Foundations of Robotics and Control Systems

Introduction

Robotics and control systems combine mechanical engineering, electrical engineering, computer science, sensing, and decision-making. A robot is not merely a programmable machine. It is an integrated system that perceives a physical environment, computes an action, and produces controlled motion or force. Control engineering provides the principles required to make that action accurate, stable, safe, and repeatable.

Industrial robots assemble vehicles, weld structures, package products, inspect components, and handle hazardous materials. Mobile robots transport goods, agricultural robots monitor crops, and medical robots assist surgeons. Across these applications, the same foundational questions arise: What is the system expected to do? What variables must be measured? How many independent motions are possible? How should commands be generated? How can errors be detected and corrected?

This chapter establishes the terminology and models needed for later study of robot kinematics, dynamics, sensors, actuators, trajectory planning, and feedback control.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Define the principal terms used in robotics and control engineering.
2. Classify robots according to structure, mobility, actuation, and application.
3. Determine and interpret the degrees of freedom of a robotic mechanism.
4. Distinguish open-loop, closed-loop, supervisory, and hierarchical automation architectures.
5. Explain feedback, error, stability, accuracy, repeatability, resolution, and response time.
6. Identify major industrial applications and select an appropriate robot type for a task.
7. Recognize safety, ethical, and implementation issues in industrial automation.

1. Systems, Robots, and Control

1.1 System terminology

A system is a set of interacting components organized to achieve a purpose. In a robotic system, the components may include a mechanical structure, motors, transmissions, sensors, a controller, software, a power supply, and a human–machine interface.

A plant is the physical process being controlled. For a robot, the plant usually includes the links, joints, actuators, gearboxes, and payload. A controller computes commands intended to make the plant follow a desired behavior. An actuator converts energy into physical action; examples include electric motors, hydraulic cylinders, and pneumatic actuators. A sensor measures a physical quantity such as position, velocity, force, temperature, or vision-based scene information.

The input is a signal applied to a system, while the output is the resulting measured or observable response. A desired output is called the reference, command, or setpoint. The difference between the reference and the measured output is the error:

Here, r(t) is the reference, y(t) is the measured output, and e(t) is the error. The equation assumes that the reference and output describe the same variable and use compatible units. For a position-control system, both may be measured in radians or metres.

A disturbance is an unwanted influence, such as friction variation, vibration, payload change, electrical noise, or an external force. A transducer converts a physical quantity into a usable signal; a rotary encoder, for example, converts shaft motion into electrical pulses or digital position data.

A robot’s work envelope or workspace is the set of positions, and sometimes orientations, that its end-effector can reach. The end-effector is the device attached to the robot wrist, such as a gripper, welding torch, drill, spray gun, or suction tool.

1.2 What distinguishes a robot?

A conventional automated machine often performs a fixed sequence using dedicated mechanisms. A robot generally has several programmable axes and can execute different motions or tasks by changing software, tooling, or task parameters. The International Organization for Standardization commonly characterizes an industrial robot as an automatically controlled, reprogrammable, multipurpose manipulator programmable in three or more axes.

This definition highlights four ideas:

• Automatic control: operation does not require continuous human motion commands.
• Reprogrammability: the sequence or motion can be changed without rebuilding the entire machine.
• Multipurpose capability: the same platform can perform more than one related task.
• Multiple axes: the mechanism has several independently controllable motions.

Not every automated device is a robot. A single-axis indexing table may be highly automated but lacks the versatility normally associated with robotics. Conversely, a robot may form only one subsystem within a larger automated production cell.

2. Robot Classification

Classification helps engineers compare mechanisms and select a suitable architecture. No single classification is sufficient; robots may be classified simultaneously by geometry, mobility, actuation, control method, and application.

2.1 Classification by mechanical configuration

Cartesian robots use three mutually orthogonal linear axes, often denoted X, Y, and Z. Their workspace is approximately rectangular. They offer simple kinematics, high positional accuracy, and straightforward programming. Gantry systems used for machining, palletizing, and large-scale handling are common examples.

Cylindrical robots combine rotary motion about a vertical axis with linear motions. Their workspace is approximately cylindrical. They have historically been used for machine loading and assembly, although many applications have shifted to articulated robots.

