TechClub Inc is seeking a Machine Vision Engineer to design and implement advanced computer vision solutions that enhance asset intelligence, operational safety, and predictive maintenance. This role focuses on vision-based systems for real-time inspection, anomaly detection, and automation across locomotives, freight cars, and yard equipment, directly impacting logistics efficiency and operational safety.
Responsibilities:
- Build robust machine vision systems for object detection, classification, and anomaly detection in rail environments
- Select cameras (sensor type, resolution, shutter type) for industrial and outdoor conditions
- Choose lenses based on focal length, depth of field, and distortion control
- Configure optimal distance to object, viewing angle, and mounting stability
- Implement lighting strategies (front, back, diffuse, structured)
- Account for glare, shadows, and color temperature variations
- Implement and optimize vision algorithms using TensorFlow, PyTorch, and OpenCV
- Customize open-source frameworks such as YOLO, Detectron2, and MMDetection
- Select and fine-tune pre-trained base models
- Build automated pipelines for training, validation, and deployment
- Monitor performance and retrain models based on accuracy and drift detection
- Work with LabelImg, CVAT, and Roboflow to maintain high-quality datasets
- Optimize for real-time, high-volume image processing
- Implement monitoring and diagnostics to ensure system uptime and accuracy
Requirements:
- 8+ years of experience in computer vision and machine learning
- Strong proficiency in Python and C++
- Expertise with OpenCV, scikit-image
- Hands-on experience with TensorFlow, PyTorch
- Experience with YOLO, Detectron2, and open-source vision models
- Knowledge of camera hardware, lighting, and industrial environments
- Familiarity with automated ML pipelines and orchestration tools
- Experience with CI/CD and real-time performance tuning
- Industrial vision systems and robotics integration
- 3D vision, LiDAR, and sensor fusion
- Cloud-based ML pipelines and distributed inference