From machines to software…
Manufacturers are investing heavily in robots, automation and industrial AI, but the source of competitive advantage is increasingly shifting away from the hardware itself and towards software, data and AI, Forrester says.
The analyst firm’s new Global Manufacturing Forecast 2025 to 2030 forecasts global manufacturing tech spend to increase 41 percent between 2024 and 2030, hitting US$683.6 billion. Commercial software investment is expected to rise 73 percent during that period, while industrial AI software spending is forecast to grow from US$908 million to US$3.25 billion.
“Hardware builds the factory; software, data and AI determine how quickly it learns.”
George Lawrie, VP and principal analyst at Forrester, says that spending reflects a broader transition already underway as the competitive advantage shifts from owning better hardware to extracting more value from broadly available hardware.
“Modern robots, sensors and automation remain essential, but competitors can increasingly buy similar equipment from the same global suppliers,” he told iStart. That means sustainable advantage now comes from the software and data layer – and the skilled workforce – that determines how quickly a manufacturer can configure the equipment, learn from it, coordinate it across plants and use it to make better decisions.
“Hardware builds the factory; software, data and AI determine how quickly it learns.”
Software-defined manufacturing
Lawrie says manufacturers are increasingly competing on their ability to build software defined operations, rather than simply deploy automation.
“Software-defined manufacturing turns fixed factories into adaptable business capabilities,” he says.
For Australian and New Zealand manufacturers, Lawrie says the shift is an opportunity as well as a risk.
“They are unlikely to outspend the largest Asian manufacturers on physical capacity, but they can compete through specialised know-how, regulated quality, trusted provenance, rapid changeovers and close integration with regional customers.”
To do so, manufacturers will need to treat operational data, software architectures and industrial AI models as strategic production assets, he says.
Companies that merely automate individual tasks may improve efficiency, but those that create a reusable digital thread across engineering, production, quality, maintenance and supply networks “can change the basis of competition”.
That capability is becoming more important as manufacturers deal with changing demand patterns, supply-chain uncertainty and workforce pressures.
“The most successful manufacturers will not simply automate existing processes,” Lawrie says. “They will build intelligent, software-defined operations that can continuously adapt to changing customer demand, supply chain conditions and geopolitical uncertainty.”
The risk of a widening gap
Lawrie warns that while Australian and New Zealand manufacturers are keeping pace with the shift to software-defined manufacturing, there is a risk of a widening gap between a small group of software leaders and a much larger middle and tail. Australian spend is on the rise at a ‘respectable’ rate that is faster than that for Japan and South Korea, but below China. Defensible growth rate data isn’t available for New Zealand. Public company evidence across both countries however shows a similar pattern, with large companies, such as Fonterra, CLS, BlueScope, Cochlear and ResMed investing in connected operations, analytics, automation and enterprise data platforms while many smaller and more domestically focused manufacturers disclose far less evidence of integrated IT/OT architectures, modern MES (manufacturing execution systems), industrial data platforms or scaled AI, he says.
“The opportunity remains credible because software-defined manufacturing reduces the importance of labor cost and scale and increases the value of flexibility, traceability, intellectual property, regulated quality and specialised know-how—areas in which A/NZ manufacturers can compete.”
But keeping pace will require treating software as a production asset rather than an IT overhead he says, urging manufacturers to modernise MES and PLM, create a common industrial data layer, integrate IT and OT, use digital twins and analytics to encode expert knowledge, and scale successful use cases across plants.
“Without that shift, the forecast’s strong growth in industrial AI, IoT, analytics and MES will translate into a capability advantage for regional competitors rather than a productivity dividend for Australian and New Zealand manufacturers.”
Don’t start with AI
When it comes to the range of technologies available, Lawrie says A/NZ manufacturers should prioritise the operating foundation before the most fashionable technology.
“For most Australian and New Zealand manufacturers, the first investment should be a clearly defined business outcome – such as higher yield, shorter changeovers, lower energy use, better traceability or improved schedule adherence – supported by reliable connectivity, contextualised operational data, cybersecurity and common process definitions.
“Without that foundation, industrial AI and digital twins will remain isolated demonstrations.”
His advice to local manufacturers is to modernise manufacturing operations management (MOM) and the industrial data layer first, using MES or a composable MOM platform to connect production, quality, maintenance, labour and materials to ERP and PLM, and an industrial data platform giving equipment and process data consistent context.
“This creates the digital thread, governance and real-time visibility required to scale every subsequent use case.
“Manufacturers with adequate MES coverage should focus less on replacement and more on standardising data models and workflows across sites.”
The next step is funding focused use cases with rapid and measurable returns, such as predictive maintenance for asset-intensive operations where unplanned downtime is costly and sufficient failure history exists, or robotics and autonomous handling where labour scarcity, safety requirements or repeatability – ‘not technology fashion’ – are proven constraints.
Only then should manufacturers look to scale digital twins and industrial AI.
“Digital twins create the most value where firms must simulate complex assets, recipes, production flows or engineering changes before acting. Industrial AI should then augment decisions across scheduling, quality, maintenance, energy and operator support,” he says.
His advice is to start with bounded, human-supervised use cases and reuse the same governed data products across plants; do not build a separate data pipeline for every model.
“Do not start with the smartest technology. Start with the foundation that lets every technology get smarter.”
For manufacturers looking to compete in an increasingly software-defined industry, Lawrie’s prescription is straightforward: “Choose the niche, build the digital backbone and compete at the speed of the customer.”
