MES integration lags, AI gains stalled by data silos
The gist
Despite near-universal MES deployment in manufacturing, stubborn data silos and weak integration are stalling AI’s promised revolution—and leaving companies exposed to new risks.
What to know
- Only 23% of manufacturers fully integrate their MES, limiting AI’s impact and keeping operations fragmented, according to Rockwell Automation’s 2026 report.
- Cybersecurity is now a boardroom priority as networked robotics and cloud-connected systems expand the attack surface, driving stricter data governance and compliance with regulations like UK GDPR and the EU AI Act.
- While giants like Hitachi, Nvidia, Siemens, and ASML showcase real industrial AI at scale, most manufacturers must still overcome cultural, workflow, and data ownership hurdles to unlock true productivity gains.
MES Integration’s Hidden Hurdles
Manufacturers remain trapped in fragmented workflows and data silos as cultural resistance and legacy systems stall the leap from MES deployment to true AI-powered transformation.
Despite an impressive 93% of manufacturers deploying Manufacturing Execution Systems (MES), full integration remains a significant hurdle, with only 23% achieving complete system integration and just 28% scaling MES enterprise-wide, according to Rockwell Automation's 2026 findings. This gap severely limits the realization of AI’s transformative potential in manufacturing workflows, as fragmented systems and siloed data prevent seamless AI application and diminish operational gains.
The root of AI readiness challenges in MES extends beyond technology to deep organizational and cultural barriers. Industry analyses reveal that many manufacturers struggle with governance, workflow redesign, and change management, which are critical for moving from mere systems integration to true business integration. As Rockwell Automation’s VP Anthony Murphy and other experts emphasize, without clarity, ownership, and cross-departmental communication, AI adoption risks becoming counterproductive rather than beneficial.
Legacy system incompatibilities and data silos continue to trap manufacturing data across rigid ERP and MES platforms, local spreadsheets, and unstructured documents, leaving nearly 62% of manufacturers stuck in 'pilot purgatory' as reported in 2026 industry analyses. Overcoming these integration challenges requires a shift from passive data visibility to active orchestration through governed AI agents layered on robust Manufacturing Operations Management (MOM) foundations, enabling seamless IT/OT convergence without risking core system stability.
Effective AI integration in MES demands a holistic approach that addresses human variability, continuous improvement, and collaborative redesign of manufacturing processes. Leaders must empower teams with autonomy and time to analyze data strategically rather than firefighting daily issues, while partnering with industry experts to reimagine workflows. This multi-faceted strategy is essential to unlock AI’s full benefits, as emphasized by manufacturing thought leaders who stress that digitizing engineering and production documentation is foundational to reducing errors and accelerating development cycles.
Cybersecurity: The AI Dealbreaker
With AI-powered robotics expanding attack surfaces, manufacturers are embedding cybersecurity and data governance at the core of every new investment to avoid costly breaches and regulatory fallout.
As AI integration within manufacturing execution systems (MES) advances, cybersecurity has rapidly ascended to a top buying priority, driven by the growing recognition of rising cyber risks. Rockwell Automation’s VP Anthony Murphy highlights that while 93% of manufacturers use MES, only 23% fully integrate it, a gap that limits AI benefits and underscores cybersecurity’s critical role in enabling trustworthy AI and robust data governance. This shift reflects a broader industry consensus that without strong cybersecurity foundations, AI readiness and deployment remain vulnerable and constrained.
The expansion of AI-powered, networked robotics and cloud-connected historians in manufacturing environments has significantly broadened the attack surface, making cybersecurity a critical operational dependency. ABB Robotics’ recent launch of an AI-powered visual platform exemplifies how manufacturers are prioritizing physical AI alongside stringent data governance to safeguard operations. Industry voices emphasize that data governance is not merely an afterthought but a foundational investment, best addressed during capital equipment purchases or system integrator engagements to avoid unreliable AI outputs and costly retrofits.
In the UK, cybersecurity’s prominence in manufacturing AI deployments is underscored by the steep financial consequences of data breaches, averaging £4.14 million per incident, and nearly £4 billion ringfenced in the 2025 Industrial Strategy for AI, cybersecurity, and advanced connectivity R&I. Compliance with governance frameworks such as UK GDPR and the EU AI Act is becoming integral to trustworthy AI, with boards focusing on audit-ready data lineage, model governance, and human-in-the-loop checkpoints from project inception. This regulatory rigor reflects a maturing approach to AI governance that balances innovation with risk mitigation.
Industrial AI: From Pilot to Scale
Industry leaders like Siemens, ASML, Hitachi, and Nvidia prove that quality data, skilled teams, and clear ownership are the real drivers behind moving AI from isolated pilots to enterprise-wide productivity gains.
At IMTS 2026, ARM Hub showcased how industrial AI is not just a futuristic concept but a practical catalyst for boosting productivity, particularly in Australian manufacturing where trusted data, incremental implementation, clear ownership, and focused management have proven essential. Siemens and ASML exemplify this success, underscoring that quality data and skilled teams are non-negotiable foundations for effective AI integration in manufacturing execution systems.
Hitachi and Nvidia's expansion of the HMAX platform with multi-agent AI capabilities marks a significant leap in industrial automation, particularly in the rail sector where this technology now scales across more than 2,000 trains. This advancement highlights a tangible evolution from isolated AI experiments to robust, scalable workflow enhancements that directly improve operational efficiency and demonstrate industry leadership in AI-driven manufacturing solutions.