12.3 Robotics and automation

Robotics and automation blend control theory, engineering, and computer science to create intelligent machines. These systems perform tasks autonomously or with minimal human input, revolutionizing industries from manufacturing to healthcare and space exploration.

Robot components include actuators, sensors, control systems, and end effectors. Understanding kinematics, dynamics, and control is crucial for designing effective robots. Path planning, sensing, and programming enable robots to navigate and interact with their environment autonomously.

Robotics overview

• Robotics is a multidisciplinary field that combines principles from control theory, mechanical engineering, electrical engineering, and computer science to design and develop intelligent machines capable of performing tasks autonomously or with minimal human intervention
• The field of robotics has seen significant advancements in recent years, with robots being used in a wide range of applications, from manufacturing and assembly to healthcare and space exploration

Types of robots

Top images from around the web for Types of robots

• 

  Japanese robotics - Wikipedia View original

  Is this image relevant?

• 

  Collaborative Robots Transforming Work and Learning - HAMK Unlimited View original

  Is this image relevant?

• 

  Cobot - Wikipedia View original

  Is this image relevant?

• 

  Japanese robotics - Wikipedia View original

  Is this image relevant?

• 

  Collaborative Robots Transforming Work and Learning - HAMK Unlimited View original

  Is this image relevant?

1 of 3

Top images from around the web for Types of robots

• 

  Japanese robotics - Wikipedia View original

  Is this image relevant?

• 

  Collaborative Robots Transforming Work and Learning - HAMK Unlimited View original

  Is this image relevant?

• 

  Cobot - Wikipedia View original

  Is this image relevant?

• 

  Japanese robotics - Wikipedia View original

  Is this image relevant?

• 

  Collaborative Robots Transforming Work and Learning - HAMK Unlimited View original

  Is this image relevant?

1 of 3

• Industrial robots are used in manufacturing and assembly lines to perform repetitive tasks with high precision and speed (welding, painting, material handling)
• Service robots assist humans in various tasks, such as cleaning, delivery, and customer service (vacuum cleaners, delivery drones, receptionist robots)
• Mobile robots are designed to navigate and operate in different environments, including wheeled robots, legged robots, and aerial robots (Mars rovers, quadruped robots, drones)
• Collaborative robots, or cobots, are designed to work safely alongside humans in shared workspaces (assembly tasks, quality inspection)

Robot components

• Actuators are the motors or hydraulic/pneumatic systems that enable robot motion and force generation (electric motors, hydraulic cylinders, pneumatic actuators)
• Sensors allow robots to perceive their environment and gather information for decision-making (encoders, force/torque sensors, cameras, LiDAR)
• Control systems process sensor data, execute algorithms, and generate commands for actuators to achieve desired robot behavior (microcontrollers, embedded systems, industrial PCs)
• End effectors are the tools or devices attached to the robot's arm to interact with the environment (grippers, suction cups, welding torches)

Degrees of freedom

• Degrees of freedom (DOF) refer to the number of independent motions a robot can perform in its workspace
• The number of DOF depends on the robot's mechanical design and the arrangement of its joints (revolute joints, prismatic joints)
• A higher number of DOF allows for greater flexibility and dexterity in performing tasks but also increases the complexity of control and path planning
• Industrial robots typically have 6 DOF, allowing them to position and orient their end effector in any desired pose within their workspace

Robot kinematics

• Robot kinematics is the study of robot motion without considering the forces and torques that cause the motion
• Kinematics deals with the relationship between the robot's joint angles and the position and orientation of its end effector in the workspace

Forward kinematics

• Forward kinematics is the process of determining the position and orientation of the robot's end effector given the joint angles
• The forward kinematics problem is solved using the robot's kinematic equations, which are derived based on the robot's mechanical structure and joint types
• Homogeneous transformation matrices are commonly used to represent the relative positions and orientations between robot links and joints
• The forward kinematics solution is unique for a given set of joint angles

Inverse kinematics

• Inverse kinematics is the process of determining the joint angles required to achieve a desired position and orientation of the robot's end effector
• The inverse kinematics problem is more complex than forward kinematics, as there may be multiple solutions or no solution at all for a given end effector pose
• Numerical methods, such as the Jacobian-based approach or optimization techniques, are often used to solve the inverse kinematics problem
• Redundant robots, which have more DOF than necessary for a task, can have an infinite number of inverse kinematics solutions

