12. Sensors, Actuators, Mechatronics, and Real-Time Integration

Introduction and Context

Machine dynamics and control become physically meaningful only when a controller can observe a machine, make a decision, and influence its motion. Sensors transform mechanical, thermal, hydraulic, pneumatic, and electrical variables into signals. Signal-conditioning and data-acquisition systems make those signals usable. Controllers compute control actions, while actuators convert commands into force, torque, displacement, flow, or pressure. Mechatronics is the engineering discipline that integrates these elements into a coordinated system rather than treating them as isolated components.

In a modern machine, the quality of the control system depends not only on its mathematical design but also on sensor placement, sampling, quantization, latency, actuator bandwidth, communication reliability, calibration, and fault response. A theoretically stable controller may perform poorly if its feedback signal is delayed or noisy. Conversely, a modest controller can perform effectively when the measurement chain and actuator are properly engineered.

This chapter develops a systems approach to sensor and actuator integration. Particular attention is given to real-time implementation, industrial controllers, programmable logic controllers (PLCs), networks, and hardware-in-the-loop (HIL) simulation. These topics connect the dynamic models studied earlier with the physical implementation of automated machinery.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Classify sensors and actuators according to their operating principles and application requirements.

2. Design a measurement chain including signal conditioning, sampling, conversion, and filtering.

3. Evaluate calibration, accuracy, resolution, noise, drift, bandwidth, latency, and reliability.

4. Select electric, hydraulic, and pneumatic actuators for motion-control applications.

5. Explain the architecture and timing requirements of embedded controllers, PLCs, and industrial networks.

6. Assess the effects of sampling, computational delay, communication delay, and actuator dynamics on closed-loop behavior.

7. Develop an integrated mechatronic architecture and validate it using simulation and HIL methods.

8. Identify safety, cybersecurity, maintenance, and professional responsibilities in automated systems.

From Physical Variable to Control Decision

A measurement system is more than a sensor. It is a chain extending from the physical variable to the control algorithm. For example, a servo drive may use an encoder to measure shaft position, differential line drivers to transmit the signal, a receiver and counter to determine position, a digital filter to suppress disturbances, and a control algorithm to calculate the required motor current. An error at any stage can affect the final motion.

A useful representation is

physical variable → sensing element → transducer → signal conditioning → data acquisition → computational algorithm → command interface → actuator → plant.

The sensing element responds to the variable of interest. A transducer converts that response into an electrical quantity such as voltage, current, resistance, frequency, or digital code. Signal conditioning may include amplification, bridge completion, linearization, isolation, excitation, anti-alias filtering, and protection. Data acquisition then samples and converts the conditioned signal. The controller processes the data and sends a command to an actuator or drive.

This chain should be specified as a complete system. Selecting a highly accurate sensor is not sufficient if the cable is exposed to electromagnetic interference, the converter has inadequate resolution, or the controller samples too slowly.

Sensor Selection for Dynamic Systems

Sensors can be classified by the variable measured, the transduction principle, and the form of output. Position may be measured using potentiometers, optical encoders, magnetic encoders, resolvers, or linear variable differential transformers. Velocity may be obtained directly from a tachometer or inferred by differentiating position, although numerical differentiation amplifies noise. Acceleration is commonly measured with piezoelectric or microelectromechanical-system (MEMS) accelerometers. Force and torque are often measured with strain-gauge bridges, piezoelectric elements, or magnetoelastic devices. Pressure is measured using diaphragm-based piezoresistive, capacitive, or piezoelectric transducers.

For a dynamic application, the important criteria are not limited to nominal accuracy. They include:

• Range: the minimum and maximum values that can be measured.
• Sensitivity: output change per unit input change.
• Resolution: the smallest distinguishable change.
• Accuracy: closeness to the reference value.
• Repeatability: agreement among repeated measurements under the same conditions.
• Linearity: departure from an ideal proportional relationship.
• Hysteresis: dependence of output on the direction of input change.
• Bandwidth: frequency range over which the sensor represents the input adequately.
• Drift: gradual output change caused by time, temperature, ageing, or environment.
• Loading: disturbance of the measured system caused by sensor installation.

Accuracy and resolution are distinct. A high-resolution encoder may repeatedly report small increments while still having an offset or scale error. Similarly, a sensor with excellent static accuracy may be unsuitable for vibration measurement if its bandwidth is too low.

