Industrial AI researcher and engineering leader focusing on AI Safety, Industrial Safety OS, Industrial Spatial Safety, Vision Operating Systems, AI Agents, Industrial Vision, Embodied / Physical AI, CUDA acceleration, and Intelligent Manufacturing.
My current research explores how AI can move beyond perception and prediction toward risk-aware, fail-safe and evidence-driven industrial safety systems that connect sensing, spatial reasoning, risk assessment, decision making, machine control, event traceability, V&V and Safety Case engineering.
2022 – 2026
Led smart factory transformation with Vision Operating System, AI Agents, and real-time quality inspection systems.
View LinkedIn →2015 – 2022
Designed end-to-end Vision Operating System and CUDA-accelerated industrial vision pipelines.
View LinkedIn →Building trustworthy industrial safety systems that integrate functional safety, spatial intelligence, AI perception, risk reasoning, deterministic control, lifecycle governance, V&V and Safety Case evidence.
Research on how AI perception, multimodal models and intelligent agents can enhance industrial safety while explicitly managing uncertainty, model limitations, degraded operation and residual risk.
Explore AI Safety →An industrial spatial safety operating system connecting 3D perception, AGV, Robot and PLC data with spatial modeling, AI risk assessment, Safety Supervisor, deterministic control and event traceability.
Explore Safety OS →Runtime safety orchestration covering Normal, Warning, Slowdown, Pause, E-Stop Request, Degraded, Recovery and Fault states, with safer-action priority and controlled recovery.
View Architecture →Evidence-driven lifecycle engineering connecting hazards, safety goals, safety requirements, design controls, test cases, evidence and residual-risk acceptance.
Explore Safety Engineering →Explore the Industrial AI Knowledge Base through four interconnected hubs.
Industrial AI, Vision Operating System, AI Agent, Embodied AI, Industrial Safety, AI Safety, Future Manufacturing.
Explore →Engineering Documentation, CUDA, Camera SDK, Algorithm SDK, Industrial Vision SDK, Vision Kernel.
Explore →CameraSDK, AlgorithmSDK, VisionOS, Industrial AI Platform, Engineering Practice.
Explore →Industrial AI runtime architecture for real-time perception, reasoning, decision and execution.
Read More →A system-level research framework integrating functional safety, industrial AI, spatial perception, uncertainty-aware risk assessment, deterministic safety supervision, V&V and lifecycle evidence management.
Read More →A closed-loop industrial spatial safety architecture for lithium battery Pack lines, integrating Camera, AGV, Robot, PLC, MES, Digital Twin, and AI Agent.
Read the topic →Industrial AI Agent ecosystem integrating LLM, VLM, workflow and manufacturing intelligence.
Read More →3D Spatial Intelligence is the ability of artificial intelligence systems to perceive, interpret, and reason about the physical three-dimensional world, moving beyond flat 2D images and text data to understand depth, physics, geometry, and spatial relationships.
Read More →Real-time event-driven industrial vision runtime with GPU acceleration.
Read More →Cross-platform industrial camera SDK supporting GigE, USB3, GMSL, MIPI, Python, C++, Zero Copy.
View Project →Industrial Vision SDK including Calibration, Measurement, 3D, AI, CUDA and Python bindings.
View Project →Industrial AI operating environment integrating AI, Vision, Workflow, Agent, and Runtime.
View Project →Industrial Safety Operating System for people–vehicle–robot– equipment spatial safety. Integrates 3D ToF sensing, AGV, Robot and PLC interfaces, spatial modeling, AI Risk Engine, Safety Supervisor, Event Bus, Recorder and Safety Dashboard.
View Safety OS →From industrial perception to AI safety and intelligent agents — connecting runtime systems, safety engineering and physical-world execution.
Dr. Junyang Zhou is an Industrial AI researcher, engineering leader, and architect with deep expertise in Industrial AI, Industrial Safety, Vision Operating Systems, AI Agents, and Intelligent Manufacturing.