Denavit-Hartenberg parameters

• The Denavit-Hartenberg (DH) convention is a systematic method for assigning coordinate frames to the links of a robot and describing their relative positions and orientations
• DH parameters consist of four variables for each link: link length, link twist, joint offset, and joint angle
• The DH convention simplifies the derivation of the robot's kinematic equations by following a set of rules for assigning coordinate frames
• The DH parameters are used to construct the homogeneous transformation matrices between adjacent links, which can be concatenated to obtain the overall forward kinematics solution

Robot dynamics

• Robot dynamics is the study of the forces and torques that cause robot motion, taking into account the robot's mass, inertia, and external forces acting on it
• Understanding robot dynamics is crucial for designing effective control systems and optimizing robot performance

Lagrangian formulation

• The Lagrangian formulation is an energy-based approach to deriving the equations of motion for a robot
• It considers the difference between the robot's kinetic energy and potential energy, called the Lagrangian
• The Lagrangian equations of motion are obtained by applying the Euler-Lagrange equation to the Lagrangian function
• This approach is well-suited for robots with complex structures and multiple degrees of freedom

Newton-Euler formulation

• The Newton-Euler formulation is a recursive method for deriving the equations of motion based on Newton's laws of motion and Euler's equations for rigid body dynamics
• It involves computing the forces and torques acting on each link of the robot, starting from the base and propagating towards the end effector
• The Newton-Euler formulation is computationally efficient and suitable for real-time control applications
• It provides a systematic way to account for the effects of gravity, friction, and external forces on the robot's motion

Equations of motion

• The equations of motion describe the relationship between the joint torques and the resulting motion of the robot
• They take the form of a set of second-order differential equations, relating the joint angles, velocities, and accelerations to the applied joint torques
• The equations of motion include terms for the robot's mass matrix, Coriolis and centrifugal forces, gravity forces, and friction forces
• These equations are used to design controllers that compute the required joint torques to achieve desired robot motions while compensating for dynamic effects

Robot control

• Robot control involves designing algorithms and systems to regulate the motion and force output of a robot to achieve desired tasks
• Control theory concepts, such as feedback control, stability analysis, and optimization, are applied to robot control problems

Joint space control

• Joint space control focuses on controlling the individual joint angles of the robot to track desired joint trajectories
• PID (Proportional-Integral-Derivative) control is a common approach for joint space control, where the controller computes joint torques based on the error between the desired and actual joint angles
• Feedforward control can be added to compensate for the robot's dynamics, such as gravity and friction, to improve tracking performance
• Joint space control is suitable for tasks that primarily involve motion planning in the robot's joint space, such as point-to-point movements

Operational space control

• Operational space control, also known as Cartesian space control, focuses on controlling the position and orientation of the robot's end effector in the task space
• The control law is formulated in the operational space, and the computed forces are then mapped to joint torques using the robot's Jacobian matrix
• Operational space control allows for more intuitive specification of tasks and can handle constraints in the task space, such as obstacle avoidance
• Examples of operational space control include impedance control, where the robot behaves like a mass-spring-damper system, and hybrid position/force control

Hybrid position/force control

• Hybrid position/force control is a technique that allows a robot to simultaneously control the position and force of its end effector in different directions
• The task space is decomposed into position-controlled and force-controlled subspaces, and separate control laws are applied to each subspace
• This approach is useful for tasks that involve contact with the environment, such as assembly, grinding, or polishing
• The control law ensures that the desired position trajectories are followed in the position-controlled directions while maintaining the desired contact forces in the force-controlled directions

Path planning

• Path planning is the process of generating a collision-free path for a robot to move from its current configuration to a desired goal configuration in its workspace
• It involves representing the robot's environment, identifying obstacles, and searching for feasible paths that satisfy certain criteria, such as shortest distance or minimum energy consumption

Configuration space

• Configuration space, or C-space, is a mathematical representation of all possible configurations (positions and orientations) that a robot can attain in its workspace
• The dimensions of the C-space correspond to the degrees of freedom of the robot
• Obstacles in the workspace are mapped into the C-space, creating regions that the robot must avoid
• Path planning algorithms search for paths in the C-space that connect the start and goal configurations while avoiding obstacles

