Advanced Industrial Control Systems: PLC, Robotics & Sensors

Introduction to Industrial Control Systems

Control systems are fundamental to industrial automation, with a primary focus on Distributed Control Systems (DCS). This document covers the core components, variables, and applications of these systems.

Learning Outcomes

• Describe the components of a control system.
• Define the three types of variables associated with a control system.
• Provide examples of common control systems.
• Explain distributed control systems and their industrial applications.
• List and define components of DCS, including SCADA, communication, and alarms.

Components of an Automation System

A modern automation system is composed of several interconnected components:

• Control System: The brain of the operation, processing inputs and making decisions.
• SCADA (Supervisory Control and Data Acquisition): The interface for human operators to monitor and control the process.
• Communications: The network that allows all components to exchange data.
• Sensors: Devices that measure physical properties of the system.
• Actuators: Devices that perform physical actions based on commands from the control system.
• Physical System: The actual machinery and process being controlled.

What is a Control System?

A control system provides a desired output response by controlling an output variable. It is a vast field of engineering. A common example is a thermostat in a home, which maintains a desired temperature by controlling the heating or cooling system.

Important Terminology

• Controlled Variables: These variables, also called output variables, quantify the performance or quality of the final product. Example: The water temperature in a shower.
• Manipulated Variables: These variables are adjusted dynamically to keep the controlled variables at their desired setpoints. Example: The positions of the hot and cold water valves in a shower.
• Disturbance Variables: Also known as “load” variables, these are input variables that can cause controlled variables to deviate from their setpoints. Example: Fluctuations in hot or cold water pressure and temperature.

Types of Process Control

Continuous Process Control

Continuous control involves an analog (continuous) output to an actuator, making many small variations to maintain the controlled variable. It is implemented through open or closed-loop controllers using elements like scaling coefficients, integrators, and differentiators (e.g., PID controllers). This type of control is mathematically complex, often involving Laplace transforms and state-space models. Example: Regulating a motor's speed to maintain a set output pressure.

Discontinuous Process Control (Discrete Control)

Discrete control uses a discrete (on/off) output to an actuator, often called a “bang-bang” controller. It is implemented through state machines and can be mathematically simpler than continuous control. Example: The operation cycle of a dishwasher.

Industrial Applications of Control Systems

• Conveyance Systems: Most controllers use a mix of continuous and discontinuous control to regulate conveyor speed, route objects, validate quality, and manage safety.
• Wastewater Treatment Plants: Numerous control systems are deployed in every subsection of a treatment plant to manage complex chemical and physical processes.
• Robotics: Control systems are essential for multi-axis motor controllers, stage controllers, grip controllers, and dynamic positioning systems.

Distributed Control Systems (DCS)

A Distributed Control System (DCS) is a network of autonomous controllers distributed throughout a system. This course focuses primarily on PLC-based automation control systems.

Programmable Logic Controller (PLC)

PLCs are rugged, industrial computers that form the backbone of many automation systems. They are designed for reliable control in harsh environments.

Remote Telemetry Units (RTUs)

Also known as Remote Terminal Units, RTUs are similar to PLCs but are designed for extremely remote or hostile environments. They are extra rugged but often not compatible with the common IEC 61131-3 programming standards.

FPGA/MCU-Based Systems

• MCUs (Microcontrollers): These are application-specific, low-cost computing devices for general-purpose, low-speed control. They require custom implementation and are typically programmed in C/C++.
• FPGAs (Field Programmable Gate Arrays): FPGAs are used for application-specific, highly versatile, and complex control applications. They require custom hardware and software, making development expensive, but are ideal for high-speed or high I/O applications.

PC-Based Control

Computer applications like LabVIEW and MATLAB/Simulink provide versatile and powerful control functions. However, the complexity of general-purpose computers increases the probability of failure compared to dedicated industrial controllers.

Programmable Automation Controller (PAC)

Sometimes called industrial PCs, PACs combine the ruggedness of PLCs with the computational and software capabilities of a PC. They support advanced applications like Simulink projects, computer vision, and other machine learning or AI tasks.