Contact sensors are often inexpensive and simple, but they may suffer from wear and mechanical loading. Non-contact sensors reduce wear and can support high-speed motion, but they may require careful alignment and may be sensitive to magnetic, optical, or environmental conditions. Redundant sensing is used when a single measurement failure could create unacceptable risk. In safety-related applications, the architecture must also consider diagnostic coverage and the possibility of common-cause failure.

Calibration and Uncertainty

Calibration establishes the relationship between sensor output and a reference standard. A practical calibration may identify zero offset, sensitivity, nonlinearity, hysteresis, and temperature dependence. For a linear model,

y = Kx + b + e,

where x is the physical input, y is the measured output, K is sensitivity, b is offset, and e represents random and unmodelled error. Calibration estimates K and b, but it does not eliminate all uncertainty.

A sound calibration procedure defines the reference, range, environmental conditions, loading, number of points, direction of testing, acceptance limits, and recalibration interval. Calibration records should be traceable and associated with a particular instrument and software configuration. In production systems, calibration constants should be protected from unauthorized alteration and version-controlled.

Signal Conditioning and Data Acquisition

Raw sensor signals are rarely suitable for direct use by a controller. A strain gauge, for example, produces a small differential voltage and requires bridge excitation, instrumentation amplification, filtering, and often temperature compensation. A thermocouple requires cold-junction compensation and an appropriate low-level amplifier. Industrial current-loop signals such as 4–20 mA provide improved transmission over long distances and allow open-circuit detection when the lower and upper limits are interpreted appropriately.

Signal conditioning commonly performs five functions. First, it adapts signal magnitude to the input range of the converter. Second, it removes unwanted frequency components. Third, it protects the acquisition hardware against overvoltage, transients, and wiring errors. Fourth, it electrically isolates field equipment from sensitive computation hardware. Fifth, it linearizes or compensates the sensor response.

Sampling and Aliasing

A continuous signal sampled at frequency fs can represent frequencies reliably only when the input bandwidth is restricted below half the sampling frequency. This limit is the Nyquist frequency, fs/2. Frequency components above this limit can be folded into lower frequencies, producing aliasing. Once aliasing has occurred, digital processing cannot recover the original signal.

An anti-alias analog low-pass filter is therefore placed before the analog-to-digital converter. The cutoff frequency must be selected with regard to the desired control bandwidth, sampling rate, filter roll-off, and allowable phase delay. Oversampling provides greater separation between the signal band and the Nyquist limit, making filtering easier.

Quantization introduces an additional error because a converter maps a continuous amplitude to discrete numerical levels. For an ideal converter with N bits and input span Vspan, the quantization step is approximately Vspan/2^N. Increasing resolution does not automatically improve the measurement if electrical noise, reference instability, or sensor drift is larger than one quantization step.

Noise and Filtering

Noise may originate in the sensor, power supply, cabling, grounding, switching devices, or the environment. Shielding, twisted-pair wiring, differential measurement, appropriate grounding, isolation, and physical separation from power cables are often more effective than software filtering alone.

A moving-average filter is simple but introduces delay and attenuates rapid changes. A first-order digital low-pass filter may be represented as

y[k] = αy[k−1] + (1−α)x[k],

where x[k] is the current measurement and α determines smoothing. A larger α gives stronger smoothing but greater delay. In feedback control, that delay can reduce phase margin. Filtering must therefore be designed together with the controller rather than added as an afterthought.

Actuators and the Physical Realization of Control

An actuator converts a command into a physical effect. The actuator may be a motor producing torque, a hydraulic cylinder producing force, a pneumatic valve regulating flow, a solenoid changing position, or a piezoelectric stack producing a small displacement. The selected actuator must satisfy force or torque, speed, stroke, duty cycle, controllability, efficiency, environmental, and safety requirements.

Electric Actuators

DC motors are easily controlled through armature voltage or current and are useful in laboratory and industrial servo systems. Brushless DC and permanent-magnet synchronous motors provide high efficiency and reduced mechanical wear, but require electronic commutation. Induction motors are robust and widely used with variable-frequency drives. Stepper motors can provide convenient open-loop positioning under suitable load conditions, although loss of synchronism is possible when the demanded torque exceeds available torque.

