Connected, Automated and Software-Defined Vehicles

Introduction and Context

Automobile engineering is being transformed from the design of predominantly mechanical machines into the design of cyber-physical systems. A contemporary vehicle combines mechanical structures, electric powertrains, embedded controllers, communication networks, cloud services, artificial intelligence (AI), and continuously updated software. This transformation is described by three related ideas:

• Connected vehicles exchange information with other vehicles, infrastructure, mobile networks, cloud platforms, and service providers.
• Automated vehicles use sensing, computation, and control to assist or replace portions of the driving task.
• Software-defined vehicles (SDVs) implement an increasing proportion of vehicle functions through software that can be configured, monitored, and upgraded throughout the vehicle life cycle.

These ideas overlap but are not identical. A vehicle may be connected without being automated, or automated in a limited function without being connected to a cloud. The engineering challenge is to integrate these capabilities while preserving safety, security, reliability, affordability, legal compliance, and public trust.

Learning Objectives

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

1. Explain ADAS and the levels of driving automation.
2. Describe the operating principles and limitations of cameras, radar, lidar, ultrasonic sensors, GNSS, and inertial sensors.
3. Explain sensor fusion, localization, planning, and vehicle control.
4. Distinguish V2X communication, telematics, and cloud-connected services.
5. Explain software-defined vehicle architecture and over-the-air updating.
6. Analyze applications of AI and the safety implications of machine learning.
7. Apply the principles of functional safety and automotive cybersecurity.
8. Evaluate ethical, regulatory, and societal issues associated with automated mobility.

1. Advanced Driver-Assistance Systems and Driving Automation

1.1 ADAS functions

Advanced driver-assistance systems (ADAS) support the driver by sensing the environment, interpreting traffic situations, and influencing steering, braking, or propulsion. Typical functions include:

• forward-collision warning (FCW);
• automatic emergency braking (AEB);
• adaptive cruise control (ACC);
• lane-departure warning and lane-keeping assistance;
• blind-spot monitoring;
• traffic-sign recognition;
• driver monitoring;
• parking assistance and automated parking.

ADAS is not a single technology. It is a system comprising sensors, electronic control units (ECUs), software, actuators, human–machine interfaces, and diagnostic functions. A system can be technically capable of controlling a vehicle but still require continuous driver supervision. Therefore, the name and operating design must communicate the driver’s responsibility clearly.

1.2 Levels of driving automation

The widely used SAE classification describes automation from Level 0 to Level 5:

The classification concerns the dynamic driving task, not whether a vehicle has connectivity, remote monitoring, or advanced infotainment. Operational design domain (ODD) is critical. An ODD specifies the conditions under which a function is designed to operate, including road type, geography, speed, weather, traffic, lighting, and time of day.

A Level 2 system, for example, may control steering and speed on a motorway, but the driver must remain attentive. A Level 4 shuttle may operate without driver supervision, but only within a mapped campus and at low speed. Confusing capability with responsibility is a major safety and legal risk.

2. Perception Sensors

Perception converts physical phenomena into estimates of objects, road geometry, free space, and motion. No individual sensor is perfect; each has a characteristic range, resolution, field of view, update rate, and failure mode.

2.1 Cameras

Cameras measure reflected visible or infrared light. Monocular cameras provide rich appearance information, including lane markings, signs, traffic lights, and object categories. Stereo cameras estimate depth from the difference between corresponding image points. The approximate depth relationship for a rectified stereo pair is:

where Z is object depth in metres, f is focal length in pixels, b is the baseline between cameras in metres, and d is disparity in pixels. The relationship assumes calibrated cameras and reliable correspondence. Depth uncertainty increases as disparity becomes small, so distant objects are difficult to measure accurately.

Cameras are comparatively inexpensive and information-rich, but performance is degraded by darkness, glare, fog, rain, dirt, shadows, and ambiguous visual patterns. AI-based object detection can classify vehicles, pedestrians, cyclists, and road features, but it may fail on unusual objects or conditions poorly represented in training data.

2.2 Radar

Automotive radar transmits radio-frequency waves and measures reflected energy. The Doppler effect provides relative radial velocity. For a simple monostatic radar, the approximate Doppler relationship is:

where f_D is Doppler frequency shift in hertz, v_r is radial relative velocity in metres per second, and λ is wavelength in metres. The factor 2 occurs because the wave travels to the target and back. The sign convention depends on whether closing or separating motion is defined as positive.