SCADA, Communications, and Alarms

SCADA Introduction

A Supervisory Control and Data Acquisition (SCADA) system is used for plant oversight, serving as the management interface between operators and equipment.

User Experience (UX) in SCADA

A well-designed SCADA interface is crucial for efficient operation. Key considerations include:

• Monitoring: Prioritize critical measurements and highlight alarms.
• Control: Provide access to all required setpoints with appropriate user access levels.
• Layout: Organize the interface by process subsystem, often mimicking a Piping and Instrumentation Diagram (P&ID).
• Consistency: Use consistent colors and symbols (e.g., Red for Stop, Green for Go, Blinking for an alert).

Communications

Industrial plants typically use many individual control subsystems that must communicate sensor data, control variables, and system status. This communication can span multiple buildings and even hundreds of kilometers.

Alarms

Alarms notify users of critical control system events. Important considerations include:

• Notification: Local or remote call-out? Audible or visual?
• Acknowledgement: Who acknowledges an alarm and how?
• Intervention: Is manual intervention required?
• Logging and Escalation: Are alarms logged, and is there an escalation procedure?

Timely and effective alarms can significantly reduce plant downtime and improve safety.

Cloud and Value-Added SCADA

Modern systems leverage cloud applications for distributed SCADA and Smart Manufacturing, integrating plant-level control with enterprise-level data systems, often visualized as the "automation pyramid."

Control Theory Fundamentals

Control theory provides the mathematical foundation for designing and analyzing control systems. The focus here is on the system-level principles relevant to industrial automation.

Controller Types

• Open-Loop Controllers: These controllers use an actuating device to control a process without using feedback. The system's outputs have no effect on the control inputs. This can be inexact as it cannot correct for changes in the process.
• Closed-Loop Controllers: Also known as feedback controllers, these systems use output measurements to compare with the input (setpoint) and adjust the control action accordingly. This allows the system to self-correct. Closed-loop systems are far more common in industrial applications.
• Multivariable Controllers: These are complex systems with multiple inputs and multiple outputs (MIMO) that manage interconnected processes.

Requirements for a Good Control System

• Accuracy: The system's accuracy is limited by sensor accuracy and the stability of the control system itself.
• Sensitivity: The system should be sensitive to the input command only, not to changes in the plant or process.
• Noise: The design must minimize or manage noise in all variables.
• Bandwidth: A larger frequency bandwidth allows for a faster response.
• Speed: The system should have a high-speed response with good management of transients.
• Oscillations: Stability is critical and is often a trade-off against speed.

System Performance and Analysis

Laplace Domain and Transfer Functions

A transfer function is a mathematical model of a system's output for all possible inputs, assuming the system is linear and time-invariant. The Laplace domain is a powerful tool used to convert complex differential equations (representing system dynamics) into simpler algebraic equations, making it easier to analyze system response and stability.

Frequency Response and Bode Plots

Frequency response analyzes how a system responds to sinusoidal inputs across a range of frequencies. A Bode plot is a standard way to visualize this, showing the system's gain (magnitude) and phase shift as a function of frequency. This is crucial for stability analysis.

Performance Metrics

• Settling Time: The time it takes for the system to reach and stay within a target band around the new setpoint after a change.
• Rise Time: The time taken to go from 10% to 90% of the new setpoint.
• Steady-State Error: The difference between the controlled variable and the setpoint after the system has settled.
• Overshoot: The amount by which the controlled variable exceeds the final setpoint during a transient response.
• Damping: Describes how a system's oscillations decay after a disturbance. Systems can be overdamped (slow, no oscillation), critically damped (fastest response without overshoot), or underdamped (oscillates before settling).

PID Controllers

The Proportional-Integral-Derivative (PID) controller is a classic and widely used feedback control loop mechanism. It covers approximately 90% of industrial automation control scenarios.

Controller Components

A PID controller calculates an error value as the difference between a measured process variable and a desired setpoint. The controller attempts to minimize the error by adjusting a manipulated variable.