The mechanical behavior of a motor-driven load is often described by

J(dω/dt) = Tm − Tl − Bω,

where J is equivalent inertia, ω is angular velocity, Tm is motor torque, Tl is load torque, and Bω represents viscous resistance. The drive electronics add current-loop dynamics, voltage limits, switching effects, and protection functions. A motor should therefore be selected with its drive, encoder, gearbox, thermal behavior, and mechanical transmission as one system.

Hydraulic and Pneumatic Actuators

Hydraulic actuators provide high force density and are suitable for presses, mobile equipment, and heavy machinery. Their performance depends on pump capacity, valve flow characteristics, compressibility, leakage, pressure variation, and fluid temperature. Servo valves and proportional valves permit closed-loop control, but hydraulic systems require careful contamination control and maintenance.

Pneumatic systems are clean, relatively inexpensive, and capable of rapid motion. However, air compressibility makes precise positioning more difficult, and friction, stick-slip, and pressure variation can produce nonlinear behavior. Pneumatic systems are often used for clamping, pick-and-place operations, and discrete automation rather than high-precision servo positioning.

Actuator Saturation and Protection

Every actuator has limits. Voltage, current, pressure, flow, travel, temperature, and rate limits must be represented in the control design. Saturation can create integral windup: the integral component continues to accumulate error while the actuator is unable to provide the requested command. Anti-windup methods, command limiting, and bumpless transfer reduce recovery time and undesirable transients.

Actuator protection includes overcurrent detection, thermal monitoring, pressure relief, end-of-travel limits, emergency stopping, and safe energy isolation. A control command should never be treated as a guarantee that the physical actuator will produce the requested output.

Embedded Controllers, PLCs, and Industrial Networks

An embedded controller normally includes a processor, memory, input-output interfaces, timers, communication ports, and firmware. Real-time performance means that a task produces its result within a specified deadline, not merely that it executes quickly on average. A control loop with a nominal period of 1 ms is unsuitable if computation sometimes takes 4 ms without a defined handling strategy.

A typical real-time sequence is: acquire inputs, validate and timestamp them, execute the control algorithm, apply limits and safety logic, update outputs, and record diagnostic information. Deterministic scheduling, interrupt priorities, watchdog timers, and fault-handling states are essential. Time stamps are especially important when data from multiple devices arrive at different times.

PLCs are designed for industrial automation, discrete sequencing, and robust field operation. They commonly support ladder logic, function-block diagrams, structured text, and sequential function charts. Motion-capable PLCs can coordinate axes and communicate with drives, while safety PLCs execute certified safety functions. General-purpose embedded controllers may offer more flexible numerical computation and lower latency, but require greater responsibility for software architecture, diagnostics, and lifecycle support.

Industrial networks connect sensors, controllers, drives, remote I/O, and supervisory systems. The design choice involves determinism, bandwidth, topology, synchronization, diagnostics, interoperability, and security. A network that is adequate for slow supervisory data may be unsuitable for tightly synchronized multi-axis control. Distributed clocks and synchronized sampling reduce timing uncertainty in coordinated motion.

Communication delay can be represented approximately as a time delay in the feedback or command path. Delays reduce phase margin and may destabilize a system that was stable in a delay-free model. The engineer should account for sensor conversion time, filtering, processor execution, network transmission, drive update, and actuator response as separate contributions to total latency.

Integrated Mechatronic Design

Mechatronic design begins with the required machine function rather than with a preferred component. The designer defines performance measures such as settling time, positioning error, throughput, force, energy consumption, noise, and allowable fault response. The physical plant, measurement system, controller, actuator, communication architecture, and human interface are then developed concurrently.

A useful workflow is:

1. Define operating modes, requirements, constraints, and hazards.

2. Develop a plant model that includes dominant mechanical, electrical, hydraulic, and actuator dynamics.

3. Select measurable variables and sensor locations based on observability, bandwidth, installation, and reliability.

4. Select actuators and drives using force, torque, speed, duty-cycle, and saturation requirements.

5. Specify signal conditioning, sampling rates, timing, and data formats.

6. Implement control, sequencing, diagnostics, and safety functions as separate but coordinated layers.

7. Verify individual components, integrated subsystems, and complete machine behavior.

8. Validate performance under nominal, boundary, and fault conditions.

Separation of functions is important. A performance controller may regulate position, while an independent safety function monitors overspeed or access doors. Safety should not depend solely on the same software or measurement channel whose failure it is intended to detect.