Sampling-based methods

• Sampling-based methods are a class of path planning algorithms that rely on randomly sampling the C-space to build a graph or tree of feasible configurations
• These methods are particularly effective for high-dimensional C-spaces and complex environments
• Examples of sampling-based methods include Rapidly-exploring Random Trees (RRT), Probabilistic Roadmaps (PRM), and their variants
• These algorithms incrementally expand the graph or tree by adding new configurations and connecting them to existing ones until a path is found

Optimization-based methods

• Optimization-based methods formulate the path planning problem as an optimization problem, seeking to minimize a cost function while satisfying constraints
• The cost function can represent various criteria, such as path length, energy consumption, or smoothness
• Constraints include collision avoidance, joint limits, and dynamic feasibility
• Examples of optimization-based methods include gradient-based techniques, such as sequential quadratic programming (SQP), and stochastic optimization, such as particle swarm optimization (PSO)
• These methods can generate optimal or near-optimal paths but may be computationally expensive for high-dimensional problems

Sensing and perception

• Sensing and perception enable robots to gather information about their environment and internal states, which is essential for decision-making, control, and interaction
• Sensors can be classified into proprioceptive sensors, which measure the robot's internal states, and exteroceptive sensors, which measure the external environment

Proprioceptive sensors

• Proprioceptive sensors provide information about the robot's internal states, such as joint angles, velocities, and torques
• Encoders are commonly used to measure joint angles and velocities by tracking the rotation of the robot's joints (optical encoders, magnetic encoders)
• Force/torque sensors measure the forces and torques applied to the robot's joints or end effector, enabling force control and collision detection
• Inertial Measurement Units (IMUs) combine accelerometers and gyroscopes to estimate the robot's orientation and motion in space

Exteroceptive sensors

• Exteroceptive sensors gather information about the robot's external environment, such as the presence of objects, distances, and visual features
• Cameras are widely used for visual perception, providing rich information about the environment (RGB cameras, depth cameras, stereo cameras)
• LiDAR (Light Detection and Ranging) sensors use laser beams to measure distances and create 3D point clouds of the environment
• Ultrasonic sensors and infrared sensors are used for proximity detection and obstacle avoidance
• Tactile sensors, such as pressure-sensitive skin or capacitive sensors, enable robots to sense contact and forces when interacting with objects

Sensor fusion

• Sensor fusion is the process of combining data from multiple sensors to obtain a more accurate and comprehensive understanding of the environment
• Probabilistic techniques, such as Kalman filters and particle filters, are commonly used for sensor fusion
• These techniques estimate the robot's state (position, orientation, velocity) by integrating information from different sensors and accounting for their uncertainties
• Sensor fusion can help overcome the limitations of individual sensors and provide a robust perception of the environment, even in the presence of noise or occlusions

Robot programming

• Robot programming involves developing software and algorithms to control the behavior and actions of robots
• It encompasses a wide range of tasks, from low-level motion control to high-level task planning and decision-making

Robot operating system (ROS)

• The Robot Operating System (ROS) is an open-source framework for robot software development
• ROS provides a set of libraries, tools, and conventions that simplify the process of creating complex robotic systems
• It is based on a publish-subscribe architecture, where nodes (software modules) communicate with each other through messages published on topics
• ROS supports a wide range of programming languages (C++, Python) and has a large community and ecosystem of packages for various robotic applications

Robot programming languages

• Various programming languages are used for robot programming, depending on the level of abstraction and the specific requirements of the application
• C++ is commonly used for low-level control and real-time applications due to its performance and direct access to hardware
• Python is popular for high-level scripting, prototyping, and rapid development, thanks to its simplicity and extensive libraries
• MATLAB and Simulink are widely used in academia and research for robot modeling, simulation, and control design
• Domain-specific languages, such as VAL (Variable Assembly Language) or RAPID, are used for industrial robot programming and provide a higher level of abstraction