Proportional (P) Control

The P component provides an output that is proportional to the current error. It offers a fast response but requires an error to generate a response, which can lead to a permanent steady-state error.

Integral (I) Control

The I component sums the error over time. This term is designed to eliminate steady-state error by continuing to adjust the output as long as an error persists. However, it can introduce overshoot and slow down the response. A common issue is integrator windup, where a large change in setpoint causes the integral term to accumulate a significant error, leading to large overshoot. Anti-windup strategies are used to mitigate this.

Derivative (D) Control

The D component's output is proportional to the rate of change of the error. It is predictive, attempting to reduce overshoot and improve transient stability. However, it is very sensitive to measurement noise and is less commonly used in manually tuned systems.

PID Tuning

Tuning is the process of selecting the optimal coefficients (Kp, Ki, Kd) for the P, I, and D terms. The goal is to achieve fast command tracking and stable regulation with minimal overshoot.

• Trial and Error: A manual method where Kp is increased until oscillation occurs, then Ki is added to remove offset, and finally, Kd is adjusted to speed up the response.
• Ziegler-Nichols Method: A heuristic method where the system is made to oscillate by increasing Kp (now called Ku). The period of oscillation (Tu) and Ku are then used in specific formulas to calculate the PID parameters.

Advanced Controller Types

While PID is dominant, other controller types are used for specific applications.

• Bang-Bang Controllers: A simple on/off controller, often with hysteresis (a deadband) to prevent rapid switching or "chattering."
• Model Predictive Control (MPC): Uses a dynamic model of the process to predict future behavior and calculate an optimal sequence of control actions. It is powerful but computationally intensive.
• Fuzzy Logic Controllers: Excellent for poorly defined systems, using expert-defined rules and "fuzzy" logic to simulate human decision-making.
• Adaptive Controllers: Automatically adjust their parameters in real-time based on a reference model or adaptation law to cope with changing process dynamics.

PLC Hardware and Implementation

Basic PLC Components

• CPU (Central Processing Unit): Stores and executes the PLC program.
• I/O Modules: Connect the PLC to sensors (inputs) and actuators (outputs). Modules can be analog or discrete.
• Power Module: Provides DC voltage for the PLC modules.
• Racks/Buses: A chassis or motherboard that allows for the interconnection of individual PLC modules.

Sensors and Actuators

• Sensors: Devices that detect events or changes in the environment and provide that information to the PLC. They can be indicators (show data locally) or transmitters (send data electronically).
• Actuators: Control endpoints that perform an action, such as motor controllers, solenoids, and alarms. They can be continuous (operating over a range) or discrete (on/off).

Signal Types

• Voltage (0-5V, 0-10V): Simple but susceptible to noise.
• Current (4-20mA): More robust against noise and very common in industrial settings. The 4mA baseline allows the sensor to be loop-powered and indicates a live connection.
• Digital: Communication via an industrial protocol like Modbus.

Latch Principle and Circuits

• Latch Principle: Refers to discrete outputs. A Set command latches an output in the energized (on) state, while a Reset command latches it in the de-energized (off) state.
• Series and Parallel Circuits: Used extensively in automation logic. A series circuit requires all contacts to be closed for current to flow (used for safety permissives). A parallel circuit requires only one contact to be closed (used for OR logic, like a start button or a holding contact).

PLC Programming

The IEC 61131-3 standard defines several programming languages for PLCs, with Ladder Logic being the most common.

Ladder Logic (LD)

Ladder Logic is a graphical programming language that mimics electrical relay circuits. A program consists of rungs of logic that are read from left to right and top to bottom.

• Contacts and Coils: The basic elements are normally open (NO) and normally closed (NC) contacts (inputs) and coils (outputs).
• Tags and Memory: PLCs use memory locations called tags to store discrete (bit) or continuous (integer, real) values.
• Function Blocks: Provide advanced functions like math operations, timers, counters, loops, and data manipulation (e.g., COPY).
• Subroutines: Break up a large program into smaller, manageable pieces. A main routine calls subroutines to execute specific functions, improving organization and reusability.

