Interview-grade
Practice
Real, interview-grade questions — anonymized & vetted. Subjective answers get AI feedback.
312 questions
- hardWhy is the Kalman filter optimal — and when does it stop being optimal?Kalman / EKF / UKF· Subjective · AI-gradedPractice question
- mediumBehavior of the Kalman gainKalman / EKF / UKF· Multiple choicePractice question
- easyEliminating steady-state errorControl Theory (PID / LQR / MPC)· Multiple choicePractice question
- easyEKF prediction covariance dimensionKalman / EKF / UKF· NumericalPractice question
- mediumWhy loop closure mattersSLAM· Subjective · AI-gradedPractice question
- mediumAdmissible heuristics in A*Motion Planning· Multiple choicePractice question
- easyEigenvalues and Eigenvectors of a Rotation MatrixLinear Algebra· Subjective · AI-gradedPractice question
- easyDeterminant of a Valid Rotation MatrixLinear Algebra· NumericalPractice question
- easyPosterior probability of an obstacle from a noisy proximity sensorProbability & Statistics· NumericalPractice question
- easyExplain the bias-variance tradeoff for an estimatorProbability & Statistics· Subjective · AI-gradedPractice question
- mediumGradient Descent vs. Gauss-Newton for Robotics Least-SquaresCalculus & Optimization· Subjective · AI-gradedPractice question
- mediumConvergence Behavior at a Saddle Point of the Cost SurfaceCalculus & Optimization· Multiple choicePractice question
- mediumWhy a 4x4 homogeneous transform instead of separate R and t?Rigid-Body Motion & Transforms· Subjective · AI-gradedPractice question
- mediumNet rotation angle of two composed 90-degree rotationsRigid-Body Motion & Transforms· NumericalPractice question
- mediumCamera Calibration and the Pinhole ModelComputer Vision· NumericalPractice question
- mediumRobustness of Vision Perception to Real-World ConditionsComputer Vision· Subjective · AI-gradedPractice question
- mediumComputing F1 from a confusion matrixML Fundamentals· NumericalPractice question
- mediumDiagnosing and addressing the bias-variance tradeoffML Fundamentals· Subjective · AI-gradedPractice question
- mediumSingularities in Inverse Kinematics: Detection and HandlingKinematics (FK / IK)· Subjective · AI-gradedPractice question
- mediumForward Kinematics of a Planar 2-Link ArmKinematics (FK / IK)· NumericalPractice question
- mediumProcess vs. Measurement Noise in Kalman Filter TuningBayesian Filtering· Subjective · AI-gradedPractice question
- mediumDiscrete Bayes Filter: One Measurement UpdateBayesian Filtering· NumericalPractice question
- hardDiagnosing and fixing exploding gradients in a deep recurrent/Transformer training runDeep Learning (PyTorch)· Subjective · AI-gradedPractice question
- hardCounting parameters in a 2D convolution layer (with bias)Deep Learning (PyTorch)· NumericalPractice question
- hardSkew-symmetry of the manipulator dynamics and the passivity propertyDynamics· Subjective · AI-gradedPractice question
- hardReflected inertia through a gear reductionDynamics· NumericalPractice question
- hardDeadly Triad and Off-Policy Value DivergenceReinforcement Learning· Subjective · AI-gradedPractice question
- hardGAE: Effect of the lambda Parameter on the Bias-Variance TradeoffReinforcement Learning· Multiple choicePractice question
- hardScalar LQR: solving the Riccati equation for the optimal gainControl Theory (PID / LQR / MPC)· NumericalPractice question
- hardMPC vs LQR: when the extra machinery of MPC earns its costControl Theory (PID / LQR / MPC)· Subjective · AI-gradedPractice question
- hardParticle Deprivation and the Cost of ResamplingParticle Filters· Subjective · AI-gradedPractice question
- hardComputing Effective Sample SizeParticle Filters· NumericalPractice question
- hardEKF Consistency, Linearization Error, and Choosing Between EKF and UKFSensor Fusion· Subjective · AI-gradedPractice question
- hardStatic 1D Fusion: Optimal Inverse-Variance Weighting and Fused VarianceSensor Fusion· NumericalPractice question
- mediumSampling-based planners: RRT vs. RRT* and the role of optimalityMotion Planning· Subjective · AI-gradedPractice question