Spherical or polar robots use rotary joints combined with a radial extension. Their workspace resembles a spherical sector. They are historically important but less common in modern general-purpose manufacturing.

SCARA robots—Selective Compliance Assembly Robot Arms—usually have two parallel rotary joints for planar motion, a vertical linear axis, and sometimes a wrist rotation. They are compliant in the horizontal plane but relatively rigid vertically. This makes them effective for high-speed insertion, electronics assembly, and packaging.

Articulated robots contain rotary joints arranged like a human arm. Six-axis models can normally control three-dimensional position and three-dimensional orientation of the tool. Their flexibility makes them dominant in welding, painting, assembly, and material handling.

Parallel robots support the moving platform through multiple kinematic chains. Delta robots are a well-known example. They can achieve high speed and low moving mass, although their workspace and singularity behavior require careful analysis.

2.2 Classification by mobility

A fixed-base manipulator is anchored to a structure and operates within a defined workspace. A mobile robot can move its base through the environment. Mobile robots may use wheels, tracks, legs, or hybrid locomotion. An autonomous mobile robot typically combines localization, mapping, perception, planning, and motion control.

A collaborative robot, or cobot, is designed to operate in proximity to people under specified safety conditions. Collaboration does not mean that all protective measures can be removed. Risk assessment, speed and force limits, workspace design, tool selection, and validated safety functions remain essential.

2.3 Classification by actuation

Electric actuation is dominant in modern industrial robots because motors are clean, efficient, controllable, and compatible with digital drives. Hydraulic actuation offers high force and power density and remains useful in heavy-duty or outdoor applications. Pneumatic actuation is inexpensive and fast for simple motion, but compressed air is relatively inefficient and pneumatic systems are less suitable for precise continuous-position control.

2.4 Classification by control capability

Robots may be described as having:

• Point-to-point control, where only selected target positions matter.
• Continuous-path control, where the entire trajectory between points is controlled.
• Force or impedance control, where contact force or dynamic interaction is regulated.
• Adaptive control, where controller behavior changes in response to estimated system properties.
• Autonomous control, where perception and planning determine actions with limited direct instruction.

The appropriate category depends on the task. Spot welding may require accurate point-to-point positioning, whereas spray painting requires smooth continuous-path motion.

3. Degrees of Freedom and Robot Configuration

3.1 Meaning of degrees of freedom

A degree of freedom (DOF) is an independent coordinate required to specify the configuration of a mechanical system. A rigid body moving freely in three-dimensional space has six degrees of freedom: translation along X, Y, and Z, and rotation about those three axes. These are often represented as three Cartesian coordinates and three orientation angles.

A robot’s joint variables describe its internal configuration. Revolute joints contribute an angular variable, usually measured in radians or degrees. Prismatic joints contribute a linear displacement, usually measured in metres or millimetres.

The number of mechanical joints does not always equal the number of independent degrees of freedom. Mechanical constraints, coupled joints, or redundant joints can alter the relationship. For a serial robot with independently actuated, independent joints, the DOF is normally equal to the number of independent joint variables.

A robot with six independent DOF can generally place and orient a rigid tool arbitrarily in three-dimensional space, provided that the requested pose lies within the workspace and the configuration is not singular. Robots with fewer than six DOF cannot generally achieve every arbitrary position and orientation. Robots with more than six DOF are redundant for general spatial pose control and can use the additional freedom to avoid obstacles, limits, or singularities.

3.2 Mobility of mechanisms

For a mechanism, mobility can be estimated using the Grübler–Kutzbach criterion. For a spatial mechanism with lower pairs, a commonly used form is:

Here, M is the mobility, n is the number of links including the fixed link, j is the number of joints, and fᵢ is the number of relative freedoms permitted by joint i. The expression assumes ideal rigid links, correctly identified constraints, and independent joint constraints. Special geometries or dependent constraints can make the simple formula misleading.

For planar mechanisms, the corresponding form is:

The coefficient changes from 6 to 3 because a planar rigid body has three possible motions: two translations and one rotation.

Worked Example: Planar four-bar mechanism

Consider a planar four-bar mechanism with four links, including the fixed link, and four revolute joints. Each revolute joint permits one relative freedom, so fᵢ = 1. Therefore:

The mechanism has one degree of freedom. One input link is sufficient to determine the positions of the remaining links, subject to assembly configuration and mechanical limits.

