MT Magazine May/June 2025

FEATURE STORY

THE EMERGING MARKETS ISSUE

26

through predictive maintenance), and waste reduction (with AI enhancing CNC programming efficiency). Barnes underscored the importance of data integration and contextualization in democratizing insights and upskilling workers. For AMT members, Barnes recommended starting small with AI by assessing workflow optimization opportunities and identifying incremental steps for implementation. 3. AI Trends Shaping Manufacturing in 2025 With a career spanning two decades across operational and informational manufacturing technology, Praveen Rao from Google Cloud walked AMT members through AI’s ability to accelerate innovation by up to 10 years and automate 50% of tasks, enabling teams to accomplish more with fewer resources.

However, implementing AI is not without challenges. Chouhan pointed out that cost, latency, and accuracy are the top three hurdles, but Nvidia’s industry partnerships with leading technology companies such as Siemens and Google are helping businesses overcome these barriers to achieve a positive ROI. His presentation underscored that the key to successful AI integration is leveraging a robust partner ecosystem. Chouhan concluded with a look into the future of physical AI, which comprises three facets: (1) digital intelligence with physical systems, (2) moving beyond traditional robotics to include humanoids for complex tasks, and (3) autonomous facility management solutions. As AMT members asked thought-provoking questions about AI’s role in additive manufacturing (AM) and data security, it became clear that AI isn’t just about automation; it’s about rethinking how we innovate and operate. With two decades of experience developing cutting-edge industrial technologies at Cisco and Microsoft, Bryce Barnes explored AI’s transformative impact on manufacturing. He traced AI’s evolution from the early days of the Perceptron to the emergence of generative AI and LLMs, emphasizing how deep learning and neural networks have revolutionized cognitive functions in machines. 2. AI’s Impact on Manufacturing — Practice Use Cases

A key concept Rao introduced is multimodal AI, which processes text, images, video, and audio to produce highly contextual and intuitive outputs. This grounding approach leverages data from multiple sources and formats, from enterprise systems to audio and visual feeds, to ensure more robust insights and accurate results. In addition to the multimodal AI concept, which can be seen as an “input” system, Rao brought up five key AI agent use cases, which can be seen as the “output” models for business applications: 1. Customer agents: personalized, real-time support and feedback analysis 2. Creative agents: generating visual and verbal content 3. Data agents: data normalization, predictive modeling, and anomaly detection 4. Code agents: code generation, optimization, and automated testing 5. Security agents: real-time cybersecurity threat detection Finally, he tied these AI capabilities to Google Cloud’s Industry 5.0 framework, outlining how connected data infrastructures and cloud computing can break down traditional operations technology and information technology (OT-IT) silos. AMT members could benefit from reviewing how their operations technology converges with information technology. From Rao’s presentation, we learned that by intentionally building data and technology infrastructures that connect the

Taking the newfound AI-powered intelligence into practical use, Barnes highlighted key use cases: safety management (with AI agents optimizing incident response), factory operations (where AI enhances productivity