Advantech has recently unveiled its MIC-AI series, which is based on NVIDIA Jetson Thor and is specifically designed to support agentic AI workloads at the industrial edge. This series includes the MIC-743, MIC-742, and MIC-741 systems, along with the MIB-741 and MIB-742 development boards tailored for NVIDIA Jetson T4000 and T5000 modules.
For readers of eeNews Europe, this announcement holds significance as agentic AI is transitioning from cloud-based orchestration to local deployment in various industrial settings such as factories, inspection systems, and autonomous machines. This shift has the potential to impact how engineers approach the design of edge compute, sensor fusion, and industrial automation architectures.
The MIC-AI platforms are engineered to execute generative and agentic AI models at the edge, reducing the reliance on remote cloud infrastructure. Advantech highlights that these systems can integrate NVIDIA Nemotron 3, OpenClaw, and the NVIDIA OpenShell secure runtime through NVIDIA NemoClaw on Jetson Thor.
Within an industrial environment, this capability enables an authorized AI system to access standard operating procedures, maintenance records, and spare parts information to make real-time adjustments to production sequences. Advantech provides an example where a system detects low component stock, contacts suppliers, monitors progress, reschedules production, and assigns work orders to engineers.
While this closed-loop automation showcases the potential of agentic AI, its successful implementation relies heavily on policy control, system integration, and validation. Nevertheless, it offers insights into the broader applications of agentic AI beyond standalone inference tasks, particularly in smart manufacturing and automated inspection.
The MIC-743, MIC-742, and MIC-741 systems are compatible with NVIDIA Jetson T4000 and T5000 modules. Advantech reveals that the Jetson T5000 series is already in mass production, with the Jetson T4000 series scheduled to commence phased production by the end of April 2026.
These systems deliver up to 2,070 FP4 TFLOPS of AI performance, while the MIB-741 and MIB-742 development boards offer up to 1,200 FP4 TFLOPS. Additionally, the platforms feature high-bandwidth sensor interfaces and industrial connectivity to facilitate real-time perception, synchronized multi-sensor processing, and low-latency decision-making.
Advantech positions the MIC-AI series as a pivotal step towards physical AI deployment, where AI models interact directly with machines, production lines, and sensor networks. By integrating agentic AI frameworks into Jetson Thor-based hardware, the company aims to empower enterprises to experiment with and implement autonomous AI systems in close proximity to equipment and data sources.