables real-time decisions that reduce costs and boost efficiency.

2. Digital twins: Virtual clones of your assets

They simulate environments, predict failures, and optimize workflows—without disrupting production or creating downtime.  

3. Predictive analytics: The insight engine

It taps massive data lakes to forecast demand, inventory and maintenance needs — unlocking intelligence never before possible.

But here’s the catch: none of this works without a fast, autonomous and secure network — one that can process huge volumes of data from numerous devices and systems with extremely low latency and very high bandwidth. It requires moving away from old-school connectivity methods like multiprotocol label switching (MPLS) to more advanced solutions like private 5G, which offer lightning-fast speeds and minimal lag, perfect for real-time data processing and action.  

• ALSO READ → The new frontier: Why agentic AI should be on every CIO’s roadmap

Security and visibility: The non-negotiables

Networks must be built with zero trust, persona-based access, as well as full-stack observability. Anything less is a risk.

Legacy monitoring tools fall short in providing the network observability you need to detect and respond to threats in real time. Poor segmentation allows ransomware to jump from IT to OT, and a lack of visibility of encrypted traffic leaves blind spots.

• A zero trust security framework assumes no user or device inside or outside the network should be trusted by default. Instead, it continuously verifies every access request based on identity context and risk level.
• Persona-based access aligns user permissions with roles, responsibilities and risk profiles. Individuals access only what they need to perform their jobs.
• Full-stack observability provides real-time, end-to-end visibility across the entire IT and OT environment, from infrastructure and applications to user behavior and network traffic.

Together, these ensure that only the right people access the right resources, every action is monitored in real time and threats are stopped before they spread. Anything less leaves your network exposed.

The IT/OT divide: A growing threat

One major pain point for manufacturers is the intersection of IT and OT. When legacy OT environments, including your network, become outdated and unstable, it’s tough to make changes and achieve the agility you need to stay competitive. Modernization is crucial for seamless integration and smooth operations.

Furthermore, consistent security policies and visibility across both OT and IT are paramount, to prevent attacks like ransomware from jumping from IT networks to production systems. As long as IT and OT systems operate in silos, achieving unified security and visibility will remain extremely challenging, if not impossible.

To close the IT/OT divide, manufacturers must modernize legacy systems, unify security policies and implement shared visibility across both domains.

Measuring ROI: Beyond SLAs

How will you know if your investment in network modernization is paying off? Traditionally, SLAs have been the go-to for ensuring network reliability. Outcome-based metrics, however, guarantee specific performance and availability outcomes at each manufacturing site. This shift means that the network is both meeting basic standards and actively contributing to your business goals.

To measure your ROI, focus on key performance indicators (KPIs) that include:

• Network availability and performance: How consistently is your network up and running, and how well does it perform under load? High availability and reliable performance are critical to keeping you operating without costly interruptions.
• Network visibility: How well can you see and understand what’s happening across your network, including devices, applications and traffic patterns? Improved visibility helps you detect anomalies faster, contextualize alerts and optimize performance for both human and machine users.
• Security-incident response times: How quickly can your team detect, assess and respond to security threats? Faster response times reduce downtime, limit damage and strengthen overall resilience.
• Operational efficiency: Is your network helping to streamline workflows and eliminate bottlenecks? This KPI reflects how effectively your infrastructure supports your day-to-day operations and continuous improvement.
• Cost savings from reduced security breaches: What financial impact are you seeing from fewer breaches and better threat prevention? Lower incident rates translate directly into reduced recovery costs and improved ROI.

These KPIs show whether your network is enabling or hindering your intelligent manufacturing goals.

Real-world impact

At NTT DATA, we’re already working with manufacturers around the world to prepare their networks for intelligent new technologies.  

• Liantis, an entrepreneurial HR company in Belgium, implemented a managed SASE service to simplify their network management and security, allowing them to move to a more secure and future-proof network.
• A major automaker adopted private 5G networks globally to protect against connectivity disruptions that could lead to a shutdown of their manufacturing lines — and significant revenue loss.
• And a multinational soft-drinks manufacturer transformed their wide area network architecture to a more software-defined model, integrating zero trust network access to improve security. These changes support their manufacturing processes and deliver better insights and intelligence, all while minimizing downtime.

As these manufacturers have realized, the future of intelligent manufacturing is bright — but without the right network infrastructure, it will remain nothing but a dream.

Network services that cover far more than connectivity

For manufacturers, managed network services are an ideal route to operational efficiency and stronger security. They also give you access to the latest skills and best practices — a key consideration at a time when networks and IT estates are becoming increasingly complex, with distributed locations for data, applications, users and other entities.

NTT DATA’s holistic approach to networking starts with consulting and continues with ongoing service improvement. Through a discovery process, we understand the current state of your network and identify what needs to change. This includes assessing the fitness of your existing applications and infrastructure for the new network environment.

Our focus on industry-specific architectures, including through our Smart AI Agent™ Ecosystem, simplifies the adoption of agentic AI and other advanced technologies. By providing preconfigured solutions and best practices, we reduce the time and effort required to implement these technologies in your organization, ultimately leading to faster and more widespread adoption.

We don’t just connect manufacturing environments. We transform them — and our comprehensive Managed Network Services hold the key to this future. Let us help you innovate, work more efficiently and stay competitive.

• ALSO READ → The invisible backbone: Why your workplace is only as good as its network