▸case-01 We are building a path planner for an Ackermann-steering autonomous vehicle operating in a tight parking lot. A team member suggested using standard RRT or PRM in a 2D position space (x, y) because sampling-based planners easily handle obstacles. Explain why standard Euclidean 2D sampling is inadequate for car-like dynamics and specify the appropriate planning algorithm approach for nonholonomic car-like robots. | pass→pass | 22,831 | 55,615 | +144% | 1 | 1 | 0% | 3,296 | 4,798 | +46% | 0 | 0 | — |
▸case-02 A developer is configuring a single-query planner for a 6-DOF robotic manipulator arm. They selected standard RRT-Connect because it finds a solution path quickly, but the resulting paths are excessively long and erratic. The team requires path length to converge to optimal length as planning time increases. Specify the planner algorithm to use instead of RRT-Connect and explain its key algorithmic mechanism for optimality. | pass→pass | 10,909 | 16,928 | +55% | 1 | 1 | 0% | 2,075 | 2,916 | +41% | 0 | 0 | — |
▸case-03 In a robotic assembly cell, a 6-DOF robot arm executes thousands of pick-and-place motions per hour between changing start and goal poses in a completely static workspace with fixed obstacles. The current setup runs RRT-Connect from scratch for every motion query, causing CPU bottlenecks. Recommend an alternative sampling-based planning approach optimized for static multi-query scenarios and explain how it improves query speed. | pass→pass | 15,284 | 16,269 | +6% | 1 | 1 | 0% | 2,386 | 3,044 | +28% | 0 | 0 | — |
▸case-04 When refining a joint-space trajectory for an industrial arm moving through a narrow obstacle window, the optimization process using functional gradient descent frequently fails because the trajectory passes through thin obstacle corners when collision weights compete with path smoothness weights. Recommend an optimization method that enforces exact collision constraints rather than unconstrained gradient penalties. | pass→pass | 18,159 | 21,132 | +16% | 1 | 1 | 0% | 2,911 | 3,502 | +20% | 0 | 0 | — |
▸case-05 An open-loop kinematic path generator outputs a sequence of discrete joint angle waypoints (q0, q1, ..., qn) for a 7-DOF arm. A developer attempts to send these waypoints directly to the motor drives at fixed time intervals. Explain why fixed time-stepping fails and specify the algorithm required to convert geometric waypoints into continuous time parameterization respecting velocity and acceleration limits. | pass→pass | 20,257 | 17,468 | -14% | 1 | 1 | 0% | 3,739 | 3,529 | -6% | 0 | 0 | — |
▸case-06 A developer is implementing a state validity checker method isValid(const State* state) in C++ for a high-DOF manipulator moving near an OctoMap dynamic point cloud. They currently measure Euclidean distance from joint centers to simple spheres, causing false-negative collision reports. Name the standard collision checking library to integrate within isValid for mesh-to-pointcloud broadphase collision detection. | pass→pass | 6,964 | 4,648 | -33% | 1 | 1 | 0% | 1,016 | 1,044 | +3% | 0 | 0 | — |
▸case-07 A sampling-based planner returns a raw piecewise linear sequence of joint positions. Sending this path to the low-level controller causes severe robot shaking at path vertices. A team member suggests applying Gaussian blur smoothing over joint coordinates. Explain why naive spatial blurring is dangerous and specify the standard two-phase path post-processing procedure for sampling-based paths. | fail→fail | 20,245 | 18,978 | -6% | 1 | 1 | 0% | 2,941 | 3,626 | +23% | 0 | 0 | — |
▸case-08 We need to optimize joint trajectories in an environment where cost gradients cannot be computed analytically due to black-box voxel cost maps and non-differentiable torque penalty functions. Standard gradient-based trajectory optimizers like CHOMP fail. Recommend a trajectory optimization algorithm that functions without cost derivatives and explain its sampling mechanism. | pass→pass | 19,241 | 22,854 | +19% | 1 | 1 | 0% | 2,822 | 4,456 | +58% | 0 | 0 | — |
▸case-09 A pick-and-place manipulator needs to find any collision-free path as quickly as possible without caring about path length or smoothness. An engineer suggests configuring RRT* with a high maximum planning time. Explain why RRT* is inefficient for pure fast feasibility and specify the fastest single-query sampling algorithm. | pass→pass | 12,460 | 14,546 | +17% | 1 | 1 | 0% | 2,039 | 2,582 | +27% | 0 | 0 | — |
▸case-10 We require an optimal path planner for high-dimensional state space navigation that retains asymptotic optimality like RRT*, but evaluates sample connections in batch using dynamic programming over pre-sampled points to achieve significantly faster runtime. Specify the planning algorithm that fulfills these requirements. | pass→pass | 11,885 | 12,399 | +4% | 1 | 1 | 0% | 2,037 | 2,439 | +20% | 0 | 0 | — |
