At Embedded World 2025, David Steele from Arcturus presented a line inspection demonstration developed in collaboration with 42 Technology (42T). The system operates on Synaptics’ SL1680 processor, integrating advanced object detection and color analysis to identify pills and detect contaminants on a conveyor belt. The SL1680’s architecture includes a quad-core Arm® Cortex®-A73 CPU and a multi-TOPS Neural Processing Unit (NPU), facilitating efficient edge AI processing.
—
Synaptics is my Embedded World 2025 video coverage sponsor, check out my Synaptics videos here: https://www.youtube.com/playlist?list=PL7xXqJFxvYvhAbQoe9YN4c84SqXxIY3fQ
—
The demonstration utilizes a YOLOv8 model for object detection, specifically trained on a custom dataset to recognize pills. By leveraging Synaptics’ toolkit, the model’s preprocessing is executed directly on the NPU, reducing CPU load and optimizing system performance. This allocation allows the CPU to handle visualization and presentation tasks more effectively.
Integrating natural language processing (NLP) capabilities, such as ChatGPT, into such systems could enhance user interaction by enabling queries about anomaly correlations or failure statistics. This would involve adding another model to the existing ensemble, a feasible extension given the system’s modular design.
Arcturus employs an MLOps framework that facilitates continuous model improvement directly at the edge. This process involves collecting data from the operational device, building new datasets, training updated models, and comparing their performance against existing ones. Successful models can then be deployed in real-time, ensuring the system adapts to evolving operational conditions.
The application of AI in industrial inspection is poised to enhance analysis capabilities significantly. By deploying cost-effective models on efficient hardware like the SL1680, industries can achieve substantial improvements in inspection accuracy and speed. This advancement represents a shift from legacy algorithms to more sophisticated AI-driven solutions.
Arcturus is actively engaged in multiple AI projects, with approximately six initiatives currently underway. The rapid pace of AI innovation presents both challenges and opportunities, and the team is enthusiastic about exploring new use cases, particularly in edge computing environments.
Check out all my Embedded World 2025 videos in this playlist: https://www.youtube.com/playlist?list=PL7xXqJFxvYvjgUpdNMBkGzEWU6YVxR8Ga
This video was filmed using the DJI Pocket 3 ($669 at https://amzn.to/4aMpKIC using the dual wireless DJI Mic 2 microphones with the DJI lapel microphone https://amzn.to/3XIj3l8 ), watch all my DJI Pocket 3 videos here https://www.youtube.com/playlist?list=PL7xXqJFxvYvhDlWIAxm_pR9dp7ArSkhKK
Join https://www.youtube.com/charbax/join for Early Access to my videos and to support my work, or you can Click the “Super Thanks” button below the video to send a highlighted comment!
