Module 5: AI in Robotics

Sensors that send out energy into the environment and measure how this energy is reflected back. They provide more information but have higher power consumption and potential interference when multiple sensors are used.

Sensors that passively capture signals already present in the environment, such as light or sound waves.

Sensors that measure distances to objects, providing information about the robot's surroundings.

Sensors that capture images of the environment, enabling the robot to see and analyze its surroundings.

Sensors that provide information about the robot's own internal state, such as its position, orientation, and joint angles.

A type of range sensor that emits sound waves and measures the time it takes for the sound to return after bouncing off an object. Used for underwater and near-range collision avoidance.

A type of range sensor that uses electromagnetic waves to measure distances. Used for long-range applications like aircraft navigation.

A type of range sensor that uses a laser beam to measure distances. Provides accurate measurements for both short and long distances.

Physical agents that perform tasks by interacting with the physical world. They have effectors (like arms and legs) to manipulate objects and sensors to perceive their environment.

Robot parts responsible for physical interaction, like arms, legs, wheels, and grippers. They exert forces on the environment.

Robot components that allow them to sense their environment, like cameras, ultrasonic sensors, gyroscopes, and accelerometers.

Robots that are fixed in place, usually used in industrial settings or hospitals. They feature controllable joints for precise movements.

Robots that can move under their own power, such as wheeled robots, walking robots, and drones. They are used for tasks like delivery, transportation, and exploration.

Robots that combine both manipulation and locomotion, like humanoid robots and prosthetics. They combine the capabilities of manipulators and mobile robots.

A real-world challenge for robots where they can't see everything in their surroundings, like objects behind walls or corners.

A real-world challenge for robots where uncertainties and errors exist in movement, like friction and gear slips. This means that robots need to be robust and adaptable.

A locomotion system where all wheels always point in the same direction and move with the same speed.

Robots that move using legs, capable of navigating rough terrain. However, they are slower on flat surfaces and mechanically complex.

A type of legged robot that can remain upright while hopping without needing to move its legs.

Robots that rely on propellers or turbines for locomotion, like drones and blimps.

Robots that utilize thrusters to move underwater, such as submarines and autonomous underwater vehicles (AUVs).

The process by which robots interpret sensor measurements and create internal representations of their environment.

Internal representations in robot perception must have enough information for decisions, be efficiently updatable, and correspond to natural variables in the world.

A model used in robot perception to analyze the relationship between actions, observations, and the state of the environment. It helps in updating the robot's understanding of its environment over time.

The way a robot's position changes over time based on control inputs like speed and direction.

How a robot understands its environment and itself through its sensors.

Determining where objects or the robot is in the environment.

A problem where the robot's initial location is unknown and it needs to figure out where it is.

A problem where the robot's initial location is known, and it needs to keep track of its position over time.

A problem where the robot is moved to an unknown location without knowing where it is.

Stable, recognizable features in the environment used by a robot to figure out its position.

The robot's orientation and position in the environment, described by its Cartesian coordinates (x, y) and its heading (θ).

Sensors that measure forces in both translational and rotational directions. Used in handling delicate objects and adapting to unknown object shapes.

The number of independent movements a robot's joint or body can make. Each axis of movement counts as one DOF.

A robot, like an underwater vehicle (AUV), has six DOFs: three for its position (x, y, z) and three for its orientation (yaw, pitch, roll).

Robots with more effective DOFs (possible movements) than controllable DOFs (movements the robot can control).

Robots with equal effective DOFs and controllable DOFs. Easier to control but mechanically complex.

A robot with two independently actuated wheels (or tracks) on each side, like a tank. Controls movement direction.

A robot that can move forward or backward and rotate, like a car.

This sensor is used to measure the distance to objects and usually provides a vector of range values along multiple beam directions.

This method estimates a robot's location using particles that represent possible robot states. Particles are updated based on the robot's motion and sensor readings.

This filter assumes the robot's belief about its position is a Gaussian distribution and updates the belief using motion and sensor data.

This challenging problem involves a robot simultaneously figuring out its location and creating a map of an unknown environment.

This method uses an Extended Kalman Filter to represent the robot's state and map as a Gaussian distribution, tracking uncertainties and correlations between robot pose and landmarks.

The process of using sensor data to improve a robot's estimate of its position and orientation. It involves updating the robot's belief about its state based on sensor readings.

This process involves selecting particles with higher weights in Monte Carlo Localization to represent a more accurate estimate of the robot's location.