Series 16 – Episode 6 – Generative AI: an industrial use case

As Generative AI (GenAI) advances from generating content to revolutionising industrial operations, enterprises seek to understand how it might align with their objectives.

Di Segni stated: “Generative AI has massive implications for product design and manufacturing. Using advanced algorithms, companies can rapidly create and evaluate multiple design iterations, optimising factors such as cost, materials, and performance. This approach accelerates the innovation cycle and enables more efficient, cost-effective products.”

In manufacturing, GenAI’s benefits are substantial. Di Segni explained how it optimises complex processes through large-scale data analysis, facilitating more efficient operations by suggesting improvements. Predictive maintenance is a critical example, as GenAI’s ability to forecast equipment failures before they occur reduces downtime and cuts maintenance costs.

Additionally, Di Segni highlighted supply chain optimisation: “GenAI can simulate various scenarios to anticipate disruptions and recommend adjustments, making enterprises more resilient and adaptable to market shifts.”

This predictive capability allows enterprises to better manage supply chains by simulating potential disruptions and suggesting preventive measures, adding a layer of adaptability in volatile markets.

GenAI also enables enterprises to offer personalised solutions at scale. SECO has focused on integrating GenAI with Edge computing to drive real-time data processing and decision-making, which is essential for responsive and safe industrial settings. Di Segni noted: “By running AI models directly at the Edge, we minimise latency and improve responsiveness. This approach is crucial in industrial environments where immediate action can significantly impact productivity and safety.”

In practice, this integration supports tailored products or services, enhancing customer satisfaction by enabling quick adjustments based on real-time data.

Despite its potential, GenAI’s integration into industrial systems presents significant challenges, such as data quality and availability. Industrial data is often fragmented, unstructured, or siloed, complicating its use for AI model training.

Computational resources are another hurdle. GenAI models are computationally intensive, requiring considerable processing power and dedicated hardware. SECO has addressed this through Edge computing solutions optimised for AI workloads. However, deploying GenAI at the Edge still demands robust infrastructure, an investment not all enterprises are prepared to make.

Legacy systems also pose a challenge, as many industrial setups were not designed with AI integration in mind. Ensuring interoperability between new AI technologies and existing systems often requires retrofitting, which can be both costly and technically challenging.

Security and privacy concerns add further complexity. Integrating AI solutions introduces new cybersecurity risks, necessitating a secure-by-design approach. SECO has developed multi-layered security solutions to safeguard data, from secure boot mechanisms in hardware to data encryption protocols.

Looking forward, SECO aims to make GenAI a widespread feature in industrial devices within five years. Di Segni envisioned a future where devices powered by GenAI are not just tools but intelligent partners capable of optimising operations and anticipating user needs. These advancements could transform sectors such as retail, where GenAI-enabled smart shelves provide personalised recommendations, and healthcare, where devices monitor patient health in real time and offer personalised care recommendations.

To hear more about Generative AI and much more, you can listen to Electronic Specifier’s interview with Fausto di Segni on Spotify or Apple podcasts.