- mediumHeuristic admissibility and A* optimalityMotion Planning· Multiple choicePractice question
- mediumChoosing and tuning DDS QoS for a ROS 2 sensor streamROS / ROS2· Subjective · AI-gradedPractice question
- mediumSizing a ROS 2 subscriber queue from publish rate and processing timeROS / ROS2· NumericalPractice question
- hardConsistency and the Observability of EKF-SLAMSLAM· Subjective · AI-gradedPractice question
- hardInformation-Form Fusion of Two Range MeasurementsSLAM· NumericalPractice question
- mediumDesigning the FBD and EOM for a friction-driven robot wheelClassical Mechanics· Subjective · AI-gradedPractice question
- easyAngular speed after a robot arm retracts a mass (conservation of angular momentum)Classical Mechanics· NumericalPractice question
- mediumChoosing a sample rate and anti-aliasing for an IMUSignal Processing· Subjective · AI-gradedPractice question
- easyAliased frequency from undersamplingSignal Processing· NumericalPractice question
- mediumWhy an IMU alone cannot be trusted for longSensors & How They Work· Subjective · AI-gradedPractice question
- easyHeading error from a constant gyro biasSensors & How They Work· NumericalPractice question
- mediumSizing a motor and gear ratio for a robot jointActuators & Motors· Subjective · AI-gradedPractice question
- easyOutput torque of a geared DC motor at a given currentActuators & Motors· NumericalPractice question
- mediumPriority inversion in a real-time control loopEmbedded & Real-Time· Subjective · AI-gradedPractice question
- easyUART throughput sizingEmbedded & Real-Time· NumericalPractice question
- mediumWhy co-train a VLA on web data instead of fine-tuning only on robot dataFoundation Models· Subjective · AI-gradedPractice question
- mediumAction tokenization vs. flow-matching action heads in VLAsFoundation Models· Multiple choicePractice question
- mediumWhy behavior cloning compounds errors and how DAgger helpsImitation Learning· Subjective · AI-gradedPractice question
- mediumChoosing a policy class for multimodal demonstrationsImitation Learning· Multiple choicePractice question
- hardToken binning vs. continuous action decoding in VLAsVision-Language-Action (VLA)· Subjective · AI-gradedPractice question
- mediumWhy internet pretraining helps VLA generalizationVision-Language-Action (VLA)· Multiple choicePractice question
- mediumWhy latent imagination is sample-efficientWorld Models & Sim-to-Real· Subjective · AI-gradedPractice question
- mediumWhat domain randomization fundamentally trades offWorld Models & Sim-to-Real· Multiple choicePractice question
- mediumDesigning a multi-robot training data formatData & ML Infrastructure· Subjective · AI-gradedPractice question
- mediumWhy GPU-parallel simulation changes RL training economicsData & ML Infrastructure· Multiple choicePractice question
- mediumWhy a reliable subscriber misses a latched topicMiddleware & System Architecture· Subjective · AI-gradedPractice question
- easyWhat the RMW layer is forMiddleware & System Architecture· Multiple choicePractice question
- mediumTime complexity of a Kalman filterKalman / EKF / UKF· Subjective · AI-gradedInterview
- mediumExplain the Kalman-filter update stepKalman / EKF / UKF· Subjective · AI-gradedInterview
- mediumAssumptions behind the Kalman filterKalman / EKF / UKF· Subjective · AI-gradedInterview
- mediumKF vs EKF vs UKFKalman / EKF / UKF· Subjective · AI-gradedInterview
- mediumProcess noise: meaning and effect of low vs high QKalman / EKF / UKF· Subjective · AI-gradedInterview
- hardEstimating bias and scale inside the filterKalman / EKF / UKF· Subjective · AI-gradedInterview
- mediumTransforming a state's covariance by a rotation and translationKalman / EKF / UKF· Subjective · AI-gradedInterview
- easyCovariance under a pure rotation — and is rotation affine?Kalman / EKF / UKF· Subjective · AI-gradedInterview
- easyWhere does the fused estimate lie between prediction and measurement?Kalman / EKF / UKF· Subjective · AI-gradedInterview
- mediumInitializing the state and covarianceKalman / EKF / UKF· Subjective · AI-gradedInterview
