Computer Vision Engineer

Required Skills: C++, Computer Vision, PyTorch, Tensor Country: United States Role Summary We are seeking a highly skilled Computer Vision Engineer to design, develop, and deploy real-time defect detection systems for our industrial automation pipelines. In this role, you will bridge the gap between advanced machine learning models and physical manufacturing hardware. You will be responsible for integrating high-speed cameras, Industrial PCs (IPCs), and Programmable Logic Controllers (PLCs) to capture, process, and label data at high throughput, ensuring zero-defect quality control on the production floor. Key Responsibilities Design and implement robust object detection and image classification algorithms specifically tuned for identifying micro-defects on fast-moving production lines. Integrate and calibrate high-speed industrial cameras, specialized lighting, and optical setups to ensure optimal image quality for the vision models. Establish low-latency communication pipelines between the computer vision systems and PLCs to trigger physical sorting or rejection mechanisms in real-time. Optimize deep learning models for edge deployment using hardware acceleration to meet strict cycle-time requirements. Develop and maintain automated data collection and labeling pipelines to continuously feed and improve the accuracy of the defect detection models. Troubleshoot system performance on the manufacturing floor, addressing issues related to camera triggering, lighting fluctuations, and model drift. Required Programming Languages C++: Essential for writing highly optimized, multithreaded applications necessary for real-time, low-latency image processing and inferencing on the edge. Python: Required for rapidly prototyping, training, testing, and refining deep learning models, as well as managing the data processing and labeling scripts. Structured Text / Ladder Logic (Familiarity): While not strictly required to be an expert, a working knowledge of IEC 61131-3 PLC programming languages is crucial for effectively communicating and designing handshakes with automation engineers. Required Experience & Technical Skills High-Speed Imaging Hardware: Proven hands-on experience with industrial camera interfaces (GigE Vision, CoaXPress, Camera Link), frame grabbers, and industrial optics. Industrial Communication Protocols: Experience seamlessly passing data between Windows/Linux edge computers and PLCs using protocols like OPC UA, Modbus TCP/IP, Ethernet/IP, or PROFINET. Computer Vision & Deep Learning Frameworks: Deep practical knowledge of traditional computer vision (OpenCV) alongside modern deep learning frameworks (PyTorch, TensorFlow) and architectures suited for fast inference (e.g., YOLO variants). Hardware Acceleration & Edge AI: Proven ability to deploy models onto industrial hardware (NVIDIA Jetson, x86 IPCs) using optimization toolkits like TensorRT or OpenVINO to maximize frame rates. Real-Time System Design: Experience dealing with the complexities of concurrent processing, memory management, and preventing frame drops in continuous, 24/7 manufacturing environments. Data Pipeline Management: Familiarity with tools and best practices for versioning datasets, managing synthetic data generation, and coordinating efficient image labeling workflows for quality control. AI Products See the AI platforms we're taking to market Explore what Augusta is building and sharing #J-18808-Ljbffr Augusta Hitech