Radar performs well in darkness, rain, and dust and directly measures relative speed. Its angular resolution and object classification are generally weaker than those of high-resolution cameras or lidar. Multipath reflections and stationary roadside objects can also complicate interpretation.

2.3 Lidar

Lidar estimates range by measuring the travel time of laser pulses. For a pulse traveling to a target and returning, the range is:

where R is range in metres, c is the speed of light in metres per second, and Δt is round-trip time in seconds. The equation assumes the pulse travels approximately in a straight path and that the return has been correctly associated with the transmitted pulse.

Lidar creates a three-dimensional point cloud with accurate geometry. It can support mapping, object detection, and free-space estimation. Limitations include cost, moving mechanical components in some designs, reduced performance in heavy precipitation, contamination of optical surfaces, and the challenge of processing large data volumes.

2.4 Ultrasonic sensors

Ultrasonic sensors emit sound above the audible range and measure echo time. They are useful at short range for parking and low-speed obstacle detection. They are inexpensive but have limited range, relatively poor angular resolution, and sensitivity to surface orientation, temperature, and acoustic interference.

2.5 Localization sensors

Global Navigation Satellite Systems (GNSS) provide global position, but urban canyons, tunnels, multipath, atmospheric effects, and signal blockage reduce accuracy. An inertial measurement unit (IMU) measures acceleration and angular velocity. Integrating these measurements gives motion estimates, but small biases accumulate over time, producing drift. Wheel-speed sensors and steering-angle sensors provide useful vehicle-motion information, although wheel slip and tire-radius variation introduce errors.

3. Sensor Fusion and Environment Modeling

Sensor fusion combines measurements to produce a more reliable estimate than one sensor can provide. Fusion may occur at different levels:

• Raw-data fusion: combines sensor signals before feature extraction; it can be powerful but demands precise calibration and synchronization.
• Feature-level fusion: combines detected edges, point-cloud features, image features, or radar tracks.
• Object-level fusion: combines independently generated object lists; it is modular and easier to validate but may discard information.

A common probabilistic estimator is the Kalman filter. For a linear discrete-time model:

Here x_k is the state vector at time k, u_k is a known input, z_k is a measurement, F_k is the state-transition matrix, B_k is the input matrix, and H_k maps the state into measurement space. w_k and v_k represent process and measurement noise, commonly modeled as zero-mean random variables with covariance matrices Q_k and R_k. The standard Kalman filter assumes linear models and approximately Gaussian noise. Extended or unscented variants are used for nonlinear vehicle motion and sensor models.

The prediction equations are:

The measurement update is:

where P is state-estimate covariance, K is Kalman gain, and the superscript T denotes matrix transpose. In practical vehicles, fusion must also handle outliers, delayed data, sensor faults, uncertain object identity, and non-Gaussian errors. A robust design includes innovation monitoring, plausibility checks, redundancy, and graceful degradation.

Figure: Simplified closed-loop architecture of an automated vehicle.

4. Localization, Planning, and Control

4.1 Localization

Localization estimates the vehicle pose—position, orientation, and sometimes velocity—relative to a map or coordinate frame. Methods include GNSS/IMU integration, lidar map matching, visual odometry, and simultaneous localization and mapping (SLAM). High-definition maps may contain lane boundaries, traffic signs, road topology, and landmarks. However, maps become stale because roads, signs, and lane markings change. A capable system must compare map information with live perception rather than treating the map as unquestionable truth.

4.2 Prediction and planning

Prediction estimates how road users may move. A pedestrian may stop, cross, or change direction; a vehicle may brake or change lanes. Planning then selects a safe and legally permissible trajectory. Planning is usually hierarchical:

1. route planning selects a road-level route;
2. behavioral planning chooses actions such as following, yielding, or changing lanes;
3. local trajectory planning generates a time-dependent path;
4. control tracks the selected trajectory.

Safety constraints include collision avoidance, road boundaries, vehicle stability, comfort, actuator limits, and uncertainty. A mathematically optimal trajectory is not necessarily acceptable if it is difficult for surrounding human drivers to interpret.

4.3 Vehicle control

A controller converts the planned trajectory into steering, braking, and propulsion commands. Longitudinal control regulates speed and distance; lateral control regulates path position and heading. A simple proportional-integral-derivative (PID) controller is:

where e(t) is the tracking error, u(t) is the control command, and K_p, K_i, and K_d are proportional, integral, and derivative gains. The integral term accumulates error and can remove steady-state offset, while the derivative term responds to error rate. Practical implementation requires sampling, actuator saturation, anti-windup, filtering, and careful treatment of measurement noise.

