Download Autonomous Navigation Systems Exam and more Exams Technology in PDF only on Docsity!
Exam
Question 1. Which of the following best defines an autonomous navigation system? A) A system that requires constant human intervention to operate B) A system capable of perceiving, planning, and acting independently in its environment C) A system that only uses GPS for navigation D) A system limited to pre-programmed routes Answer: B Explanation: An autonomous navigation system is designed to perceive its environment, plan routes or actions, and execute those actions independently, requiring minimal or no human intervention. Question 2. Who is considered a pioneer in the development of self- driving vehicles? A) Nikola Tesla B) Sebastian Thrun C) Alan Turing D) Steve Jobs Answer: B
Exam
Explanation: Sebastian Thrun led the Stanford team that won the 2005 DARPA Grand Challenge, a pivotal event in autonomous vehicle development. Question 3. Which application domain commonly utilizes autonomous navigation systems? A) Online banking B) Self-driving cars C) E-commerce D) Social media Answer: B Explanation: Self-driving cars are a prominent example of autonomous navigation systems in real-world applications. Question 4. What is the main difference between Level 2 and Level 5 autonomy in vehicles? A) Level 2 requires full human oversight, Level 5 is fully autonomous B) Level 2 has no automation, Level 5 requires partial automation C) Level 2 is only for aerial vehicles, Level 5 is for ground vehicles
Exam
B) Body frame C) ECEF (Earth-Centered, Earth-Fixed) D) UTM (Universal Transverse Mercator) Answer: C Explanation: The ECEF frame’s origin is at the Earth's center and its axes are fixed relative to the Earth's rotation, commonly used in GPS calculations. Question 7. Euler angles are used to represent: A) Linear velocities B) Rotational orientations C) Map scales D) Sensor range Answer: B Explanation: Euler angles describe the orientation of a rigid body in 3D space using three rotational angles. Question 8. Which mathematical model is typically used for ground vehicle kinematics?
Exam
A) Bicycle model B) Double integrator model C) Helicopter dynamic model D) Pendulum model Answer: A Explanation: The bicycle model is a simplified kinematic model commonly used for ground vehicle motion analysis and planning. Question 9. The mean and variance are key parameters of which probability distribution? A) Binomial B) Exponential C) Gaussian (Normal) D) Poisson Answer: C Explanation: The Gaussian (normal) distribution is fully characterized by its mean and variance, and is widely used in state estimation. Question 10. Which sensor provides absolute global position data?
Exam
Question 12. Which sensor is most affected by poor lighting conditions? A) Radar B) IMU C) Monocular camera D) GPS Answer: C Explanation: Monocular cameras rely on ambient light, making them susceptible to poor lighting conditions, unlike radar or GPS. Question 13. What is the primary limitation of GPS in autonomous navigation? A) High cost B) Inability to work indoors or under dense foliage C) Low data rate D) Heavy weight Answer: B Explanation: GPS signals can be blocked or degraded in indoor environments, urban canyons, or under dense foliage, limiting their effectiveness.
Exam
Question 14. What is sensor calibration? A) Cleaning the sensor B) Adjusting sensor parameters to improve accuracy C) Replacing the sensor battery D) Installing new firmware Answer: B Explanation: Sensor calibration involves adjusting the sensor's internal parameters to ensure its measurements are accurate and reliable. Question 15. Which type of noise commonly affects IMU readings? A) Gaussian white noise B) Salt-and-pepper noise C) Poisson noise D) Compression noise Answer: A Explanation: IMU sensors are typically subject to Gaussian white noise due to random electronic fluctuations.
Exam
Question 18. Semantic segmentation in perception systems is used for: A) Dividing sensor data into time intervals B) Classifying each pixel in an image into a category C) Filtering out noise D) Compressing image data Answer: B Explanation: Semantic segmentation assigns a class label to every pixel in an image, aiding in detailed scene understanding. Question 19. Occupancy grid mapping is primarily used for: A) Storing 3D point clouds B) Representing free and occupied space in an environment C) Estimating sensor noise D) Calculating GPS coordinates Answer: B Explanation: Occupancy grids divide the environment into cells, each representing the probability of being occupied or free, useful for path planning and mapping.