3.3 Configuration space and singularities

The configuration space is the set of all allowable joint-variable combinations. A robot may have a large Cartesian workspace but still contain inaccessible regions or configurations where control becomes difficult.

A singularity is a configuration in which the robot loses instantaneous motion capability in one or more Cartesian directions, or in which very large joint velocities would be required to produce a desired end-effector velocity. Singularities are related to the Jacobian matrix, which maps joint velocities to end-effector velocities:

Here, ẋ is the end-effector velocity vector, J(q) is the Jacobian evaluated at joint configuration q, and q̇ is the vector of joint velocities. This relationship is a local velocity model and assumes differentiable kinematics.

4. Automation Architectures

4.1 Open-loop control

In an open-loop system, the controller issues commands without measuring the output to determine whether the command achieved its objective. A timed conveyor sequence is a simple example. Open-loop control can be inexpensive and adequate when the plant is predictable, disturbances are small, and accuracy requirements are modest.

Its principal weakness is lack of automatic correction. If a motor stalls or a part is misplaced, the controller may continue operating as though the task were successful.

4.2 Closed-loop control

A closed-loop system measures the output and uses the error to modify the command. A position servo may compare commanded joint angle with encoder-measured angle and increase or decrease motor effort accordingly.

The feedback loop consists conceptually of a reference, comparator, controller, actuator, plant, sensor, and feedback path. Negative feedback is normally used because it tends to reduce error, reject disturbances, and reduce sensitivity to plant variation. Excessive gain, delays, poorly filtered measurements, or unmodeled dynamics can nevertheless produce oscillation or instability.

For a simple proportional controller:

where u(t) is the control signal, Kₚ is proportional gain, and e(t) is error. This controller is useful when immediate correction is desired, but it may leave a nonzero steady-state error under constant load.

A proportional–integral–derivative controller is commonly written:

Here, Kᵢ is integral gain, K_d is derivative gain, and τ is an integration variable. The integral term can remove steady-state error for suitable systems, while the derivative term can improve anticipation of changing error. In practice, derivative action is filtered because measurement noise is amplified by differentiation, and integral action is limited to prevent windup when actuators saturate.

4.3 Hierarchical and distributed automation

Industrial systems are often organized into levels:

1. Field level: sensors, actuators, drives, and safety devices.
2. Cell level: robot controllers, programmable logic controllers, fixtures, and machine tools.
3. Supervisory level: production monitoring, scheduling, data collection, and human–machine interfaces.
4. Enterprise level: manufacturing execution, quality systems, inventory, and business planning.

A hierarchical architecture clarifies responsibilities, but modern factories increasingly use distributed networks and edge computing. Real-time motion control remains close to the machine, while higher-level analytics may run on industrial computers or cloud-connected platforms. Time synchronization, cybersecurity, deterministic communication, and failure containment are important design concerns.

5. Performance Requirements

Robot performance must be expressed in measurable engineering terms rather than general claims such as “fast” or “accurate.”

Accuracy is the closeness of the achieved position or orientation to the commanded value. Repeatability is the closeness of repeated results to one another. A robot can be highly repeatable but inaccurate if its calibration is incorrect. Resolution is the smallest command or measurement increment that the system can distinguish. Settling time is the time required for the response to enter and remain within a specified tolerance band around the final value.

Rise time characterizes how quickly a response moves from an initial level toward its final level. Overshoot is the amount by which the response exceeds its final value. Bandwidth describes the frequency range over which a system can respond effectively. Payload is the permitted carried mass under stated conditions; it is not equivalent to maximum static load in every posture.

Other requirements include cycle time, reach, stiffness, backlash, energy consumption, noise, thermal capacity, reliability, maintainability, and safety integrity. These requirements can conflict. Increasing speed may reduce settling quality, while increasing payload may reduce acceleration. A suitable design therefore results from trade-off analysis rather than maximizing one metric.

For a simple first-order response, a time constant τ indicates the characteristic response speed. The normalized step response is:

This model assumes a stable first-order system with a unit step input and zero initial output. At t = τ, the response reaches approximately 63.2% of its final value. Real robot axes often exhibit higher-order behavior, friction, saturation, flexible modes, and digital sampling, so this equation is a conceptual approximation rather than a complete robot model.