- hardTracking an object that switches motion modelsKalman / EKF / UKF· Subjective · AI-gradedInterview
- hardBringing machine learning into a state estimator (learning the Kalman gain)Kalman / EKF / UKF· Subjective · AI-gradedInterview
- hardWrite the Bayes-filter equationsBayesian Filtering· Subjective · AI-gradedInterview
- easyState estimation vs localizationBayesian Filtering· Subjective · AI-gradedInterview
- hardTesting, validating, and knowing a state estimate is goodBayesian Filtering· Subjective · AI-gradedInterview
- mediumKalman filter vs particle filter — when to use whichParticle Filters· Subjective · AI-gradedInterview
- mediumIncorporating measurement noise in a particle filterParticle Filters· Subjective · AI-gradedInterview
- hardHandling old / delayed (out-of-sequence) measurementsSensor Fusion· Subjective · AI-gradedInterview
- hardA confident but wrong measurement (small covariance, large error)Sensor Fusion· Subjective · AI-gradedInterview
- hardSystem design: tracking an object dropped from a drone under a parachuteKalman / EKF / UKF· Subjective · AI-gradedInterview
- mediumChallenges in designing a state estimatorBayesian Filtering· Subjective · AI-gradedInterview
- mediumHyperparameters of a state estimatorBayesian Filtering· Subjective · AI-gradedInterview
- hardSystem design: design a state estimation system (state, sensors, rotation)Kalman / EKF / UKF· Subjective · AI-gradedInterview
- hardState estimation in GPS-denied environmentsSensor Fusion· Subjective · AI-gradedInterview
- hardEstimating vehicle heading from IMU yaw rate and camera lane linesSensor Fusion· Subjective · AI-gradedInterview
- hard2D vs 3D ego-state estimation, and IMU-yaw drift over a 1 km loopSensor Fusion· Subjective · AI-gradedInterview
- mediumHow do you know your measurement noise is Gaussian?Bayesian Filtering· Subjective · AI-gradedInterview
- mediumHandling high-bandwidth / high-rate sensors in a fusion systemSensor Fusion· Subjective · AI-gradedInterview
- hardSystem design: localization for a robot in orchard / vineyard rowsSensor Fusion· Subjective · AI-gradedInterview
- hardSystem design: odometry for an RC car doing hoops and jumpsSensor Fusion· Subjective · AI-gradedInterview
- hardObservability from one beacon vs two beacons in 2DBayesian Filtering· Subjective · AI-gradedInterview
- mediumExplain the pinhole camera modelSensors & How They Work· Subjective · AI-gradedInterview
- mediumMain parameters in camera calibrationSensors & How They Work· Subjective · AI-gradedInterview
- mediumMajor specifications when selecting a cameraSensors & How They Work· Subjective · AI-gradedInterview
- hardLidar–camera calibrationSensors & How They Work· Subjective · AI-gradedInterview
- hardCalibrating IMU, lidar, and camera togetherSensors & How They Work· Subjective · AI-gradedInterview
- hardCalibrating an IMU (intrinsics)Sensors & How They Work· Subjective · AI-gradedInterview
- mediumLidar vs camera — advantages and disadvantagesSensors & How They Work· Subjective · AI-gradedInterview
- mediumThe blooming effect in lidarSensors & How They Work· Subjective · AI-gradedInterview
- mediumImpact of retroreflectors on lidar dataSensors & How They Work· Subjective · AI-gradedInterview
- mediumCommon lidar point-cloud issuesSensors & How They Work· Subjective · AI-gradedInterview
- hardRetroreflector returns vs sensor-blockage returnsSensors & How They Work· Subjective · AI-gradedInterview
- hard1550 nm vs 905 nm lidar (and their detectors)Sensors & How They Work· Subjective · AI-gradedInterview
- mediumDetecting blockage/occlusion on a cameraSensors & How They Work· Subjective · AI-gradedInterview
- mediumDetecting blockage on a lidarSensors & How They Work· Subjective · AI-gradedInterview
- mediumFalse positives and false negatives in blockage detectionSensors & How They Work· Subjective · AI-gradedInterview
- hardChoosing lidar specs and the region of interestSensors & How They Work· Subjective · AI-gradedInterview
- hardScanning-mirror lidar vs mechanical spinning lidarSensors & How They Work· Subjective · AI-gradedInterview