More advanced vehicles use model predictive control (MPC), which repeatedly predicts vehicle behavior over a finite horizon and optimizes control commands subject to constraints. MPC can account for tire force limits and actuator constraints, but it requires a sufficiently accurate model and substantial computation.

5. Connectivity and Telematics

Vehicle connectivity links the vehicle to external systems. Important communication categories include:

• V2V: vehicle-to-vehicle communication;
• V2I: vehicle-to-infrastructure communication;
• V2P: vehicle-to-pedestrian or vulnerable-road-user communication;
• V2N: vehicle-to-network communication;
• V2X: collective term for these interactions.

Potential applications include hazard warnings, cooperative awareness, signal-phase information, electronic tolling, fleet dispatch, emergency response, and platooning. Connectivity can extend perception beyond line of sight, but received data must be authenticated, time-stamped, plausibility-checked, and treated as uncertain. Communication does not eliminate the need for onboard sensing because network latency, packet loss, malicious messages, and coverage gaps are unavoidable.

Telematics combines telecommunications and informatics to collect and transmit vehicle data. A telematics control unit may support emergency calling, stolen-vehicle recovery, fleet monitoring, remote diagnostics, usage-based insurance, and maintenance scheduling. Data may include location, speed, fault codes, energy consumption, charging events, and driver interactions.

6. Software-Defined Vehicle Architecture

Traditional vehicles contain many distributed ECUs, each designed for a relatively narrow function. SDVs increasingly use powerful domain or zonal controllers connected through high-bandwidth networks. Zonal architecture places input/output interfaces near physical vehicle zones, reducing wiring length while central computers execute shared software services.

Important SDV characteristics include:

• separation of hardware abstraction, operating systems, middleware, and applications;
• service-oriented communication between software components;
• centralized data logging and diagnostics;
• scalable computing and storage;
• software reuse across vehicle models;
• continuous deployment of validated functions;
• strong configuration and version management.

The SDV changes the engineering life cycle. Verification is no longer completed only at factory release; software behavior must be monitored and controlled throughout the vehicle’s service life. Compatibility among software versions, sensors, ECUs, cloud services, and mobile applications becomes a systems-engineering responsibility.

6.1 Over-the-air updates

An over-the-air (OTA) update transmits software or configuration changes through a wireless network. A robust process includes:

1. change identification and risk classification;
2. software development and testing;
3. cryptographic signing;
4. secure download and integrity verification;
5. staged deployment;
6. installation in a safe vehicle state;
7. rollback or recovery capability;
8. post-update monitoring and records.

A vehicle should not install a safety-critical update while driving or when interruption could disable essential functions. Dual-bank or A/B partitions allow a new image to be installed separately from the currently operating image, supporting rollback if validation fails. OTA capability increases convenience and reduces service visits, but it also creates a large cybersecurity and configuration-management responsibility.

7. Artificial Intelligence in Vehicles

AI is applied to object detection, lane and road understanding, driver monitoring, speech interaction, predictive maintenance, energy management, traffic prediction, and automated driving policy. Deep neural networks are particularly effective for image and point-cloud interpretation because they learn complex features from large datasets.

However, a high test accuracy does not guarantee safe deployment. Safety evaluation must examine:

• rare and hazardous scenarios;
• weather, lighting, and geographic variation;
• distribution shift between training and operation;
• false positives and false negatives;
• uncertainty estimation;
• dataset bias;
• adversarial or manipulated inputs;
• computational timing and resource limits;
• explainability appropriate to the safety decision.

AI components should be integrated into a broader safety architecture. A neural detector may propose an object, while independent geometric checks, temporal consistency, radar confirmation, and conservative emergency logic constrain the consequences of an incorrect prediction. The system should enter a minimal-risk condition when confidence is insufficient or critical data are unavailable.

8. Functional Safety and Cybersecurity

8.1 Functional safety

Functional safety addresses unreasonable risk arising from malfunctioning behavior of electrical and electronic systems. Automotive development commonly uses hazard analysis and risk assessment, safety goals, technical safety concepts, safety mechanisms, verification, validation, and evidence across the development life cycle. ISO 26262 defines an automotive safety integrity level (ASIL) framework from QM (quality management) through ASIL A to ASIL D. ASIL is assigned based on severity, exposure, and controllability; it is not simply a measure of how technologically advanced a function is.

