By Paul Pickering is Research Director, Semiconductors at Omdia and Saloni Gankar is Principal Analyst, Industrial Semiconductors at Omdia
As industrial AI agents move closer to real time operational control, semiconductor vendors are becoming central to securing the next generation of factory and infrastructure systems. Omdia analysts examine how Edge AI, hardware security, and emerging regulation are reshaping industrial automation.
Industrial companies are increasingly using AI systems to monitor, analyse, and even control physical operations. These AI-driven systems, often referred to as ‘agents’, are now being used to identify faults on production lines, optimise energy consumption, and coordinate complex manufacturing processes without constant human intervention.
Unlike traditional data analysis tools, these agents can make decisions on their own and are integrated directly into the equipment and systems that run factories and utilities.
For semiconductor manufacturers, this shift means creating chips that not only process AI tasks but also ensure security, reliability, and long-term performance in critical environments.
Industrial AI agents face unique challenges compared to AI systems used in offices or the Cloud. They must operate with extremely low delays (sub-millisecond), work independently of Internet connections, and remain isolated from broader IT networks for security.
For example, an AI agent might analyse vibration data from a motor, temperature readings from nearby equipment, and historical trends to recommend adjustments, all without relying on Cloud computing, which could introduce delays or security risks.
This requirement is elevating the role of specialised semiconductors. Secure microcontrollers (MCUs) and systems-on-chip (SoCs) are becoming central to industrial AI deployments, enabling AI models to run locally while protecting sensitive data and ensuring the system operates as intended.
These chips act as a safeguard, ensuring that AI agents follow strict operational rules even under stress. Without this hardware-level security, AI systems could become vulnerable, potentially causing production halts, equipment damage, or safety risks.
AI agents gain ground across industrial automation
Industrial automation leaders are already embedding AI agents into their operational platforms.
Siemens has expanded its Industrial Edge platform to combine AI functionality with certified security frameworks, enabling applications such as automated defect detection and intelligent energy balancing across production systems. The company’s approach reflects the growing need to integrate AI without compromising operational control or compliance requirements.
Similarly, Rockwell Automation is using AI to support industrial operators through functions such as troubleshooting and interface configuration. Importantly, these AI capabilities are designed to operate alongside human personnel while maintaining offline functionality, reinforcing the importance of resilient Edge based processing.
Schneider Electric is also advancing industrial AI through systems that optimise heating, cooling, and equipment maintenance in energy intensive facilities. Such use cases depend on a combination of predictive analytics and highly reliable execution, again underscoring the need for secure, industrial-grade semiconductor platforms.
The semiconductor industry responds to industrial AI demand
To support this transition, semiconductor manufacturers are developing increasingly specialised hardware aimed specifically at industrial AI workloads.
NXP Semiconductors is focusing on chips designed for real-time communications and secure factory operations, enabling AI agents to coordinate machinery and sensor networks with minimal latency.
Infineon Technologies, meanwhile, is concentrating on secure and standards compliant semiconductors suited to highly regulated industries such as automotive and energy. The company’s emphasis on hardware based protection aligns closely with the tightening cybersecurity requirements emerging across industrial markets.
In Asia Pacific, Renesas Electronics is developing semiconductors optimised for distributed intelligence, where AI capabilities are spread across multiple connected devices. This architecture is particularly relevant for robotic assembly lines and sensor-heavy manufacturing environments that require coordinated Edge intelligence.
Long product lifecycles are also shaping semiconductor strategies. Microchip Technology continues to position itself around long-term reliability, offering industrial chips designed to maintain security and performance for up to 15 years. For industries operating infrastructure with extended operational lifespans, this longevity is increasingly valuable.
How industrial AI priorities differ across global markets
Regional approaches to industrial AI adoption are also beginning to diverge.
In North America, the emphasis remains on accelerating workflows and managing increasingly large data volumes. Companies such as Rockwell Automation and NVIDIA are prioritising high-performance Edge computing platforms capable of supporting demanding manufacturing and logistics applications.
Europe, by contrast, is placing greater emphasis on security, governance, and regulatory compliance. Siemens and Schneider Electric are aligning their industrial AI deployments with evolving cybersecurity and safety standards, particularly in response to the European Union’s Cyber Resilience Act (CRA).
Asia Pacific continues to focus heavily on scale and manufacturing efficiency. Companies including Renesas and TSMC are driving hardware-centric AI strategies aimed at supporting high-volume electronics manufacturing, smart sensors, and tightly coordinated production systems.
The Cyber Resilience Act raises the stakes for industrial AI
The European Union’s Cyber Resilience Act (CRA), which took effect in December 2024, requires manufacturers to build cybersecurity into their products from the start. This includes documenting vulnerabilities, providing long term support, and ensuring secure updates.
For semiconductor companies, this means focusing on hardware-based security features like secure firmware updates and mechanisms to verify the integrity of devices.
AI agents make these regulations even more critical. If an AI system is not properly secured, it could cause widespread failures in connected systems, leading to significant financial and operational consequences.
Companies that adopt secure-by-design principles early will have a competitive advantage, especially as European standards influence global markets.
The next phase of industrial AI centres on security
The rise of industrial AI agents requires a unified approach where hardware, software, and operational systems work together under strict security and governance.
Semiconductors are no longer just enablers of AI; they are the foundation that ensures these systems operate safely and reliably. Companies such as NXP, Infineon, Renesas, and Microchip are leading the way by combining their expertise in security, reliability, and performance.
For industrial automation companies, the challenge is to create AI systems that enhance human oversight without compromising control. While regional differences in priorities remain, global standards such as the CRA are driving the industry toward secure and reliable AI systems. The result will be a new era of industrial intelligence in which trusted silicon becomes as strategically important as the software models running on top of it.
Paul Pickering and Saloni Gankar
This article originally appeared in the May 2026 magazine issue of IoT Insider.