Real-Time Control and Timing Analysis

Suppose a position controller is designed using a continuous-time model. In implementation, the controller receives a sampled position, filters it, computes a command, sends the command through a network, and waits for the actuator drive to update. The effective loop contains sampling delay, computation delay, zero-order-hold behavior, and actuator dynamics. A common engineering approximation is to treat these effects as an equivalent delay and then verify the approximation using a discrete-time model.

The sampling period should be sufficiently short relative to the fastest relevant closed-loop dynamics, but excessive sampling can waste processor resources and amplify high-frequency measurement noise. The correct choice is determined by bandwidth, required phase margin, computation time, converter performance, and timing jitter.

Jitter is variation in the timing of periodic events. Even when the average sampling period is correct, jitter can produce variable effective delay and degrade coordinated control. Timestamping, hardware triggering, real-time operating systems, and deterministic networks are methods for controlling this problem.

Hardware-in-the-Loop Validation

Hardware-in-the-loop simulation connects real control hardware to a real-time numerical model of the plant. The controller believes it is operating a physical machine, while the HIL computer calculates plant response and returns simulated sensor signals. HIL is valuable before a prototype is available and when testing dangerous or expensive conditions.

A credible HIL setup must match the relevant plant dynamics, I/O timing, scaling, sensor behavior, actuator limits, and fault conditions. If the real plant has a nonlinear valve, backlash, saturation, dead zone, or communication delay, these features should be included when they influence the test objective. The numerical model must execute within its real-time step; a mathematically accurate model that misses deadlines is not a valid real-time plant.

HIL testing can examine sensor disconnection, encoder faults, overspeed, loss of communication, actuator saturation, emergency stops, and abnormal operating sequences. It does not replace physical testing because mechanical compliance, electromagnetic interference, thermal effects, installation errors, and unmodelled wear may not be represented adequately. It is best understood as one level in a verification hierarchy: software-in-the-loop, processor-in-the-loop, HIL, subsystem testing, and full-machine validation.

Illustrative Case Study: Servo Positioning of an Automated Inspection Axis

Background

An illustrative manufacturing system uses a motor-driven linear axis to move a camera across a component. The axis must travel 400 mm, settle within a specified positioning tolerance, and repeat the inspection cycle continuously during an eight-hour shift. The original design used an open-loop stepper motor and a limit switch at each end.

Situation and Problem

At higher speeds, the axis occasionally lost position. The inspection software also reported inconsistent camera locations. Investigation showed that the motor could stall during acceleration, the limit switches had mechanical bounce, and the camera trigger was issued by a supervisory computer with unpredictable communication delay.

Analysis

The system lacked closed-loop position feedback and had no reliable relationship between axis position and image acquisition time. Increasing motor size alone would not solve the timing problem. The engineering team identified four requirements: direct position measurement, controlled acceleration, deterministic trigger timing, and fault detection.

Approach and Solution

A linear optical encoder was installed on the moving axis rather than relying only on motor-step count. The encoder signal was transmitted differentially to reduce interference and was read by a motion controller at a fixed period. A brushless servo motor with a drive-resident current loop and velocity loop replaced the open-loop stepper arrangement. The motion controller executed the position loop and applied acceleration, velocity, and following-error limits.

The end switches were retained as safety and homing devices, but their signals were debounced and monitored independently. The camera trigger was generated from a position event in the motion controller rather than from a non-real-time supervisory application. A watchdog monitored communication and controller execution. Before installation on the production machine, the controller and encoder interface were tested using HIL, including simulated encoder loss, excessive following error, and delayed camera acknowledgment.

Outcome

The redesigned system provided a defined relationship between position and image capture. The controller detected following-error conditions instead of silently accumulating position error. HIL testing exposed a trigger-timing defect before commissioning, and physical testing then focused on mechanical vibration, cable routing, and thermal behavior.

Learning Points

The case illustrates that mechatronic performance is an integration problem. A more powerful actuator would not correct uncertain triggering. Limit switches are useful but cannot substitute for continuous feedback. Supervisory software and real-time motion control should be assigned different responsibilities. Finally, fault detection and timing analysis must be included from the beginning rather than added after performance problems appear.

Challenges, Risks, and Professional Responsibilities

Common integration mistakes include selecting sensors only by static accuracy, ignoring bandwidth, filtering without considering phase delay, using software differentiation on noisy signals, neglecting actuator saturation, assuming nominal network timing, and failing to test abnormal conditions. Another frequent error is mixing safety logic with ordinary sequence logic so that a software fault can defeat the protective function.