Typical safety mechanisms include redundancy, watchdog timers, range checks, end-to-end communication protection, independent monitoring, safe-state transitions, diagnostic trouble codes, and fail-operational or fail-safe design. A steering system for automated operation may need continued operation after a single fault, whereas a nonessential infotainment feature may safely shut down.

Safety of the intended functionality (often associated with SOTIF concepts) addresses hazards that occur without a conventional component failure—for example, an object detector that behaves inadequately in an unusual but foreseeable situation. This distinction is crucial for automated driving, where limitations of perception and interpretation can be hazardous even when hardware works correctly.

8.2 Cybersecurity

Connected vehicles present attack surfaces in wireless interfaces, diagnostic ports, mobile applications, cloud services, charging systems, third-party software, and supplier components. Threats include unauthorized unlocking, data theft, denial of service, manipulation of sensor or positioning data, ransomware, and compromise of safety-related networks.

A cybersecurity engineering process includes:

• asset identification;
• threat analysis and risk assessment;
• secure architecture and segmentation;
• authentication and authorization;
• encryption where confidentiality is required;
• secure boot and hardware-backed keys;
• intrusion detection;
• vulnerability management;
• incident response and forensic logging;
• secure decommissioning.

Safety and cybersecurity interact but are distinct. A safety failure may result from random hardware failure; a cybersecurity failure is often intentional and adaptive. Security controls must not introduce unacceptable latency or prevent emergency operation. Conversely, a safety monitor can detect effects of a cyberattack without identifying its cause.

9. Ethical, Legal, and Governance Issues

Automated and connected vehicles raise issues extending beyond technical performance.

Privacy: Continuous collection of location, cabin, biometric, and behavioral data can reveal sensitive personal information. Data minimization, informed consent, purpose limitation, retention controls, and access governance are essential.

Accountability: When a function makes a harmful decision, responsibility may involve the driver, manufacturer, software supplier, fleet operator, maintenance provider, or infrastructure authority. Clear operating instructions and event records are necessary but do not by themselves resolve legal responsibility.

Fairness: Perception systems may perform differently for pedestrians with different clothing, body types, mobility aids, skin tones, or lighting conditions. Testing must include representative populations and environments.

Transparency: Users should understand what a system can and cannot do, what data it collects, and when control is transferred. Marketing terms that imply full autonomy can encourage misuse.

Employment and access: Automation may alter professional driving jobs while improving mobility for people unable to drive. Policy should consider retraining, accessibility, affordability, and equitable deployment.

Moral decision-making: Engineers should not reduce complex social values to simplistic “trolley problem” algorithms. The primary design goal is prevention, lawful behavior, risk reduction, and human oversight within a defensible governance framework.

10. Illustrative Case: Connected Automated Campus Shuttle

Background

A university deploys a battery-electric shuttle on a 4-km route containing marked crossings, mixed pedestrian traffic, two signalized intersections, and a short section shared with bicycles. The shuttle is intended to operate at 25 km/h with a safety operator during the initial phase.

Problem

The operator requires reliable detection of pedestrians and cyclists, accurate stopping at designated points, remote health monitoring, and safe operation during intermittent wireless coverage. The system must also tolerate a blocked camera and prevent unauthorized remote commands.

Analysis

A camera supplies classification and lane information, radar supplies relative motion and range, and lidar supports geometric detection of obstacles. GNSS, IMU, wheel speeds, and map matching provide localization. The ODD excludes heavy snow, flooded roads, and roads outside the mapped route. The safety analysis identifies hazards including failure to detect a pedestrian, unintended acceleration, loss of localization, communication loss, and cyberattack on the remote-control interface.

Approach and solution

The shuttle uses object-level fusion with time synchronization and confidence checks. An independent low-speed emergency braking channel monitors near-field obstacles. If localization uncertainty exceeds a defined limit, the shuttle reduces speed and stops at a safe location. Loss of cellular communication does not directly disable local driving, but it prevents dispatch of new missions and triggers operator notification. Remote commands require mutual authentication, authorization by role, command sequencing, and event logging. Software updates are tested in simulation, on a closed track, and in a limited pilot before staged deployment.

Outcome

During the pilot, the shuttle completes normal trips but stops conservatively when a cyclist approaches from an occluded angle. Investigation shows that the stop is appropriate but creates schedule delays. Engineers improve prediction and route-side sensing without weakening the emergency stop threshold. A camera contamination alert is also added after controlled testing demonstrates degraded detection.

