KairosAD: A SAM-Based Model for Industrial Anomaly Detection on Embedded Devices

2025-05-30

AI Summary

To address the challenge of deploying industrial image anomaly detection models on resource-constrained edge devices in small- and medium-sized manufacturing enterprises—where existing methods suffer from excessive parameter counts, high computational overhead, and poor hardware adaptability—this work pioneers the integration of MobileSAM into a supervised anomaly detection framework, yielding a lightweight and efficient model. Methodologically, we design an edge-optimized architecture based on the Mobile Segment Anything Model, specifically tailored for real-time inference on NVIDIA Jetson NX and AGX platforms. Experimental evaluation demonstrates a 78% reduction in model parameters, a 4× speedup in inference latency, and state-of-the-art AUROC performance. The approach is rigorously validated on standard benchmarks (MVTec-AD, ViSA) and on a real-world production line at ICE Lab, University of Verona, Italy. Our core contribution lies in achieving a unified trade-off among high detection accuracy, ultra-low resource consumption, and practical industrial deployability.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Adversarial Attacks & Robustness
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

In the era of intelligent manufacturing, anomaly detection has become essential for maintaining quality control on modern production lines. However, while many existing models show promising performance, they are often too large, computationally demanding, and impractical to deploy on resource-constrained embedded devices that can be easily installed on the production lines of Small and Medium Enterprises (SMEs). To bridge this gap, we present KairosAD, a novel supervised approach that uses the power of the Mobile Segment Anything Model (MobileSAM) for image-based anomaly detection. KairosAD has been evaluated on the two well-known industrial anomaly detection datasets, i.e., MVTec-AD and ViSA. The results show that KairosAD requires 78% fewer parameters and boasts a 4x faster inference time compared to the leading state-of-the-art model, while maintaining comparable AUROC performance. We deployed KairosAD on two embedded devices, the NVIDIA Jetson NX, and the NVIDIA Jetson AGX. Finally, KairosAD was successfully installed and tested on the real production line of the Industrial Computer Engineering Laboratory (ICE Lab) at the University of Verona. The code is available at https://github.com/intelligolabs/KairosAD.

Problem

• Detects industrial anomalies on resource-constrained embedded devices
• Reduces model size and computation for SMEs' production lines
• Maintains performance while optimizing speed and parameter efficiency

Innovation

• Uses MobileSAM for anomaly detection
• 78% fewer parameters than SOTA
• 4x faster inference on embedded devices