Alternatives to Ladder Logic

• Function Block Diagram (FBD): A graphical language that connects function blocks to represent logic flow.
• Structured Text (ST): A high-level, text-based language similar to Pascal or C, used for complex algorithms.

State Machines

A state machine is a model of behavior composed of a finite number of states and transitions between those states. They are excellent for controlling sequential processes like a washing machine. There are two main types:

• Moore Machine: The outputs are determined solely by the current state.
• Mealy Machine: The outputs are determined by both the current state and the current inputs.

State machines can be implemented in a PLC using a state variable (a tag) and logic rungs that execute only when the machine is in a specific state.

Programming and Debugging Best Practices

• Structure: Use subroutines to organize code by function (e.g., analog inputs, pump control).
• Documentation: Comment code thoroughly and use clear, descriptive tag names.
• Diagnostics: Design the program with diagnostics in mind. Avoid constructs like for-loops that can be difficult to debug online.
• Going Online: Connecting to a live PLC provides a real-time view of the program's execution, which is invaluable for troubleshooting.

Special PLC Applications

Motion Controllers

Motion controllers are specialized modules or standalone devices that provide semi-autonomous control over motors, often based on feedback from an encoder. They offload the complexity of generating motion profiles from the main PLC, allowing for faster and finer control, especially in multi-axis systems.

Real-Time Control

Real-time control systems are those where the correctness of the system depends not only on the logical result of a computation but also on the time at which the results are produced. They require guaranteed low latency and low jitter (variation in latency). This is critical for applications like CNC machine control, where missing a deadline can cause system failure.

Permissives and Safety

A permissive is a condition or set of conditions that must be met before a piece of equipment is allowed to start or operate. They are a critical part of a system's safety design.

• Electrical Permissives: Hardwired safety circuits (e.g., an emergency stop button) that physically disconnect power and do not rely on the PLC.
• Programmatic Permissives: Logic-based checks within the PLC program (e.g., checking if a tank level is high enough before starting a pump).

HMI and SCADA

Human-Machine Interface (HMI)

An HMI is typically a local display panel that allows an operator to view process variables and perform basic control actions. It communicates directly with a PLC and is designed for a specific machine or process area.

SCADA vs. HMI

SCADA is a more centralized system, typically PC-based, that supervises an entire plant or multiple sites. It offers advanced features like:

• Historians: A database for long-term storage and aggregation of process data.
• Advanced Analytics: Tools for analyzing performance and identifying trends.
• Remote Access: The ability to monitor and control the system from a remote location.
• Cloud Integration: Modern SCADA systems can leverage cloud platforms for data storage, analytics, and lower infrastructure costs.

Industrial Communications

Communication protocols are the languages that allow industrial devices to exchange data. A well-designed automation system relies on a robust and standardized communication architecture.

The Purdue Model

The Purdue Enterprise Reference Architecture is a model for industrial control system security that segments the network into logical levels, from the physical process (Level 0) up to the enterprise business network (Level 5), with firewalls and security zones in between.

Common Protocols

• Modbus (RTU and TCP): A simple, open, and widely supported serial (RTU) and Ethernet (TCP) protocol. It uses a master-slave architecture.
• EtherNet/IP: An industrial Ethernet protocol managed by ODVA and widely used by Rockwell Automation/Allen-Bradley. It is based on the Common Industrial Protocol (CIP).
• OPC UA (Unified Architecture): A modern, secure, and platform-independent standard for data exchange. It acts as a universal translator between different devices and software applications.

Cybersecurity

As industrial systems become more connected, cybersecurity is an essential consideration. Protecting the Operational Technology (OT) network from unauthorized access requires a multi-layered approach, including firewalls, VPNs for remote access, and secure network design.

Industrial Robotics

An industrial robot is an automatically controlled, reprogrammable, multipurpose manipulator programmable in three or more axes.