Exam
Question 20. What is the main benefit of sensor fusion in autonomous navigation? A) Increases computational complexity B) Provides more robust and accurate state estimation C) Reduces sensor cost D) Eliminates the need for sensors Answer: B Explanation: Sensor fusion combines data from multiple sensors to improve the reliability, robustness, and accuracy of navigation and perception. Question 21. Which filter is most suitable for nonlinear state estimation in navigation systems? A) Linear Kalman Filter B) Extended Kalman Filter (EKF) C) Moving Average Filter D) Butterworth Filter Answer: B
Exam
Answer: B Explanation: Particle filters excel in scenarios with non-Gaussian, multimodal probability distributions, where Kalman filters may fail. Question 24. What is covariance intersection used for in sensor fusion? A) Sensor fault detection B) Fusing correlated estimates without known cross-correlations C) Reducing computational burden D) Increasing measurement noise Answer: B Explanation: Covariance intersection allows combining estimates from different sources when their correlations are unknown, preventing overconfidence. Question 25. Which localization method uses known landmarks and map matching? A) Relative localization B) Global localization C) Dead reckoning
Exam
D) Odometry Answer: B Explanation: Global localization determines the system's position relative to known landmarks or maps, unlike relative methods like odometry. Question 26. Which is a key limitation of odometry-based localization? A) It is highly accurate over long distances B) Errors accumulate over time due to sensor drift C) It requires external infrastructure D) It provides absolute position Answer: B Explanation: Odometry accumulates errors due to sensor noise and drift, making it unreliable for long-term localization without correction. Question 27. Markov Localization is characterized by: A) Using a single hypothesis for position B) Maintaining a probability distribution over possible positions C) Only working with GPS
Exam
B) Feature map C) Occupancy grid map D) Semantic map Answer: C Explanation: Occupancy grid maps divide the environment into a grid of cells, each with a probability of being occupied or free. Question 30. What is a semantic map? A) A map showing only geometric features B) A map enriched with high-level labels (e.g., road, pedestrian, vehicle) C) A map with only GPS coordinates D) A map containing only grid cells Answer: B Explanation: Semantic maps include both geometric and semantic information, assigning labels to different regions or objects for better scene understanding. Question 31. What is the main objective of SLAM (Simultaneous Localization and Mapping)?
Exam
A) Only to build a map B) Only to localize the robot C) To estimate both the map and the robot’s pose simultaneously D) To minimize computational cost Answer: C Explanation: SLAM simultaneously estimates the robot’s location and builds a map of the environment, handling the chicken-and-egg problem. Question 32. Which SLAM approach uses a graph structure to represent poses and constraints? A) Markov SLAM B) Graph-based SLAM C) Particle Filter SLAM D) Kalman Filter SLAM Answer: B Explanation: Graph-based SLAM represents robot poses and constraints as a graph, enabling efficient optimization and loop closure detection.
Exam
Question 35. Backend optimization in SLAM is used for: A) Real-time sensor fusion B) Refining the map and trajectory estimates using optimization techniques C) Calibrating the IMU D) Visualizing the environment Answer: B Explanation: Backend optimization adjusts the entire map and pose graph using optimization techniques (e.g., least squares) to minimize errors and improve accuracy. Question 36. Which algorithm guarantees finding the shortest path in a graph with non-negative edge weights? A) Dijkstra’s algorithm B) RRT* C) Breadth-First Search D) Particle Filter Answer: A
Exam
Explanation: Dijkstra’s algorithm guarantees the shortest path in graphs with non-negative weights by systematically exploring all nodes. Question 37. What is the main improvement of A* over Dijkstra's algorithm? A) It can handle negative weights B) Uses a heuristic to guide the search, reducing computation C) It is slower D) It does not guarantee optimality Answer: B Explanation: A* uses a heuristic to prioritize nodes closer to the goal, making the search faster while still guaranteeing the shortest path if the heuristic is admissible. Question 38. Which sampling-based algorithm is widely used for high- dimensional path planning? A) Breadth-First Search B) Dijkstra’s algorithm C) Rapidly-exploring Random Trees (RRT)