Learning points

The case illustrates that deployment is an operational system, not merely an algorithm. A defined ODD, redundant sensing, conservative fallback, human oversight, cybersecurity controls, monitoring, and iterative validation are all required. Performance should be evaluated not only by average trip completion but also by hazardous-event detection, false-stop rate, recovery time, update integrity, and user understanding.

11. Challenges, Limitations, and Best Practices

Important limitations include sensor occlusion, adverse weather, incomplete maps, ambiguous human behavior, communication latency, software complexity, energy consumption, hardware obsolescence, insufficient training data, and difficulty proving safety for rare events. Fully automated operation cannot be justified by demonstrations of normal conditions alone.

Best practice is to:

1. define the ODD precisely;
2. design for graceful degradation and minimal-risk conditions;
3. combine complementary sensors rather than relying on a single modality;
4. maintain calibration, time synchronization, and data quality;
5. use independent safety monitors;
6. separate safety-critical networks and enforce least privilege;
7. test simulation, hardware-in-the-loop, closed-track, and public-road scenarios;
8. include rare, adversarial, and degraded conditions;
9. maintain traceable software, data, and configuration records;
10. monitor field performance and support secure incident response;
11. communicate driver responsibilities unambiguously;
12. consider privacy, fairness, accessibility, and environmental impact from the design stage.

Contemporary Developments

Current engineering trends include high-performance centralized computers, zonal electrical architectures, solid-state lidar, imaging radar, cooperative perception, vehicle-to-grid communication, digital twins, foundation models for multimodal perception, and increasingly automated validation. These developments may improve capability, but they also increase software complexity, data dependence, energy demand, and supply-chain exposure. The competent automobile engineer must therefore combine vehicle dynamics and controls knowledge with embedded systems, communications, data engineering, safety assurance, and responsible innovation.

Chapter Summary

Connected, automated, and software-defined vehicles are integrated cyber-physical systems. ADAS functions range from warnings to combined steering and speed control, while higher automation depends on a clearly defined operational design domain. Cameras, radar, lidar, ultrasonic sensors, GNSS, IMUs, and wheel sensors provide complementary information and must be fused with attention to calibration, uncertainty, timing, and failure modes. Localization, prediction, planning, and feedback control form a closed loop between perception and vehicle actuation.

Connectivity enables V2X services and telematics but does not replace onboard perception. SDV architectures support centralized computation, service-oriented software, diagnostics, and secure OTA updates. AI improves perception and prediction but must be evaluated under distribution shift, rare events, uncertainty, and bias. Functional safety addresses malfunctioning behavior, while cybersecurity protects systems against intentional attacks; both require structured life-cycle engineering. Ethical governance must address privacy, accountability, fairness, transparency, accessibility, and appropriate human responsibility.

Key Terms

Operational design domain: The defined conditions in which an automated function is intended to operate.

ADAS: Electronic systems that assist the driver with sensing, warning, or vehicle control.

Sensor fusion: Combination of measurements from multiple sensors to estimate states or objects.

Localization: Estimation of a vehicle’s position and orientation relative to a reference frame or map.

Perception: Interpretation of sensor data to identify objects, road features, and free space.

Lidar: Range-sensing technology using reflected laser pulses.

Radar: Radio-frequency sensing technology that can estimate range and relative velocity.

Telematics: Use of telecommunications and computing to collect, transmit, and analyze vehicle data.

V2X: Communication between a vehicle and vehicles, infrastructure, pedestrians, or networks.

Software-defined vehicle: Vehicle whose functions and behavior are substantially configured and delivered through software.

OTA update: Wireless delivery and installation of software or configuration changes.

Kalman filter: Recursive estimator combining a dynamic model and noisy measurements.

SLAM: Simultaneous localization and mapping using sensor observations.

MPC: Model predictive control that optimizes future control actions subject to constraints.

Functional safety: Reduction of unreasonable risk caused by malfunctioning electrical or electronic systems.

ASIL: Automotive Safety Integrity Level used in risk-based functional-safety development.

SOTIF: Safety consideration for hazards arising from limitations of intended functionality.

Secure boot: Startup process that permits only authenticated software to execute.

Minimal-risk condition: State entered by an automated system to reduce danger after a fault or limitation.

Sensor degradation: Reduction in sensing performance caused by faults, contamination, weather, or other conditions.

