ARM Hub says industrial artificial intelligence (AI) is delivering measurable productivity improvements in manufacturing by focusing on practical applications, rather than relying solely on widely used generative AI tools.
In a statement, ARM Hub said industrial AI combines physical and embodied AI to support decision-making and actions in industrial environments, with applications already demonstrating results in manufacturing settings overseas and in Australia.
The organisation cited findings from the University of Technology Sydney’s nine-month Turning AI into Productivity research program, which examined 14 innovation ecosystems across Europe and the Nordics. According to the research, early adopters achieved positive outcomes by combining high-quality data, clearly defined objectives and internal champions focused on business outcomes.
Examples highlighted in the research included an AI-assisted screwdriving system in Dortmund that helps improve assembly quality by reducing rework, AI-generated work instructions and robot programming support at Siemens in Munich, and the integration of AI into lithography and precision manufacturing processes at ASML in Eindhoven.
ARM Hub also pointed to its collaboration with the Air Conditioning and Mechanical Contractors’ Association (AMCA), where AI tools were developed using only AMCA-approved safety content. According to ARM Hub, the system reduced the time required to prepare a compliant Safe Work Method Statement from around four hours to less than 15 minutes while ensuring responses remained traceable to approved documentation.
The organisation said these examples demonstrate that industrial AI is most effective when general-purpose AI models are combined with an organisation’s own validated knowledge and operated by skilled workers.
ARM Hub noted that despite significant global investment in AI, productivity gains have not been universal. Referring to the UTS research, it said successful adoption depends on complementary investments in workforce capability, data governance and management practices, with many organisations encountering what the researchers describe as the “Management Chasm” when attempting to move AI projects from pilot programs into everyday operations.
The statement also highlighted the role of institutions such as Germany’s Fraunhofer network, the German Research Center for Artificial Intelligence (DFKI) and SmartFactory KL in helping manufacturers trial and implement AI technologies before wider deployment.
Drawing on the experience of its AI Adopt Centre, which ARM Hub said has worked with more than 300 Australian small and medium-sized enterprises over the past 18 months, the organisation said AI delivers the greatest value when it reduces search time, rework and reliance on individual expertise, while projects that remain disconnected from operational workflows are less likely to succeed.
ARM Hub recommended that manufacturers prioritise improving the quality of business data, implement several small AI-driven automation projects over time rather than pursuing large-scale transformations, and begin with operational problems identified by frontline staff.
According to ARM Hub, these practical approaches could help strengthen Australia’s manufacturing capability as AI adoption continues to expand globally.