6. Sensors, Actuators, and the Information Chain

A control system is only as reliable as its sensing and actuation chain. Common robot sensors include:

• Rotary encoders and resolvers for joint position.
• Tachometers or differentiated encoder signals for velocity.
• Force–torque sensors for contact interaction.
• Proximity and photoelectric sensors for part detection.
• Cameras and three-dimensional vision systems for localization and inspection.
• Current and temperature sensors for actuator monitoring.

Sensors have accuracy, resolution, noise, latency, calibration drift, and operating limits. A camera may provide rich spatial information but introduce processing delay and sensitivity to lighting. An encoder may provide precise angular measurement but cannot directly determine whether a gripper has securely grasped an object.

Actuator selection depends on torque, speed, duty cycle, stiffness, efficiency, controllability, and environmental conditions. Motor torque and angular acceleration are related through the dynamic balance:

Here, τₘ is motor torque, τ_L is load torque, τ_f is friction torque, J is equivalent rotational inertia, and θ̈ is angular acceleration. All torques are in newton-metres, inertia is in kilogram-metres squared, and angular acceleration is in radians per second squared. The equation assumes a lumped rotational model; flexible links, gear elasticity, and coupled multibody effects require a more complete model.

7. Major Industrial Applications

Material handling includes picking, sorting, palletizing, machine tending, and packaging. The main requirements are throughput, reliable gripping, collision avoidance, and integration with conveyors and sensors.

Welding uses robots for repeatable torch trajectories and consistent process parameters. Arc welding requires control of travel speed, torch orientation, voltage, current, and often seam tracking. Spot welding requires accurate positioning, electrode force, and timing.

Painting and coating benefit from consistent speed, standoff distance, and spray orientation. Robots also reduce human exposure to toxic solvents and explosive atmospheres, although the cell must be appropriately certified and ventilated.

Assembly may require vision, compliance, force control, and precise insertion. Rigid position control alone can cause jamming when parts are not perfectly aligned. Mechanical compliance or controlled impedance allows the robot to accommodate small errors.

Machining and finishing applications include drilling, milling, grinding, deburring, and polishing. These tasks demand structural stiffness, path accuracy, force regulation, and compensation for tool wear.

Inspection uses cameras, laser scanners, tactile probes, and noncontact sensors. Robot motion provides repeatable sensor placement, while software interprets measurements against quality criteria.

Robots are also used in foundries, pharmaceuticals, food processing, logistics, agriculture, construction, and hazardous-environment operations. Application suitability depends on process variation, production volume, safety, integration cost, and the value of flexibility.

8. Applied Case: Robotic Palletizing Cell

Background

Consider an illustrative medium-volume packaging plant that produces cartons in several sizes. Operators currently lift cartons from a conveyor and stack them on pallets. The task is repetitive, physically demanding, and subject to variation in carton orientation.

Problem

The plant requires higher throughput, reduced ergonomic risk, and the ability to change pallet patterns without extensive mechanical modification. A solution must identify cartons, pick them reliably, place them in a stable pattern, and stop safely when a person enters the cell.

Analysis

A six-axis articulated robot offers sufficient reach and orientation flexibility, but a four-axis palletizing robot may provide higher speed if the task needs only planar translation, vertical movement, and wrist rotation. The selection depends on carton orientation, pallet accessibility, and required cycle time.

A vacuum gripper is attractive for sealed cartons with suitable surfaces. However, porous, damaged, or irregular cartons may require mechanical fingers or a hybrid gripper. A vision sensor can identify carton position and orientation, while conveyor encoders provide motion synchronization.

The cell requires a closed-loop sequence at several levels. Sensors confirm carton presence and gripper vacuum. The robot controller follows a programmed trajectory. A programmable logic controller coordinates conveyor, robot, pallet dispenser, and safety devices. Safety-rated scanners or interlocked guarding prevent hazardous operation during human access.

Approach and solution

The proposed cell uses a robot on a fixed pedestal, an indexed conveyor, a camera above the pickup zone, a vacuum gripper with pressure sensing, and a safety-rated perimeter. Pallet patterns are stored as parameterized recipes. The controller verifies that each target location is within the robot workspace and that the pallet height remains within payload and reach limits.