Environmental factors deserve explicit treatment. Temperature changes alter sensor sensitivity and actuator characteristics. Vibration can loosen connectors or excite structural resonances. Dust, moisture, hydraulic contamination, electromagnetic fields, and radiation may affect operation. Connector, cable, enclosure, grounding, and maintenance choices are part of the control system.

Automated systems also create governance and cybersecurity responsibilities. Networked controllers should use authentication, controlled access, secure configuration, event logging, backup, and change management appropriate to the risk. Safety-related modifications require documented review and validation. Data collected from machines should be retained only as necessary and protected against unauthorized use where it can reveal production information or operator activity.

Engineers must communicate assumptions, uncertainty, operating limits, and residual risks. A system should fail in a predictable and safe state when feasible. Emergency stopping, energy isolation, restart behavior, manual override, and human-machine interface design require consideration of foreseeable misuse as well as normal operation.

Current Developments and Best Practices

Recent systems increasingly combine edge computing, time-sensitive industrial Ethernet, distributed sensing, digital twins, and condition monitoring. Edge processing can reduce communication volume and latency, while cloud systems support long-term analytics and fleet-level comparison. Machine-learning methods can identify abnormal vibration or actuator current, but they should supplement—not replace—validated safety functions and physical understanding. Their training data, false-alarm behavior, drift, and explainability must be assessed.

Effective practice includes the following:

• Specify the complete measurement and actuation chain, not only individual components.
• Match sensor bandwidth and sampling rate to the dynamics being controlled.
• Budget latency and jitter quantitatively.
• Calibrate with traceable references and maintain configuration records.
• Model saturation, dead zones, backlash, delays, and faults when they affect performance.
• Separate control, sequencing, diagnostics, and safety responsibilities.
• Use deterministic timing for coordinated motion.
• Test nominal, boundary, degraded, and recovery conditions.
• Protect field devices and networks against electrical and cyber disturbances.
• Maintain a lifecycle plan for firmware, calibration, spares, obsolescence, and skills.

Chapter Summary

A functioning control system is an integrated chain linking physical variables, sensors, signal conditioning, data acquisition, computation, communication, actuators, and the controlled plant. Sensor selection must consider bandwidth, resolution, noise, drift, loading, calibration, and environmental conditions. Sampling and filtering determine whether measurements can represent dynamic behavior without aliasing or excessive delay. Actuators must be selected with their drives, limits, thermal characteristics, and mechanical loads.

Embedded controllers and PLCs differ in emphasis but both require deterministic execution, diagnostics, and appropriate safety architecture. Industrial networks introduce timing, synchronization, and cybersecurity concerns. Latency and jitter influence closed-loop stability and must be included in design models. HIL testing provides an efficient means of validating controllers and fault responses before full-machine testing, although it does not eliminate the need for physical validation. Successful mechatronic design treats performance, reliability, safety, maintainability, and lifecycle governance as one engineering problem.

Key Terms

Sensor: A device or element that detects a physical variable.

Transducer: A device that converts a physical quantity into another form, commonly an electrical signal.

Signal conditioning: Processing that adapts, filters, isolates, protects, or linearizes a sensor signal.

Data acquisition: The process of sampling and converting physical signals into digital data.

Bandwidth: The frequency range over which a device or system responds adequately.

Resolution: The smallest input change that can be distinguished.

Accuracy: Closeness of a measured value to a reference value.

Repeatability: Agreement among repeated measurements under specified conditions.

Calibration: Comparison with a reference to establish measurement characteristics.

Drift: Gradual change in measurement output unrelated to the desired input change.

Aliasing: Misrepresentation of high-frequency content as lower-frequency content after sampling.

Quantization: Conversion of a continuous amplitude into discrete numerical levels.

Actuator: A device that converts a command into force, torque, motion, pressure, or flow.

Latency: Time between an event, its measurement or transmission, and the resulting response.

Jitter: Variation in the timing of nominally periodic events.

PLC: An industrial controller designed for robust automation and sequential control.

Real-time system: A system required to produce outputs within specified timing deadlines.

Hardware-in-the-loop: Testing in which real control hardware interacts with a real-time simulated plant.

Saturation: A condition in which an actuator or signal cannot exceed its physical or configured limit.