- hardRadar's poor cross-range (perpendicular) accuracySensors & How They Work· Subjective · AI-gradedInterview
- hardModeling GPS noise and computing DOPSensors & How They Work· Subjective · AI-gradedInterview
- hardGPS velocity noise, and deriving velocity covariance from positionSensors & How They Work· Subjective · AI-gradedInterview
- mediumModeling noise for a sensor that doesn't report its uncertaintySensors & How They Work· Subjective · AI-gradedInterview
- mediumSimulate a lidar: range to a wall, and 3D distanceSensors & How They Work· Subjective · AI-gradedInterview
- mediumHow do you fuse multiple sensory inputs?Sensor Fusion· Subjective · AI-gradedInterview
- hardEarly vs late fusion, and how each handles delayed measurementsSensor Fusion· Subjective · AI-gradedInterview
- mediumFusing measurements arriving at different frequenciesSensor Fusion· Subjective · AI-gradedInterview
- hardInterpolation-based vs filtering-based fusionSensor Fusion· Subjective · AI-gradedInterview
- mediumFusing synchronized measurementsSensor Fusion· Subjective · AI-gradedInterview
- hardFusing lidar and camera for a trackerSensor Fusion· Subjective · AI-gradedInterview
- hardFusing multiple lidars running at different frequenciesSensor Fusion· Subjective · AI-gradedInterview
- hardOrientation from IMU only — and handling driftSensor Fusion· Subjective · AI-gradedInterview
- hardSystem design: ML-based odometry for an AV (IMU + GPS + wheel odometry)Sensor Fusion· Subjective · AI-gradedInterview
- hardWheel-odometry velocity on a skid-steer robotSensor Fusion· Subjective · AI-gradedInterview
- hardDealing with unsynchronized sensor readingsSensor Fusion· Subjective · AI-gradedInterview
- hardAccumulating IMU readings in the filterSensor Fusion· Subjective · AI-gradedInterview
- mediumModeling wheel slipSensor Fusion· Subjective · AI-gradedInterview
- mediumExplain the object-tracking pipelineMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumFeatures used for object trackingMulti-Object Tracking· Subjective · AI-gradedInterview
- hardTypes of classical tracking algorithmsMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumTwo association cost functions — and when not to use themMulti-Object Tracking· Subjective · AI-gradedInterview
- hardThe Hungarian algorithm and its complexityMulti-Object Tracking· Subjective · AI-gradedInterview
- hardMultiple Hypothesis Tracking (MHT)Multi-Object Tracking· Subjective · AI-gradedInterview
- hardJoint Probabilistic Data Association (JPDA)Multi-Object Tracking· Subjective · AI-gradedInterview
- hardMHT vs JPDA — why prefer MHTMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumDifferent ways of doing data associationMulti-Object Tracking· Subjective · AI-gradedInterview
- hardQualitative and quantitative tracking metricsMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumTwo detections inside one track's gateMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumChoosing the state for a trackerMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumHyperparameters of a trackerMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumHandling varying bounding-box sizes in the trackerMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumChallenges in designing a trackerMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumML vs non-ML tracking approachesMulti-Object Tracking· Subjective · AI-gradedInterview
- hardDesign changes that improve a trackerMulti-Object Tracking· Subjective · AI-gradedInterview
- hardTesting and evaluating a trackerMulti-Object Tracking· Subjective · AI-gradedInterview
- hardModeling lidar noise for a trackerMulti-Object Tracking· Subjective · AI-gradedInterview
- hardDealing with noise in segmented objectsMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumFalse positives at the tracker input, and how to handle themMulti-Object Tracking· Subjective · AI-gradedInterview
- hardOverlapping regions in multi-lidar trackingMulti-Object Tracking· Subjective · AI-gradedInterview