Simulation environments

• Simulation environments are software tools that allow developers to model, simulate, and test robot systems in a virtual environment before deploying them on physical robots
• They provide a safe and cost-effective way to validate robot designs, algorithms, and control strategies
• Popular robot simulation environments include Gazebo (integrated with ROS), V-REP (Virtual Robot Experimentation Platform), and Webots
• These environments support physics-based simulation, sensor modeling, and realistic rendering, enabling developers to create accurate and detailed simulations of robotic systems

Industrial automation

• Industrial automation refers to the use of robots, control systems, and information technologies to automate manufacturing processes and improve efficiency, quality, and safety
• It involves the integration of various components, such as robots, sensors, actuators, and control systems, to create automated production lines and factories

Automated manufacturing systems

• Automated manufacturing systems are designed to perform various tasks in the production process, such as material handling, machining, assembly, and inspection
• They consist of robots, conveyor systems, automated guided vehicles (AGVs), and computer numerical control (CNC) machines
• These systems are programmed to execute specific sequences of operations and can adapt to changes in product designs or production volumes
• Examples of automated manufacturing systems include flexible manufacturing systems (FMS), which can handle a variety of parts and products, and transfer lines, which are dedicated to high-volume production of a single product

Programmable logic controllers (PLCs)

• Programmable Logic Controllers (PLCs) are industrial computers used to control and automate manufacturing processes
• They are designed to be robust, reliable, and capable of operating in harsh industrial environments
• PLCs execute programs in a cyclic manner, reading inputs from sensors and switches, processing the control logic, and updating outputs to actuators and machines
• They are programmed using ladder logic, a graphical programming language that represents control logic as a series of relay contacts and coils
• PLCs communicate with other devices, such as robots, sensors, and human-machine interfaces (HMIs), using industrial communication protocols (Modbus, Profibus, EtherNet/IP)

Supervisory control and data acquisition (SCADA)

• Supervisory Control and Data Acquisition (SCADA) systems are used to monitor and control large-scale industrial processes, such as manufacturing plants, power grids, and water distribution systems
• They provide a centralized view of the entire system, allowing operators to monitor real-time data, detect anomalies, and issue control commands
• SCADA systems consist of remote terminal units (RTUs) or PLCs that collect data from field devices, a communication network that transmits the data, and a central supervisory computer that processes and displays the information
• They often incorporate graphical user interfaces (GUIs) and alarm systems to alert operators of critical events or deviations from normal operating conditions
• SCADA systems enable remote monitoring and control of industrial processes, improving efficiency, safety, and decision-making

Mobile robotics

• Mobile robotics deals with the design, control, and applications of robots that can move and navigate in their environment
• Mobile robots are used in a wide range of applications, such as exploration, surveillance, transportation, and service robotics

Wheeled vs legged robots

• Wheeled robots are the most common type of mobile robots, using wheels or tracks for locomotion
• They are relatively simple to design and control, and can achieve high speeds and efficiency on flat and smooth surfaces (differential drive robots, car-like robots, omnidirectional robots)
• Legged robots, on the other hand, use articulated legs for locomotion, mimicking the movement of animals or humans
• They can navigate uneven and rough terrains, climb stairs, and adapt to different environments (biped robots, quadruped robots, hexapod robots)
• Legged robots are more complex to design and control than wheeled robots, requiring advanced algorithms for gait generation, balance, and stability control

Localization and mapping

• Localization is the process of estimating a mobile robot's position and orientation within its environment
• It is a crucial capability for autonomous navigation, as the robot needs to know where it is to plan its motions and actions
• Common localization techniques include odometry (estimating motion based on wheel encoders), inertial navigation (using IMUs), and landmark-based methods (detecting and matching visual or geometric features)
• Mapping refers to the process of creating a representation of the robot's environment, such as an occupancy grid map or a topological map
• Simultaneous Localization and Mapping (SLAM) is a popular approach that allows a robot to build a map of an unknown environment while simultaneously localizing itself within that map

Autonomous navigation

• Autonomous navigation enables mobile robots to plan and execute paths to reach desired goals while avoiding obstacles and adapting to changes in the environment
• It involves several components, such as perception (sensing and interpreting the environment), localization, mapping, path planning, and motion control
• Navigation strategies can be classified into reactive (making decisions based on current sensor data) and deliberative (planning paths based on a global map)
• Examples of autonomous navigation algorithms include potential field methods, graph-based search (A*, D*), and sampling-based planners (RRT