Glossary
Key terms in alphabetical order. `(Ch N)` indicates where the term is introduced or discussed in depth.
A
AMR (Autonomous Mobile Robot): A robot that navigates autonomously without pre-programmed paths by perceiving its environment. Used in logistics warehouses and factory material transport. (Ch 6)
D
Dark Factory: A fully autonomous factory that operates without lighting, where all processes are handled by robots and AI with minimal human intervention. (Ch 9)
Digital Twin: A 1:1 virtual replica of a physical factory, equipment, or process. Used for simulation, optimization, and anomaly detection before actual operation. (Ch 2, 5)
E
Edge AI: AI inference performed directly on field devices (cameras, robots, edge servers) rather than in the cloud. Advantageous for real-time response and data security. (Ch 2, 3)
H
HTE (High-Throughput Experimentation): A research automation technique that uses robots and automated equipment to run hundreds or thousands of experimental conditions simultaneously. (Ch 7, 9)
M
Micro-Factory: A small-scale, high-flexibility, high-turnover production cell specialized for customized and small-batch manufacturing. (Ch 9)
O
ODM (Original Design Manufacturing): A manufacturing model where the manufacturer handles design, development, and production of finished products. Cosmax is the global #1 cosmetics ODM. (Ch 8, 9)
OT (Operational Technology): Hardware and software systems that control and monitor physical equipment in factories and plants. Distinct from IT as a field technology. (Ch 3)
P
Physical AI: AI systems that perceive, reason, and act in the physical world. Unlike AI that only processes digital data, Physical AI combines robots, sensors, edge computing, and digital twins. (Ch 1)
S
SDL (Self-Driving Laboratory): A fully automated research environment where robots and AI autonomously perform hypothesis generation, experimental design, execution, and analysis. (Ch 7, 9)
Sim-to-Real: A technique for deploying AI models trained in simulation to real physical environments. Enables safe, low-cost training followed by field deployment. (Ch 2, 5)
V
VLA (Vision-Language-Action): An AI model integrating vision, language, and action. It understands images and text instructions and executes them as physical actions. Used in robot control. (Ch 1)