- hardMaking a tracker efficient (time, memory, design)Multi-Object Tracking· Subjective · AI-gradedInterview
- mediumWhy you need ego-motion compensation for tracked objectsMulti-Object Tracking· Subjective · AI-gradedInterview
- hardTracking slow-moving objectsMulti-Object Tracking· Subjective · AI-gradedInterview
- hardEvaluating velocity precision of tracksMulti-Object Tracking· Subjective · AI-gradedInterview
- hardTracking animals suddenly crossing a highway laterallyMulti-Object Tracking· Subjective · AI-gradedInterview
- hardStatic objects that appear to move due to perspective changeMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumObstacle avoidance on detections vs on tracksMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumLidar disadvantages specifically for trackingMulti-Object Tracking· Subjective · AI-gradedInterview
- hardTracking corners under a perspective shift, with orientation noiseMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumTuning process noise for a trackerMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumChoosing the maximum range to track objectsMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumFeature similarity measures beyond L1/L2Multi-Object Tracking· Subjective · AI-gradedInterview
- hardOrientation with four discrete hypotheses and Gaussian mixturesMulti-Object Tracking· Subjective · AI-gradedInterview
- hardWhen a past-frame association would have made a better trackMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumTracking a dog running to fetch a ball that has landedMulti-Object Tracking· Subjective · AI-gradedInterview
- hardTracking a dog that jumps to catch the ball mid-airMulti-Object Tracking· Subjective · AI-gradedInterview
- hardTracking a dog that somersaults to catch the ballMulti-Object Tracking· Subjective · AI-gradedInterview
- mediumTracking the trajectory of a thrown ballMulti-Object Tracking· Subjective · AI-gradedInterview
- easy3×3 matrix with two linearly dependent columns — invertible?Linear Algebra· Subjective · AI-gradedPractice question
- easyWhen can a matrix not be inverted?Linear Algebra· Subjective · AI-gradedPractice question
- mediumSingular values vs eigenvaluesLinear Algebra· Subjective · AI-gradedPractice question
- easyNull space of an invertible matrixLinear Algebra· Subjective · AI-gradedPractice question
- easyColumn space of an invertible matrixLinear Algebra· Subjective · AI-gradedPractice question
- hardSolve XA = BX for the matrix XLinear Algebra· Subjective · AI-gradedPractice question
- easyIf Ax = Ay with x ≠ y, is A singular?Linear Algebra· Subjective · AI-gradedPractice question
- easyIf AB is invertible and A is invertible, is B invertible?Linear Algebra· Subjective · AI-gradedPractice question
- easyDeterminant of a rank-deficient n×n matrixLinear Algebra· Subjective · AI-gradedPractice question
- easyDeterminant of the identity matrixLinear Algebra· NumericalPractice question
- mediumRank of AᵀALinear Algebra· Subjective · AI-gradedPractice question
- easyProduct of all eigenvalues of a square matrixLinear Algebra· Subjective · AI-gradedPractice question
- hardMaximum dimension of a subspace of singular matricesLinear Algebra· Subjective · AI-gradedPractice question
- easyRank of AᵀLinear Algebra· Subjective · AI-gradedPractice question
- easyConjugate transpose of a real matrixLinear Algebra· Subjective · AI-gradedPractice question
- hardRank of a matrix vs rank of a tensorLinear Algebra· Subjective · AI-gradedPractice question
- easyExplain eigenvectorsLinear Algebra· Subjective · AI-gradedInterview
- easyHow the determinant changes when a matrix is scaledLinear Algebra· Subjective · AI-gradedInterview
- mediumWhat is the SVD?Linear Algebra· Subjective · AI-gradedInterview
- mediumSolving Ax = 0Linear Algebra· Subjective · AI-gradedInterview
- mediumIntersection of two lines in 3D spaceLinear Algebra· Subjective · AI-gradedInterview
- hardPoint closest to multiple lines in spaceLinear Algebra· Subjective · AI-gradedInterview
- mediumValid rotation matrix and valid quaternionLinear Algebra· Subjective · AI-gradedInterview