The sequence is:

1. Detect and localize a carton.
2. Check its dimensions and orientation.
3. Move the robot along a collision-free approach path.
4. Apply vacuum and verify grip pressure.
5. Lift and transport the carton.
6. Place it at the recipe-defined position.
7. Confirm release and update the pallet state.
8. Stop or request intervention if a sensor check fails.

Outcome

The illustrative system improves consistency and reduces manual lifting, but its success depends on more than robot selection. Conveyor timing, carton quality, gripper maintenance, guarding, recipe validation, and operator training are equally important. If carton surfaces vary greatly, a vision and gripping upgrade may be more valuable than simply purchasing a faster robot.

Learning points

The case demonstrates that robotics is a systems-engineering discipline. Mechanical reach, sensing, control, software, safety, production logistics, and human factors must be designed together. Automation should be evaluated using total system performance, not isolated robot specifications.

9. Safety, Ethics, and Contemporary Developments

Robotic systems can create crushing, impact, cutting, thermal, electrical, and unexpected-start hazards. Risk assessment should identify hazards, estimate risk, implement protective measures, and verify that safeguards work under foreseeable faults. Emergency stops, protective interlocks, safe speed monitoring, restricted access, and validated safety-rated control functions are common measures. Applicable national regulations and standards must be consulted; safety is not achieved by software assumptions alone.

Ethically, automation raises questions about employment transitions, worker surveillance, algorithmic decision-making, accessibility, and responsibility when an autonomous system behaves unexpectedly. Engineers should preserve traceability, document operating limits, provide training, and design for safe human override where appropriate.

Contemporary developments include digital twins, machine-learning-based inspection, collaborative manipulation, additive manufacturing, autonomous mobile robots, and cloud-connected production analytics. Artificial intelligence can improve perception and scheduling, but it introduces risks involving dataset bias, unpredictable edge cases, cybersecurity, and explainability. Safety-critical motion should not depend on an unvalidated statistical model without independent limits and fail-safe mechanisms.

Best practice is to begin with a clearly specified task, measurable requirements, and a structured risk assessment. Engineers should then select the simplest architecture that meets those requirements, validate performance under realistic loads and disturbances, document calibration and maintenance procedures, and monitor the system throughout its life cycle.

Chapter Summary

Robotics integrates mechanisms, actuators, sensors, controllers, software, and human or supervisory systems to produce programmable physical action. Essential terminology includes plant, controller, actuator, sensor, reference, error, disturbance, end-effector, workspace, payload, accuracy, and repeatability.

Robots can be classified by geometry—Cartesian, cylindrical, SCARA, articulated, and parallel—by mobility, actuation, control capability, and application. Degrees of freedom describe independent configuration variables; six independent spatial DOF are generally required to specify arbitrary tool position and orientation. Mechanism mobility can be estimated using the Grübler–Kutzbach criterion, subject to its assumptions.

Open-loop systems do not measure output for correction, whereas closed-loop systems use feedback to reduce error and reject disturbances. Proportional, integral, and derivative actions provide basic control functions, but gain, delay, saturation, noise, and stability must be considered. Robot performance includes accuracy, repeatability, resolution, response time, payload, reach, stiffness, reliability, and safety.

Industrial applications include handling, welding, painting, assembly, machining, inspection, logistics, and hazardous-environment work. Successful automation requires integrated design of mechanics, sensing, control, communication, safety, software, maintenance, and human factors. AI and connected technologies expand capability but do not remove the need for engineering validation and risk management.

Key Terms

System: Interacting components organized to achieve a purpose.

Plant: The physical process or mechanism being controlled.

Controller: A device or algorithm that computes control actions.

Actuator: A component that converts energy into mechanical action.

Sensor: A device that measures a physical variable.

Reference: The desired value or trajectory for a controlled variable.

Error: The difference between reference and measured output.

Feedback: Returning measured output information to influence control action.

End-effector: The tool or device attached to a robot wrist.

Workspace: The set of positions and orientations reachable by a robot.

Degree of freedom: An independent coordinate needed to specify configuration.

Configuration space: The set of allowable joint-variable combinations.

