VIP20
本课程属于：005-Artificial Intelligence and Machine Learning Engineering English Technical Courses · 共 20 门相关课程，组合学习更完整。
查看所属VIP全部课程 →
← 返回所属目录 咨询 / 购买

Production AI System Design for Industrial Applications

30 Chapters
View Course Samples Online
01
Production AI system landscape
defining production-grade AI, industrial constraints, and the gap between prototypes and deployable systems.
Productionlandscape
02
Industrial data pipelines
ingesting sensor, log, and image data from PLCs, SCADA, MES, and historians with latency and reliability requirements.
Industrialpipelines
03
Feature engineering for industrial time-series and multimodal data
windowing, resampling, missing-value handling, and domain-specific transformations.
industrialtime-series
04
Model selection and training for industrial tasks
trade-offs among classical ML, deep learning, and hybrid approaches under compute and data constraints.
selectiontraining
05
Model compression and optimization
quantization, pruning, distillation, and ONNX/TensorRT conversion for edge and embedded deployment.
compressionoptimization
06
Edge inference architecture
deploying models on industrial PCs, edge gateways, and microcontrollers with real-time and safety constraints.
inferencearchitecture
07
Cloud-edge collaborative inference
splitting workloads, model synchronization, and fallback strategies for intermittent connectivity.
Cloud-edgecollaborative
08
Latency and throughput engineering
profiling inference pipelines, batching, caching, and meeting hard real-time deadlines in production lines.
throughputprofiling
09
Reliability and fault tolerance
designing for sensor drift, missing inputs, hardware failures, and graceful degradation in AI systems.
Reliabilitytolerance
10
Model monitoring and drift detection in production
data drift, concept drift, and statistical process control for industrial AI.
monitoringdetection
11
Continuous training and MLOps for industrial AI
CI/CD pipelines, model registries, versioning, and automated retraining triggers.
Continuoustraining
12
Safety and risk management
functional safety, IEC 61508, ISO 26262, and AI-specific risk assessment for industrial automation.
managementfunctional
13
Explainability and interpretability for industrial operators
SHAP, LIME, attention maps, and human-in-the-loop decision support.
Explainabilityinterpretability
14
Human-machine interaction in industrial AI
operator interfaces, alarm management, and trust calibration for AI recommendations.
Human-machineinteraction
15
Data security and privacy in industrial AI
OT/IT convergence, network segmentation, encryption, and access control for model endpoints.
securityindustrial
16
Regulatory and compliance landscape
EU AI Act, FDA, CE marking, and industry-specific standards for AI in production environments.
Regulatorycompliance
17
Predictive maintenance system design
vibration, acoustic, and thermal data pipelines, remaining useful life estimation, and maintenance scheduling.
Predictivemaintenance
18
Quality inspection and defect detection
computer vision systems for inline inspection, anomaly detection, and yield improvement.
inspectiondetection
19
Process optimization and control
reinforcement learning and model predictive control for setpoint optimization and energy efficiency.
optimizationreinforcement
20
Digital twins and simulation-to-reality transfer
co-simulation, domain randomization, and sim-to-real for industrial AI agents.
simulation-to-realitytransfer
21
Robotics and autonomous systems in production
perception, motion planning, and safe human-robot collaboration.
Roboticsautonomous
22
Supply chain and logistics AI
demand forecasting, inventory optimization, and route planning under industrial constraints.
logisticsforecasting
23
Energy management and sustainability
AI for load balancing, carbon footprint reduction, and renewable integration in factories.
managementsustainability
24
Scalability and distributed systems
microservices, message queues, Kubernetes, and orchestration for plant-wide AI deployments.
Scalabilitydistributed
25
Cost engineering for production AI
TCO analysis, cloud vs. on-prem trade-offs, and ROI modeling for industrial AI projects.
productionTCO
26
Case study
end-to-end predictive maintenance deployment in a heavy manufacturing plant.
end-to-endpredictive
27
Case study
real-time visual quality inspection on a high-speed production line.
real-timeinspection
28
Case study
AI-driven energy optimization in a continuous process industry.
AI-drivenoptimization
29
Incident response and root cause analysis for AI system failures in production
post-mortems, rollback, and corrective actions.
Incidentresponse
30
Future directions and emerging practices
foundation models for industrial time-series, federated learning, and autonomous production systems.
directionsemerging