Mechatronics: Integrated design of mechanical, electrical, control, computing, and sensing elements.

Review Questions

Short Answer Questions

1. Distinguish sensor accuracy from sensor resolution.

2. Why is an anti-alias filter placed before an analog-to-digital converter?

3. Define latency and jitter in a real-time control loop.

4. What is actuator saturation, and why can it cause integral windup?

5. State two reasons for using hardware-in-the-loop testing.

Long Answer Questions

1. Explain the complete measurement chain from a physical variable to a controller input, including calibration and signal conditioning.

2. Compare electric, hydraulic, and pneumatic actuators for industrial motion applications.

3. Discuss how sampling rate, filtering, quantization, and communication delay influence closed-loop performance.

4. Explain the respective roles of embedded controllers, PLCs, and industrial networks in an automated machine.

5. Describe a systematic workflow for designing and validating an integrated mechatronic system.

Analytical / Application Questions

1. A position sensor is sampled at 1 kHz, while machine vibration contains a strong component at 700 Hz. Explain the measurement risk and propose a revised acquisition strategy.

2. A servo axis becomes unstable after a networked feedback link is introduced. Identify the timing effects that should be measured and explain possible corrective actions.

3. Design a verification plan for a hydraulic actuator controller that includes calibration, saturation, sensor failure, communication loss, HIL testing, and final physical validation.

22. Multiple Choice Questions with Answers

Question 1

A sensor has very fine resolution but a significant constant offset. Which statement best describes it?

A. It has high resolution but may have poor accuracy

B. It has poor resolution but high accuracy

C. It necessarily has high repeatability and accuracy

D. Its bandwidth must be low

Correct Answer: A. It has high resolution but may have poor accuracy

Question 2

What is the principal purpose of an anti-alias filter in a data-acquisition system?

A. To increase actuator torque

B. To remove frequency components that could be misrepresented during sampling

C. To compensate encoder offset after conversion

D. To eliminate quantization completely

Correct Answer: B. To remove frequency components that could be misrepresented during sampling

Question 3

A control loop samples at 1 kHz. Which input frequency is most clearly above its Nyquist frequency?

A. 100 Hz

B. 250 Hz

C. 450 Hz

D. 700 Hz

Correct Answer: D. 700 Hz

Question 4

Which condition most directly produces integral windup?

A. The sensor has excessive resolution

B. The actuator is saturated while the integral term continues accumulating error

C. The network uses differential signaling

D. The controller uses a watchdog timer

Correct Answer: B. The actuator is saturated while the integral term continues accumulating error

Question 5

Why can adding a low-pass filter to a feedback signal reduce stability margin?

A. It always increases sensor range

B. It introduces phase delay in the feedback path

C. It removes actuator saturation

D. It increases sampling jitter by definition

Correct Answer: B. It introduces phase delay in the feedback path

Question 6

Which feature is most associated with a real-time controller?

A. Maximum average computation speed without deadlines

B. Guaranteed execution within specified timing limits

C. Operation without sensors

D. Dependence on cloud connectivity

Correct Answer: B. Guaranteed execution within specified timing limits

Question 7

In a positioning system, why is a linear encoder often preferable to relying only on motor step count?

A. It directly measures axis position and can reveal lost motion

B. It removes the need for an actuator

C. It guarantees zero measurement noise

D. It prevents all mechanical vibration

Correct Answer: A. It directly measures axis position and can reveal lost motion

Question 8

What is the principal role of hardware-in-the-loop simulation?

A. To replace all physical testing permanently

B. To connect real control hardware to a real-time model of the plant

C. To calibrate every sensor without a reference

D. To increase the mechanical stiffness of a machine

Correct Answer: B. To connect real control hardware to a real-time model of the plant

Question 9

Which design practice best addresses coordinated multi-axis motion over an industrial network?

A. Ignore timing because average bandwidth is sufficient

B. Use synchronized timing and account for communication latency and jitter

C. Increase sensor noise intentionally

D. Use only open-loop actuators

Correct Answer: B. Use synchronized timing and account for communication latency and jitter

Question 10

Which statement about safety functions is most appropriate?

A. They should depend entirely on ordinary sequence software

B. They should be separated or independently monitored where risk requires it

C. They are unnecessary if a system has high-resolution sensors

D. They can be omitted during commissioning tests

Correct Answer: B. They should be separated or independently monitored where risk requires it