01 PID ControlProportional-Integral-Derivative control. A fundamental feedback control algorithm widely used in robotics for maintaining desired positions, velocities, or forces.02 Feedback ControlA control strategy that uses sensor measurements to adjust outputs and maintain desired behavior despite disturbances.03 Feedforward ControlA control strategy that anticipates required outputs based on known system dynamics, without waiting for error feedback.04 Model Predictive Control (MPC)An advanced control technique that uses a model to predict future behavior and optimizes control actions over a time horizon.05 Reinforcement Learning (RL)A machine learning approach where agents learn optimal behaviors through trial and error, receiving rewards for desired actions. Increasingly used for robot control.06 Imitation LearningTraining robots by demonstrating desired behaviors, either through teleoperation or observation. Also called learning from demonstration (LfD).07 Sim-to-RealThe process of training robot controllers in simulation and transferring them to physical robots. Key challenge is bridging the 'reality gap'.08 Foundation ModelA large AI model trained on broad data that can be adapted to various downstream tasks. Emerging approach for general-purpose robot intelligence.09 Vision-Language Model (VLM)An AI model that can understand and reason about both images and text. Enables robots to follow natural language instructions about visual scenes.10 Large Language Model (LLM)An AI model trained on vast text data capable of understanding and generating human language. Increasingly integrated into robot planning and interaction.11 End-to-End LearningTraining a neural network to map directly from sensor inputs to control outputs, without explicit intermediate representations.12 Behavior CloningA form of imitation learning that directly copies demonstrated actions. Simple but can fail in states not seen during training.13 Task PlanningHigh-level reasoning about which actions to take to achieve a goal. Typically symbolic or uses LLMs for reasoning.14 Motion PrimitiveA reusable building block of robot movement that can be combined to create complex behaviors.15 SkillA learned or programmed capability that enables a robot to perform a specific task. Robots compose skills to accomplish complex goals.16 TeleoperationRemote control of a robot by a human operator. Used for dangerous tasks, data collection, and situations requiring human judgment.17 Autonomy LevelThe degree to which a robot can operate without human intervention, ranging from fully manual to fully autonomous.18 Human-Robot Interaction (HRI)The study and design of interactions between humans and robots, encompassing communication, collaboration, and safety.19 VLA (Vision-Language-Action Model)A neural network that takes visual observations and language instructions as input and directly outputs robot actions. Combines perception, language understanding, and motor control in a single model.20 Diffusion PolicyA robot control policy based on diffusion models that generates action sequences by iteratively denoising random noise. Enables learning complex, multimodal behaviors from demonstrations.21 World ModelAn internal model that predicts how the environment will change in response to actions. Enables robots to plan by simulating future outcomes before acting.22 Zero-Shot TransferThe ability of a robot to perform a task it was never explicitly trained on, often by leveraging general knowledge from foundation models or large-scale pretraining.23 Domain RandomizationA sim-to-real technique that varies simulation parameters (lighting, textures, physics) during training to produce policies robust to real-world variation.24 Embodied AIAI systems that learn and reason through physical interaction with the world, as opposed to purely digital AI. Robots are the primary platform for embodied AI research.25 Fleet LearningA paradigm where multiple robots share experiences and learn collectively, accelerating skill acquisition across an entire fleet of deployed systems.26 Dexterous ManipulationThe ability to manipulate objects with human-like dexterity using multi-fingered hands. A grand challenge in robotics requiring advanced sensing and control.