Review Questions

Short Answer Questions

1. What is the difference between a connected vehicle and an automated vehicle?
2. Why are radar and cameras complementary sensors?
3. Define an operational design domain.
4. What is the purpose of an OTA update rollback mechanism?
5. Distinguish functional safety from cybersecurity.

Long Answer Questions

1. Explain the architecture of an automated vehicle from sensing through actuation, including the roles of perception, fusion, planning, and control.
2. Compare cameras, radar, lidar, ultrasonic sensors, GNSS, and IMUs in terms of information provided and principal limitations.
3. Describe a secure and safety-conscious OTA update life cycle.
4. Discuss why AI accuracy alone is insufficient evidence of safe automated-driving performance.
5. Explain how privacy, accountability, fairness, and transparency affect connected-vehicle engineering.

Analytical / Application Questions

1. Design a sensor and fallback strategy for a low-speed automated shuttle operating near pedestrians in rain and intermittent cellular coverage. Justify the redundancy and minimal-risk behavior.
2. A vehicle’s GNSS position becomes unreliable in an urban canyon while its camera is partially occluded. Propose a localization and safety response using the remaining sensors and explain the associated uncertainties.
3. A fleet operator wants to introduce an OTA update that changes adaptive-cruise behavior. Develop a validation, cybersecurity, deployment, monitoring, and rollback plan, identifying the principal safety and governance risks.

Multiple Choice Questions with Answers

Question 1

Which statement best describes a Level 2 automated-driving system?

A. The system performs all driving everywhere without human supervision.

B. The system provides combined steering and speed assistance, while the driver supervises continuously.

C. The system only provides a visual warning and never controls the vehicle.

D. The system operates without a defined operational design domain.

Correct Answer: B. The system provides combined steering and speed assistance, while the driver supervises continuously.

Question 2

In the stereo-depth relationship Z = fb/d, what generally happens to estimated depth as disparity d becomes smaller?

A. The estimated depth becomes larger and less certain.

B. The estimated depth becomes smaller and more certain.

C. Depth becomes independent of camera baseline.

D. The object’s radial velocity becomes zero.

Correct Answer: A. The estimated depth becomes larger and less certain.

Question 3

Which sensor is particularly useful for directly measuring relative radial velocity?

A. Ultrasonic sensor

B. Radar

C. Monocular camera

D. GNSS antenna alone

Correct Answer: B. Radar

Question 4

What is the principal purpose of sensor fusion?

A. To eliminate the need for calibration

B. To replace all onboard sensing with cloud data

C. To combine complementary measurements into a more reliable estimate

D. To guarantee that every object is classified correctly

Correct Answer: C. To combine complementary measurements into a more reliable estimate

Question 5

In the lidar range equation R = cΔt/2, why is the travel time divided by 2?

A. The laser frequency is halved during reflection.

B. The measured time includes travel to the target and back.

C. Two sensors are always required.

D. The target absorbs half the pulse energy.

Correct Answer: B. The measured time includes travel to the target and back.

Question 6

Which feature is most characteristic of a software-defined vehicle?

A. All functions are permanently fixed at factory production.

B. Vehicle behavior can be delivered and improved through managed software updates.

C. The vehicle contains no electronic control units.

D. Mechanical maintenance is completely eliminated.

Correct Answer: B. Vehicle behavior can be delivered and improved through managed software updates.

Question 7

Which practice most directly protects an OTA update from unauthorized software installation?

A. Increasing tire pressure

B. Cryptographic signing and secure verification

C. Disabling all diagnostic records

D. Reducing camera resolution

Correct Answer: B. Cryptographic signing and secure verification

Question 8

What does functional safety primarily address?

A. Risks caused by malfunctioning electrical or electronic systems

B. The commercial price of cloud subscriptions

C. The visual styling of a vehicle

D. Only the confidentiality of customer data

Correct Answer: A. Risks caused by malfunctioning electrical or electronic systems

Question 9

Why can a connected vehicle still require onboard sensors?

A. Wireless communication is always faster than local sensing.

B. Network data can suffer latency, loss, spoofing, or coverage gaps.

C. Onboard sensors cannot measure vehicle speed.

D. Connectivity prevents all perception failures.

Correct Answer: B. Network data can suffer latency, loss, spoofing, or coverage gaps.

Question 10

Which is an ethical concern particularly associated with telematics?

A. Brake-pad friction

B. Collection and use of location and behavioral data

C. Wheel alignment tolerance only

D. Engine displacement

Correct Answer: B. Collection and use of location and behavioral data