▸case-11 For a mobile robot with Ackermann kinematic constraints navigating a mapped warehouse grid, an engineer proposes running standard 2D grid A* with an 8-connected neighborhood graph. Explain why standard 8-connected grid search fails and specify the discrete control set structure required for kinematically feasible A* graph search. | pass→pass | 24,729 | 19,903 | -20% | 1 | 1 | 0% | 3,678 | 3,749 | +2% | 0 | 0 | — |
▸case-12 In a ROS MoveIt manipulation setup, custom planner parameters in ompl_planning.yaml are not taking effect because all planners use global hardcoded defaults. Detail how planner parameters such as range and longest_valid_segment_fraction should be defined inside ompl_planning.yaml for a specific group. | fail→fail | 18,764 | 14,591 | -22% | 1 | 1 | 0% | 3,198 | 2,965 | -7% | 0 | 0 | — |
▸case-13 When planning trajectory paths for a dynamic quadrotor UAV with mass and thruster limits (F=ma), a planner configured with standard geometric RRT generates paths with impossible step acceleration jumps. Detail the architectural changes needed in sampling-based planning to handle dynamic control inputs. | pass→pass | 23,246 | 24,572 | +6% | 1 | 1 | 0% | 4,165 | 4,728 | +14% | 0 | 0 | — |
▸case-14 A developer is configuring an OMPL state space for a 3D flying rigid body. They instantiated RealVectorStateSpace(6) and set bounds [-3.14, 3.14] on the last 3 dimensions to represent Euler rotation angles. Explain why this state space representation causes planning failures and specify the correct state space class. | pass→pass | 13,350 | 12,047 | -10% | 1 | 1 | 0% | 2,381 | 2,223 | -7% | 0 | 0 | — |
▸case-15 Inside path-planning-algorithm.js, we need to instantiate a C++/Node binding setup for solving sampling-based planning problems. Specify the standard initialization structure using OMPL SimpleSetup, space definition, and state validity checker assignment. | pass→pass | 15,198 | 17,898 | +18% | 1 | 1 | 0% | 3,249 | 4,252 | +31% | 0 | 0 | — |
▸case-16 Inside trajectory-optimization.js, specify how to formulate a trajectory optimization cost function that balances obstacle distance margins and kinematic smoothness for a discrete sequence of N trajectory states. | pass→pass | 22,956 | 19,296 | -16% | 1 | 1 | 0% | 4,656 | 4,355 | -6% | 0 | 0 | — |
▸case-17 To determine whether RRT*, PRM*, or FMT* is best for a robot arm working in varying obstacle density scenarios, a developer wrote a manual for loop that measures execution timing with standard clock calls. Explain why this manual timing approach is inadequate for motion planning evaluations and specify the standard benchmarking tool class provided by OMPL. | fail→pass | 11,876 | 127,826 | +976% | 1 | 1 | 0% | 1,932 | 2,447 | +27% | 0 | 0 | — |
▸case-18 Inside moveit-manipulation-planning.js, describe the sequence of steps required to send a joint goal request to MoveIt's planning pipeline and handle the returned motion plan. | pass→pass | 14,834 | 14,944 | +1% | 1 | 1 | 0% | 2,895 | 3,171 | +10% | 0 | 0 | — |
▸case-19 Inside nav2-navigation-setup.js, specify how to configure the Nav2 planner server plugins to enable global path planning for a nonholonomic car-like mobile robot. | fail→fail | 16,118 | 15,049 | -7% | 1 | 1 | 0% | 3,002 | 3,168 | +6% | 0 | 0 | — |
▸case-20 During motion planning validation in MoveIt, a 6-DOF arm path is reported as valid by OMPL, but during execution the arm collides with a thin 5mm metal sheet placed between sampled waypoints. Explain why the collision checker missed the obstacle and specify the parameter to adjust. | pass→pass | 11,748 | 9,447 | -20% | 1 | 1 | 0% | 1,974 | 1,753 | -11% | 0 | 0 | — |
▸case-21 A developer wants to precompute a roadmap in a static workspace for multi-query planning, but standard PRM returns suboptimal paths that loop around obstacles inefficiently even as the roadmap size grows to infinity. Specify the variant of PRM that guarantees asymptotic path optimality and explain how its connection radius rule differs from standard PRM. | pass→pass | 13,374 | 11,706 | -12% | 1 | 1 | 0% | 2,325 | 2,374 | +2% | 0 | 0 | — |
▸case-22 We are calibrating joint drive electronics for a 6-DOF manipulator arm. Describe the Ziegler-Nichols step-response method for tuning PID loop proportional (Kp), integral (Ki), and derivative (Kd) gain constants to eliminate joint overshoot. | fail→fail | 20,166 | 19,516 | -3% | 1 | 1 | 0% | 3,451 | 3,701 | +7% | 0 | 0 | — |
▸case-23 We are defining a new robot description in ROS. Provide the exact URDF XML structure for defining link mass, inertial tensor values (Ixx, Ixy, ...), and visual mesh file paths inside link tags. | fail→fail | 7,243 | 10,215 | +41% | 1 | 1 | 0% | 1,475 | 2,338 | +59% | 0 | 0 | — |
▸case-24 Configure an Extended Kalman Filter (robot_localization node) in ROS 2 to fuse raw wheel encoder odometry (/odom) and 9-DOF IMU sensor data (/imu/data) into a filtered /odometry/filtered topic. | fail→fail | 19,140 | 23,980 | +25% | 1 | 1 | 0% | 3,672 | 5,045 | +37% | 0 | 0 | — |