- hardCoins passed around a round tableProbability & Statistics· Subjective · AI-gradedInterview
- mediumPDF of the sum of two uniformsProbability & Statistics· Subjective · AI-gradedInterview
- mediumExpected value of a coin-flip game — should you bet?Probability & Statistics· Subjective · AI-gradedInterview
- mediumJoint PDF of two independent normalsProbability & Statistics· Subjective · AI-gradedInterview
- mediumHow do you know a Gaussian is valid?Probability & Statistics· Subjective · AI-gradedInterview
- mediumShape of a 2D Gaussian for different correlationsProbability & Statistics· Subjective · AI-gradedInterview
- mediumZero covariance but dependent random variablesProbability & Statistics· Subjective · AI-gradedPractice question
- mediumConvolution of a square pulse with itselfSignal Processing· Subjective · AI-gradedInterview
- mediumFourier transform, and the transform of a sineSignal Processing· Subjective · AI-gradedInterview
- easySolve dx/dt = a·f(t) where x = f(t)Calculus & Optimization· Subjective · AI-gradedInterview
- mediumWhat is a homography, and its assumptions?Computer Vision· Subjective · AI-gradedInterview
- easyExplain stereo vision in simple termsComputer Vision· Subjective · AI-gradedInterview
- hardExplain epipolar geometryComputer Vision· Subjective · AI-gradedInterview
- mediumStereo rectification — what and whenComputer Vision· Subjective · AI-gradedInterview
- mediumImage features — detectors and descriptorsComputer Vision· Subjective · AI-gradedInterview
- mediumEstimating depth from a feature seen in two framesComputer Vision· Subjective · AI-gradedInterview
- hardFundamental vs essential matrixComputer Vision· Subjective · AI-gradedInterview
- mediumHow does RANSAC work?Computer Vision· Subjective · AI-gradedInterview
- mediumHandling outliers in geometric estimationComputer Vision· Subjective · AI-gradedInterview
- hardGeometric fitting of a bounding box to segmented pointsComputer Vision· Subjective · AI-gradedInterview
- hardClustering a lidar point cloudComputer Vision· Subjective · AI-gradedInterview
- mediumClustering algorithms and their trade-offsComputer Vision· Subjective · AI-gradedInterview
- hardDesign a stop-sign detector with classical methods (no ML)Computer Vision· Subjective · AI-gradedInterview
- mediumDesign a road/ground plane detectorComputer Vision· Subjective · AI-gradedInterview
- hardMeasuring vehicle speed from a fixed monocular cameraComputer Vision· Subjective · AI-gradedInterview
- hardMapping and localization with 2D lidar scansSLAM· Subjective · AI-gradedInterview
- hardSteps of Visual SLAM / LOAMSLAM· Subjective · AI-gradedInterview
- hardIncorporating late measurements in SLAMSLAM· Subjective · AI-gradedInterview
- mediumFinding frame-to-frame correspondences in SLAMSLAM· Subjective · AI-gradedInterview
- hardMapping in a multi-robot setupSLAM· Subjective · AI-gradedInterview
- mediumDebugging an odometry issueSLAM· Subjective · AI-gradedInterview
- hardEvaluating a SLAM system (online and offline)SLAM· Subjective · AI-gradedInterview
- mediumSLAM systems suited to online vs offline mappingSLAM· Subjective · AI-gradedInterview
- hardFusing map information in a filter — global or local frame?SLAM· Subjective · AI-gradedInterview
- hardDeciding when to update an occupancy map from new lidar dataSLAM· Subjective · AI-gradedInterview
- hardOptimization routines in Visual SLAMSLAM· Subjective · AI-gradedInterview
- mediumLayered costmapsMotion Planning· Subjective · AI-gradedInterview
- mediumAssigning costs to different objects in a costmapMotion Planning· Subjective · AI-gradedInterview
- mediumPseudocode for a planner (A*)Motion Planning· Subjective · AI-gradedInterview
- hardThree categories of planners — and when to use eachMotion Planning· Subjective · AI-gradedInterview
- hardWhy incorporate non-holonomic constraints in the plannerMotion Planning· Subjective · AI-gradedInterview
- hardLearning-based planningMotion Planning· Subjective · AI-gradedInterview
- mediumExplain SVM, margin, and its lossML Fundamentals· Subjective · AI-gradedInterview
