By Rebecca Gibson |
Manufacturing organisations integrating AI technology into their operations will gain a significant competitive advantage over those that do not, according to 93 per cent of senior manufacturing AI leaders surveyed in KPMG International’s Intelligent Manufacturing – A BluePrint for creating value through AI Driven transformation report.
KMPG found that 96 per cent of the respondents recorded operational and efficiency improvements since implementing AI, with a total of 77 per cent saying the technology has had the greatest impact in IT and R&D. In addition, 70 per cent noted significant operational improvements across the value chain once AI has become embedded into core functions. To achieve these benefits, however, manufacturing and industrial organisations must first develop a well-structured AI adoption plan. Jim Chappell, global head of AI and advanced analytics at AVEVA, explains what steps organisations should take to ensure success.
The promise of AI is real, but organisations can only capitalise on it if their data estate is ready. What can they do to prepare to scale up their AI use?
It starts with data hygiene – every time. Companies must get their data in order before investing in AI, which means building a consolidated, accessible and governed data estate. They should ensure information is catalogued and visible within a single pane of glass, so that analytics and machine learning can be confidently applied on top of it.
Can you tell us what factors are driving the success of enterprise-wide AI initiatives?
The best performing initiatives are powered by three essentials: the right foundation, the right people, and the right mindset. Fundamentally, it’s about having structured, governed data and access to multiple platforms which unify operational, customer relationship management and enterprise resource management data. When information is visible in one place, it becomes actionable.
Cross-functional collaboration is also essential. All stakeholders – from frontline operators to data scientists and business leaders – need to be fully engaged. This is because AI adoption isn’t just a technical upgrade, it’s a cultural shift. Users need to be part of the process design, so the solutions get optimally used in practice to support both workers and business goals.
Finally, innovation must be responsibly managed. Layered safeguards, clear objectives and continuous improvement cycles are non-negotiables.
Businesses should also prepare themselves to be nimble. AI strategy should not be tied to any one model or vendor. New technologies will continue to emerge – generative AI with massive language models, agentic AI, immersive interfaces – and companies require agility to pivot with the times.
How are AVEVA and Microsoft working together to empower employees through AI?
At AVEVA, we’ve always believed AI should augment people – not replace them. Our collaboration with Microsoft strengthens that vision. Through platforms like CONNECT and Microsoft Azure, we’re embedding AI into day-to-day workflows in a way that simplifies and streamlines tasks.
Take generative AI as an example. It allows all users to interact with data through natural language. Employees no longer need to be data engineers to get insights, which opens the door for everyone to participate in decision-making.
Agentic AI is the next step in the journey towards task simplification. Microsoft’s work with tools like 365 Copilot and GitHub Copilot is inspiring – it has created agents that handle IT tasks, generate reports, or retrieve data across platforms.
We’re applying that same thinking to the industrial space. We want to create an asset monitoring agent, a safety agent and a data hierarchy agent that can all coordinate to resolve a field issue in real time. But the key is human oversight. AI frees teams to focus on innovation, creativity and higher-value contributions.
Manufacturing firms can use agentic AI to automate processes and provide workers with the information they need to complete their daily tasks
Can you share an insight into how agentic AI might evolve the traditional software-as-a-service model?
Agentic AI refers to a system of autonomous agents that can work independently or together to perform specific tasks within a defined workflow. Think of it as an ecosystem of intelligent assistants – each focused on one area, such as monitoring, compliance or optimisation – that collectively deliver end-to-end automation.
For example, one agent might forecast equipment failures, another could analyse maintenance logs, and a third may handle spare part orders – all with minimal human intervention but governed oversight. This marks a major shift from passive analytics to proactive decision-making.
In contrast, traditional software-as-a-service (SaaS) offered centralised applications with standardised workflows, often requiring users to adapt their processes to the software. Agentic AI turns that model on its head. It brings a more intuitive, customised face that adapts to users, not the other way around. Agentic AI is not replacing SaaS, it is evolving it. The next generation of SaaS will be modular, context-aware and integrated with enterprise data.
What pressures will SaaS companies face as agentic AI becomes more mainstream?
The pressure is twofold: personalisation and performance. Firstly, SaaS providers must offer far more tailored, adaptive user experiences or risk becoming obsolete. Users increasingly expect the software to respond in natural language, automate repetitive tasks and learn from behaviour. Static, one-size-fits-all applications just won’t cut it.
Secondly, SaaS companies must rethink how they handle data. Agentic AI thrives on clean, accessible, governed data. SaaS platforms must become interoperable with other data environments and adhere to strong governance models. Companies that can’t, or won’t, evolve in these areas will lag behind their competition.
Discover more insights like this in the Summer 2025 issue of Technology Record. Don’t miss out – subscribe for free today and get future issues delivered straight to your inbox.