where downtime could cost millions per day. This is just one example of the critical nature of connectivity. Industrial machines can produce gigabytes of data daily, with a modern smart factory potentially relying on hundreds of such systems. For manufacturers, having network systems that can scale and adapt to meet these requirements is necessary.
If data cannot move quickly enough to support predictive maintenance, real-time analytics or safety monitoring, operations grind to a halt. The ability to sustain these capabilities ultimately depends on the resilience of the underlying network.
The cloud is evolving and so are its demands
As with many sectors, the initial wave of cloud adoption saw mass enthusiasm for offloading as much as possible to public providers. However, this approach has since matured, with manufacturers adopting multi cloud and hybrid strategies, weighing up which workloads belong in the cloud and which should remain closer to operations.
According to a recent Cloud Trends report, organisations are shifting away from public cloud, often due to cost, performance or sovereignty issues. In Europe, GDPR and initiatives like GAIA-X amplify the need for sovereignty-aware strategies.
A growing trend is the rise of so-called“ AI factories” at the edge. Siemens and NVIDIA, for example, expanded their partnership recently to build“ industrial metaverse” digital twin factories, including a FREYR Gigafactory project. These efforts demand deterministic, low-latency networking; if synchronisation lags, the digital twin loses predictive value.
This is driving repatriation of certain workloads back into private or on-premise environments, mainly citing their reasons as cost, control or security. All these factors create a more complex IT ecosystem that demands seamless connectivity between onpremises systems, edge deployments and the cloud environment chosen by the manufacturer. Many legacy networks weren’ t initially designed to handle this level of interoperability.
Incompatibility, latency, and bandwidth limitations all present risks that could undermine even the most carefully laid out plans.
The infrastructure gap is a competitive risk
While discussions around manufacturing innovation often focus on robotics or AI, the less visible infrastructure challenge carries just as much weight. Without the right digital backbone, the potential of these technologies cannot be realised.
In 2025 alone, UK and European manufacturers are projected to lose more than £ 80 billion due to downtime, according to research from IDS-INDATA. In an environment where margins are squeezed by energy volatility and supply chain risk, every minute of downtime is EBITDA at risk.
AI-driven processes, automated quality control, and robotics-enabled production lines all share a common dependency: a reliable and secure network. If the network falters individual processes or even entire operations are jeopardised. It is not tools which pose the biggest challenge, but the invisible fabric that supports them.
This is why the gap between infrastructure and capability is more than a technical issue, but a competitive risk. Manufacturers that are unable to modernise their connectivity will inevitably find themselves at a disadvantage compared to those that can. They may face longer downtime and missed opportunities, losing any competitive advantage despite large scale initiatives to gain one.
Closing the gap
The key to modern manufacturing transformation is connectivity. Manufacturers must think beyond applications and cloud strategies and address the fundamental role of network infrastructure in enabling any transformation initiative. Investment in low-latency, high-bandwidth connectivity is the foundation of performance.
Manufacturers and Managed Service Providers( MSPs) can work together to design systems that can meet the ever-increasing demands of smart factories. System requirements, levels of security and scalability must all be considered from the outset, rather than treated as afterthoughts.
The next phase of digital transformation in manufacturing will not simply be about adopting more AI or shifting more workloads to the cloud. It will be about creating an infrastructure that can sustain these advances by removing the bandwidth bottlenecks that hold progress back. Those who recognise this reality and act accordingly will be best placed to harness the full potential of digital technologies, while those who ignore it risk being left behind.
www. gtt. net
AUTOMATION, CONTROL & ENGINEERING 7