- mediumWhat is hinge loss?ML Fundamentals· Subjective · AI-gradedInterview
- mediumDifferent kinds of losses, and multi-class lossML Fundamentals· Subjective · AI-gradedInterview
- mediumHow do you select features?ML Fundamentals· Subjective · AI-gradedInterview
- mediumWhat is SGD, and why 'stochastic'?ML Fundamentals· Subjective · AI-gradedInterview
- mediumHandling a lack of labeled dataML Fundamentals· Subjective · AI-gradedInterview
- mediumWeight initialization, and why randomML Fundamentals· Subjective · AI-gradedInterview
- hardFocal lossDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- mediumVoxels — advantages and disadvantagesDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- hardWhat are PointPillars?Deep Learning (PyTorch)· Subjective · AI-gradedInterview
- mediumThe VGG architectureDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- mediumConvolution and output dimensionsDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- mediumOptimizers and how Adam worksDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- mediumVanishing gradients — what and how to identifyDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- mediumUpgrading a detector from YOLO-small to YOLO-mediumDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- hardDetecting small bounding boxesDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- hardMetrics for small-bounding-box detection when IoU is unreliableDeep Learning (PyTorch)· Subjective · AI-gradedInterview
- hardFusing RGB and thermal images for detection (wildfire example)Deep Learning (PyTorch)· Subjective · AI-gradedInterview
- hardLoss function for pose (position + orientation)Deep Learning (PyTorch)· Subjective · AI-gradedInterview
- mediumHardware fault or software: which do you check first?Embedded & Real-Time· Subjective · AI-gradedInterview
- mediumIssues when porting code to a target platformEmbedded & Real-Time· Subjective · AI-gradedInterview
- hardRunning ML algorithms on an embedded platformEmbedded & Real-Time· Subjective · AI-gradedInterview
- hardWhy avoid (or be careful with) multithreading on a target platformEmbedded & Real-Time· Subjective · AI-gradedInterview
- mediumDescribing a hard technical/algorithmic challengeSystem Design & Ownership· Subjective · AI-gradedInterview
- hardDesign a snack-delivery robot for indoor environmentsSystem Design & Ownership· Subjective · AI-gradedInterview
- mediumDeciding KPIs and their targetsSystem Design & Ownership· Subjective · AI-gradedInterview
- mediumChecking the quality of collected dataData & ML Infrastructure· Subjective · AI-gradedInterview
- mediumEnsuring tests cover everythingSystem Design & Ownership· Subjective · AI-gradedInterview
- mediumConvincing a downstream consumer team about your changesSystem Design & Ownership· Subjective · AI-gradedInterview
- hardPredicting the trajectory of an oncoming vehicle at an intersectionMotion Planning· Subjective · AI-gradedInterview
- hardDesign a perception system for a truck operating at nightSensors & How They Work· Subjective · AI-gradedInterview
- mediumBrushless (BLDC) vs brushed DC motors for a robot jointActuators & Motors· Subjective · AI-gradedPractice question
- easyBack-EMF as a motor speeds upActuators & Motors· Multiple choicePractice question
- mediumTorque from the linear torque–speed curveActuators & Motors· NumericalPractice question
- mediumCoriolis and centrifugal terms — why a rotating frame adds themClassical Mechanics· Subjective · AI-gradedPractice question
- mediumTorque to angularly accelerate a linkClassical Mechanics· NumericalPractice question
- easyRolling without slippingClassical Mechanics· Multiple choicePractice question
- mediumThe manipulator equation: what each term meansDynamics· Subjective · AI-gradedPractice question
- mediumForward vs inverse dynamicsDynamics· Multiple choicePractice question
- mediumNatural frequency of a compliant jointDynamics· NumericalPractice question
- hardWhat a robotics foundation model buys you over task-specific modelsFoundation Models· Subjective · AI-gradedPractice question