Singularity: A configuration involving loss of instantaneous motion capability or problematic velocity requirements.

Accuracy: Closeness of an achieved result to the commanded result.

Repeatability: Closeness of repeated results to one another.

Resolution: The smallest distinguishable command or measurement increment.

Payload: The permitted carried mass under specified operating conditions.

Open-loop control: Control without output measurement for automatic correction.

Closed-loop control: Control that uses measured output to modify the input.

Collaborative robot: A robot designed for controlled operation near people under specified safety conditions.

Review Questions

Short Answer Questions

1. Distinguish a plant, controller, actuator, and sensor in a robotic system.
2. What is meant by a robot’s degree of freedom?
3. State two differences between accuracy and repeatability.
4. Why is negative feedback commonly used in robot control systems?
5. Name three industrial robot configurations and one typical application of each.

Long Answer Questions

1. Explain the structure and operation of an industrial closed-loop robot control system, including the roles of reference, error, controller, actuator, plant, and sensor.
2. Compare Cartesian, SCARA, articulated, and parallel robots in terms of workspace, strengths, limitations, and applications.
3. Explain the Grübler–Kutzbach mobility criterion and discuss the assumptions that limit its use.
4. Discuss the differences among proportional, integral, and derivative control actions, including practical problems such as noise and actuator saturation.
5. Describe the major engineering, safety, and organizational factors that should be considered when introducing a robotic palletizing cell.

Analytical / Application Questions

1. A planar mechanism has four links and four one-degree-of-freedom revolute joints. Use the planar mobility equation to determine its mobility and interpret the result.
2. A manufacturing task requires high-speed insertion of small components with vertical force application but limited horizontal stiffness. Select an appropriate robot configuration and justify your choice.
3. A robot is highly repeatable but places every part 4 mm away from the commanded location. Diagnose the likely performance issue and propose engineering actions to improve the process.

Multiple Choice Questions with Answers

Question 1

Which component converts a control signal or supplied energy into physical motion?

A. Sensor

B. Actuator

C. Reference

D. Comparator

Correct Answer: B. Actuator

Question 2

A robot with three orthogonal linear axes is generally classified as a:

A. Cartesian robot

B. SCARA robot

C. Parallel robot

D. Polar robot

Correct Answer: A. Cartesian robot

Question 3

For a free rigid body moving in three-dimensional space, how many degrees of freedom are required to specify its pose?

A. 3

B. 4

C. 6

D. 9

Correct Answer: C. 6

Question 4

In a closed-loop control system, the error is normally defined as:

A. The sum of plant and actuator outputs

B. The reference minus the measured output

C. The controller gain divided by the input

D. The disturbance minus the actuator command

Correct Answer: B. The reference minus the measured output

Question 5

Which robot is particularly associated with fast planar assembly and vertical insertion?

A. SCARA robot

B. Spherical robot

C. Tracked mobile robot

D. Hydraulic excavator

Correct Answer: A. SCARA robot

Question 6

A robot can be highly repeatable but inaccurate when it:

A. Produces scattered results centered on the target

B. Produces consistent results offset from the target

C. Has no measurable output

D. Uses a closed-loop controller

Correct Answer: B. Produces consistent results offset from the target

Question 7

For a planar four-bar mechanism with four links and four one-degree-of-freedom joints, the mobility is:

A. 0

B. 1

C. 2

D. 4

Correct Answer: B. 1

Question 8

Which control action is most directly intended to eliminate steady-state error for a suitable system?

A. Integral action

B. Derivative action

C. Feedforward measurement noise

D. Dead-zone action

Correct Answer: A. Integral action

Question 9

Which of the following is a principal function of a robot end-effector?

A. To define the factory’s electrical frequency

B. To interact physically with the workpiece or environment

C. To replace every system sensor

D. To determine the number of links in the robot

Correct Answer: B. To interact physically with the workpiece or environment

Question 10

Why must collaborative robot applications still include risk assessment and validated safeguards?

A. Because collaboration automatically increases payload

B. Because proximity to people eliminates all hazards

C. Because speed, force, tooling, faults, and workspace conditions can still create hazards

D. Because collaborative robots cannot use feedback

Correct Answer: C. Because speed, force, tooling, faults, and workspace conditions can still create hazards