Robot Types

• Articulated Robots: The most common type, with a “robotic arm” design offering 6 or more degrees of freedom (axes).
• SCARA Robots: Fast and precise for pick-and-place operations, with a cylindrical working envelope.
• Cartesian/Gantry Robots: Move along three linear axes (X, Y, Z), often used for CNC and 3D printing.
• Cobots (Collaborative Robots): Designed to work safely alongside humans, with enhanced sensors and limited power.

Key Components

• Manipulator/End Effector: The “hand” of the robot, which can be a gripper, welder, paint gun, or other tool.
• Controller: The robot's brain, which interprets code and sends motion commands to the motors.

Integration with Automation Systems

Traditionally, robots operated as standalone islands of automation. Modern trends are toward tighter integration with the plant-wide control system. Robot controllers now often support industrial protocols like EtherNet/IP and Modbus/TCP, allowing a PLC to monitor the robot's status and coordinate its actions with other machinery.

Sensors and Transduction

A sensor is a device that detects changes in a physical parameter and converts that information into an electronic signal. This conversion process is called transduction.

Sensor Performance

• Accuracy vs. Precision: Accuracy is how close a measurement is to the true value. Precision is how repeatable the measurements are.
• Resolution: The smallest change in the measured variable that the sensor can detect.
• Bandwidth: The range of frequencies over which the sensor can operate effectively.
• Hysteresis: The difference in output for the same input when approached from opposite directions.
• Drift: A gradual change in the sensor's characteristics over time.

Strain Gauges and Accelerometers

Wheatstone Bridge

The Wheatstone bridge is a circuit used to measure an unknown electrical resistance by balancing two legs of a bridge circuit. It is the basis for many sensors, including strain gauges, because it can detect very small changes in resistance while canceling out common errors like temperature effects.

Strain Gauges

A strain gauge measures mechanical strain. They are used in:

• Load Cells: To measure force or weight.
• Pressure Sensors: A diaphragm deforms under pressure, which is measured by a strain gauge.
• Torque Sensors: To measure twisting force.

Accelerometers

Accelerometers measure acceleration, typically using a small, elastically mounted proof mass. They are used for:

• Vibration Analysis: To detect faults in rotating machinery.
• Shock Detection: To measure impacts.
• Position Tracking: By integrating acceleration twice over time, one can calculate displacement. However, this is prone to significant integration error and drift.

Position Measurement

• Limit and Proximity Switches: Simple discrete sensors that detect the presence or absence of an object at a specific point.
• Linear Encoders: Measure linear displacement along an axis using optical or magnetic principles.
• LVDT (Linear Variable Differential Transformer): A highly accurate and robust transducer for measuring linear displacement.
• Ultrasonic and Radar Sensors: Non-contact methods that measure distance by timing the reflection of an acoustic (ultrasonic) or electromagnetic (radar) wave.
• Float Switches: A simple and reliable method for detecting liquid level in a tank.

Temperature Sensing

• Thermistors: Inexpensive resistors whose resistance changes significantly with temperature. They are sensitive but non-linear.
• RTDs (Resistance Temperature Detectors): More accurate and linear than thermistors, typically made of platinum (e.g., Pt100). They are the standard for precise process temperature measurement.
• Thermocouples: Based on the thermoelectric (Seebeck) effect, where a voltage is produced at the junction of two dissimilar metals. They are self-powered, have a very wide temperature range, but are less accurate than RTDs.

Encoders and Flow Measurement

Flow Measurement

Measuring the rate of fluid movement is critical in process industries.

• Turbines: A mechanical impeller whose rotation speed is proportional to fluid velocity.
• Electromagnetic Meters: Use Faraday's Law to measure the velocity of conductive fluids without obstructing flow.
• Ultrasonic Meters: Non-intrusive meters that measure flow using the Doppler effect or transit time of sound waves.
• Coriolis Meters: Highly accurate devices that directly measure mass flow rate by sensing the Coriolis forces on an oscillating tube.

Rotary Encoders

Rotary encoders track the angular position or speed of a rotating shaft. Quadrature encoding uses two output signals, typically called A and B, which are 90 degrees out of phase. This allows the controller to determine not only the amount of rotation but also its direction.