Advantech Advances Physical AI and Robotics With Integrated Functional Safety

Advantech's platform expansion targets the growing need for safe, validated, and deployable intelligent robots. As AI moves into physical spaces, robots are evolving into "Physical AI" systems, capable of perceiving, processing, deciding, and interacting with their surroundings. This shift necessitates that safety becomes an integral part of the system architecture, particularly as robots increasingly operate beyond controlled environments and interact with people and machines.

To achieve this, Advantech has integrated Functional Safety (FuSa) into its robot platform, employing a five-layer architecture that spans perception to deployment. A key aspect of this approach is separating the AI compute path from an independent safety path. This allows the AI to manage inference, vision, and control tasks, while the dedicated safety path monitors system health, detects faults, and facilitates application-defined safe-state handling. This architectural choice aims to ensure that AI advancements do not compromise fundamental safety.

The platform incorporates specific hardware like the MIC-735-IT, powered by NVIDIA IGX Thor and configurable with a Renesas Safety MCU, along with diagnostic mechanisms and safety documentation, including FMEDA-related evidence. Advantech states these integrations help Robot OEMs reduce the complexity and effort involved in system development, integration, and validation. By providing platform-level design and support, Advantech assists customers in addressing functional safety standards such as IEC 61508 and ISO 13849, streamlining the path towards real-world robot deployment.