NXP has expanded its agentic AI-based autonomous systems ecosystem in the edge AI market.

NXP officially announced its 'eIQ Agentic AI Framework', the core of its next-generation edge AI strategy, at CES 2026, marking the start of its full-scale attack on the autonomous edge intelligence market.

This framework aims to implement agentic AI that performs real-time decision-making in industrial, medical, and smart building environments where low latency, high reliability, and data privacy are essential.

Through this announcement, NXP presented its vision of “AI that makes decisions and takes action directly at the edge.” The eIQ agentic AI framework can perform multi-model inference, decision-making, and control on the device itself without relying on the cloud, making it advantageous in areas requiring immediate response, such as factory equipment control, medical monitoring, and risk detection.

NXP emphasized that agentic AI will become a central technology for next-generation automation.

The new framework features a deterministic real-time decision engine, multi-model orchestration capabilities, and intelligent scheduling to distribute workloads across CPUs, NPUs, and accelerators.

This allows for the simultaneous execution of vision, audio, time series, and control models, supporting a variety of real-time environments such as robotics, industrial automation, smart buildings, and transportation systems.

“The eIQ agentic AI framework is a key component that empowers edge AI devices,” said Charles Dax, senior vice president and general manager of NXP. “It is designed to enable both experienced and novice developers to quickly build and deploy agentic AI.”

NXP, in collaboration with GE Healthcare, unveiled an anesthesia delivery and neonatal monitoring concept at CES. This is an example of implementing safe on-device AI in the medical field, and includes functions that analyze patient conditions in real time and provide immediate notifications to medical staff.

Honeywell also plans to expand the application of agentic AI in building automation and industrial control based on its long-term partnership with NXP.

“NXP’s next-generation AI agent will accelerate the creation of more autonomous and resilient systems,” said Suresh Venkatarayalu, CTO of Honeywell.

The eIQ agentic AI framework supports i.MX 8 and i.MX 9 processors and Ara discrete NPUs, and is compatible with open agentic standards such as A2A and MCP. This enables developers to easily deploy complex multi-step AI workflows to edge environments.

NXP also unveiled its cloud-based eIQ AI hub.

Developers can instantly access the latest AI tools in the cloud to accelerate prototyping and remotely deploy them to actual hardware boards to validate performance. Various tools, including eIQ Time Series Studio and eIQ GenAI Flow, are also available through the hub.

NXP emphasizes security as a core element of agentic AI. The framework includes features to prevent prompt injection, adversarial input, and model spoofing, and combines with NXP hardware security technologies such as secure boot, runtime isolation, and root-of-trust to support deployment in high-trust environments.