1|Page
Certified Professional in Autonomous Systems
(CPAS) Exam | Latest Verified Questions and
Detailed Answers
OVERVIEW DESCRIPTION:
This Comprehensive set of multiple choice questions is designed for the Certified
Professional in Autonomous Systems (CPAS) exam, a rigorous, multidisciplinary credential
that validates a professional’s ability to design, integrate, verify, and validate safe
autonomous and unmanned systems. The questions span the exam’s six fundamental
domains: Perception & Sensor Fusion, Localization & Mapping, Decision Making & Path
Planning, Control Systems & Actuation, Verification, Validation & Safety (V&V), and
Cybersecurity & Ethics, assessing both foundational knowledge and practical reasoning
across the full spectrum of autonomous system engineering.
QUESTION 1
Which sensor is most susceptible to multipath reflections causing ghost detections in
autonomous driving?
A) LiDAR
B) Radar
C) Monocular camera
D) Ultrasonic sonar
CORRECT ANSWER: B
EXPERT RATIONALE: Radar waves readily reflect off surfaces, creating false targets at
incorrect ranges. LiDAR and cameras are less prone to coherent multipath artifacts.
,2|Page
QUESTION 2
In a Kalman filter, the Kalman gain determines:
A) The initial state estimate
B) The weight given to measurement versus prediction
C) The process noise covariance
D) The control input matrix
CORRECT ANSWER: B
EXPERT RATIONALE: The Kalman gain optimally blends the predicted state with the
incoming measurement based on their respective uncertainties.
QUESTION 3
What is the primary advantage of early sensor fusion over late fusion?
A) Lower computational cost
B) Retention of raw data correlations before feature extraction
C) Simpler synchronization
D) Less bandwidth
CORRECT ANSWER: B
EXPERT RATIONALE: Early fusion combines raw sensor streams, preserving mutual
information that may be discarded after independent feature extraction in late fusion.
,3|Page
QUESTION 4
LiDAR point cloud intensity values primarily represent:
A) Object color
B) Surface reflectivity at the laser wavelength
C) Distance measurement accuracy
D) Temperature
CORRECT ANSWER: B
EXPERT RATIONALE: Intensity corresponds to the strength of the returned laser pulse,
which depends on the target material's reflectivity.
QUESTION 5
Which data association algorithm maintains multiple hypotheses over time using a
probabilistic framework?
A) Global Nearest Neighbor (GNN)
B) Joint Probabilistic Data Association (JPDA)
C) Multiple Hypothesis Tracking (MHT)
D) Nearest Neighbor (NN)
CORRECT ANSWER: C
, 4|Page
EXPERT RATIONALE: MHT propagates branching association hypotheses and defers
decisions until future measurements resolve ambiguities.
QUESTION 6
The Extended Kalman Filter (EKF) handles process nonlinearities by:
A) Using Monte Carlo sampling
B) Linearizing the system model around the current estimate
C) Applying a particle filter
D) Using neural networks
CORRECT ANSWER: B
EXPERT RATIONALE: EKF computes Jacobians to linearize the state transition and
measurement functions at each time step.
QUESTION 7
In camera calibration, which parameters describe lens distortion?
A) Extrinsic parameters
B) Intrinsic parameters including radial and tangential distortion coefficients
C) Rotation matrix
D) Translation vector
CORRECT ANSWER: B
Certified Professional in Autonomous Systems
(CPAS) Exam | Latest Verified Questions and
Detailed Answers
OVERVIEW DESCRIPTION:
This Comprehensive set of multiple choice questions is designed for the Certified
Professional in Autonomous Systems (CPAS) exam, a rigorous, multidisciplinary credential
that validates a professional’s ability to design, integrate, verify, and validate safe
autonomous and unmanned systems. The questions span the exam’s six fundamental
domains: Perception & Sensor Fusion, Localization & Mapping, Decision Making & Path
Planning, Control Systems & Actuation, Verification, Validation & Safety (V&V), and
Cybersecurity & Ethics, assessing both foundational knowledge and practical reasoning
across the full spectrum of autonomous system engineering.
QUESTION 1
Which sensor is most susceptible to multipath reflections causing ghost detections in
autonomous driving?
A) LiDAR
B) Radar
C) Monocular camera
D) Ultrasonic sonar
CORRECT ANSWER: B
EXPERT RATIONALE: Radar waves readily reflect off surfaces, creating false targets at
incorrect ranges. LiDAR and cameras are less prone to coherent multipath artifacts.
,2|Page
QUESTION 2
In a Kalman filter, the Kalman gain determines:
A) The initial state estimate
B) The weight given to measurement versus prediction
C) The process noise covariance
D) The control input matrix
CORRECT ANSWER: B
EXPERT RATIONALE: The Kalman gain optimally blends the predicted state with the
incoming measurement based on their respective uncertainties.
QUESTION 3
What is the primary advantage of early sensor fusion over late fusion?
A) Lower computational cost
B) Retention of raw data correlations before feature extraction
C) Simpler synchronization
D) Less bandwidth
CORRECT ANSWER: B
EXPERT RATIONALE: Early fusion combines raw sensor streams, preserving mutual
information that may be discarded after independent feature extraction in late fusion.
,3|Page
QUESTION 4
LiDAR point cloud intensity values primarily represent:
A) Object color
B) Surface reflectivity at the laser wavelength
C) Distance measurement accuracy
D) Temperature
CORRECT ANSWER: B
EXPERT RATIONALE: Intensity corresponds to the strength of the returned laser pulse,
which depends on the target material's reflectivity.
QUESTION 5
Which data association algorithm maintains multiple hypotheses over time using a
probabilistic framework?
A) Global Nearest Neighbor (GNN)
B) Joint Probabilistic Data Association (JPDA)
C) Multiple Hypothesis Tracking (MHT)
D) Nearest Neighbor (NN)
CORRECT ANSWER: C
, 4|Page
EXPERT RATIONALE: MHT propagates branching association hypotheses and defers
decisions until future measurements resolve ambiguities.
QUESTION 6
The Extended Kalman Filter (EKF) handles process nonlinearities by:
A) Using Monte Carlo sampling
B) Linearizing the system model around the current estimate
C) Applying a particle filter
D) Using neural networks
CORRECT ANSWER: B
EXPERT RATIONALE: EKF computes Jacobians to linearize the state transition and
measurement functions at each time step.
QUESTION 7
In camera calibration, which parameters describe lens distortion?
A) Extrinsic parameters
B) Intrinsic parameters including radial and tangential distortion coefficients
C) Rotation matrix
D) Translation vector
CORRECT ANSWER: B