- mediumWhy cross-embodiment training helpsFoundation Models· Multiple choicePractice question
- hardRunning a large VLA at control rateFoundation Models· Subjective · AI-gradedPractice question
- mediumCovariate shift in behavior cloningImitation Learning· Subjective · AI-gradedPractice question
- mediumWhy diffusion policies handle multimodal actionsImitation Learning· Multiple choicePractice question
- hardInverse RL vs behavior cloningImitation Learning· Subjective · AI-gradedPractice question
- mediumDenavit–Hartenberg parametersKinematics (FK / IK)· Subjective · AI-gradedPractice question
- mediumWhat happens at a kinematic singularityKinematics (FK / IK)· Multiple choicePractice question
- easyEnd-effector distance for a right-angle elbowKinematics (FK / IK)· NumericalPractice question
- hardDDS discovery and its cost at scaleMiddleware & System Architecture· Subjective · AI-gradedPractice question
- mediumQoS reliability for a high-rate lidar streamMiddleware & System Architecture· Multiple choicePractice question
- hardROS 2 executors and callback groupsMiddleware & System Architecture· Subjective · AI-gradedPractice question
- mediumOn-policy vs off-policy RLReinforcement Learning· Subjective · AI-gradedPractice question
- mediumWhy PPO clips the probability ratioReinforcement Learning· Multiple choicePractice question
- hardReward shaping without changing the optimal policyReinforcement Learning· Subjective · AI-gradedPractice question
- mediumRotation representations: Euler vs quaternion vs matrixRigid-Body Motion & Transforms· Subjective · AI-gradedPractice question
- easyDegrees of freedom of a free rigid bodyRigid-Body Motion & Transforms· NumericalPractice question
- mediumInverting a homogeneous transformRigid-Body Motion & Transforms· Subjective · AI-gradedPractice question
- mediumtf2: the transform tree and extrapolation errorsROS / ROS2· Subjective · AI-gradedPractice question
- easyTopics vs services vs actionsROS / ROS2· Multiple choicePractice question
- hardZero-copy and intra-process communicationROS / ROS2· Subjective · AI-gradedPractice question
- mediumAction chunking in VLAsVision-Language-Action (VLA)· Subjective · AI-gradedPractice question
- mediumFast head, slow backboneVision-Language-Action (VLA)· Multiple choicePractice question
- hardLanguage grounding in a VLA and its failure modesVision-Language-Action (VLA)· Subjective · AI-gradedPractice question
- mediumModel-based vs model-free RLWorld Models & Sim-to-Real· Subjective · AI-gradedPractice question
- mediumWhy train a policy in imagination (Dreamer)World Models & Sim-to-Real· Multiple choicePractice question
- hardCompounding error over a world-model rolloutWorld Models & Sim-to-Real· Subjective · AI-gradedPractice question
- mediumConvex vs non-convex optimizationCalculus & Optimization· Subjective · AI-gradedPractice question
- mediumNewton's method vs gradient descentCalculus & Optimization· Multiple choicePractice question
- hardControllability and observabilityControl Theory (PID / LQR / MPC)· Subjective · AI-gradedPractice question
- easyPoles and stabilityControl Theory (PID / LQR / MPC)· Multiple choicePractice question
- hardImportance sampling and the proposal distributionParticle Filters· Subjective · AI-gradedPractice question
- mediumIIR vs FIR filters for real-time roboticsSignal Processing· Subjective · AI-gradedPractice question
- hardBuilding an evaluation harness for a perception modelData & ML Infrastructure· Subjective · AI-gradedPractice question
- mediumCurating a large robot demonstration datasetData & ML Infrastructure· Subjective · AI-gradedPractice question
- mediumHow to structure a robotics system-design answerSystem Design & Ownership· Subjective · AI-gradedPractice question
- hardDesign the autonomy stack for a sidewalk delivery robotSystem Design & Ownership· Subjective · AI-gradedPractice question
- hardDesign a warehouse bin-picking robotSystem Design & Ownership· Subjective · AI-gradedPractice question
- mediumChoosing a YOLO model variant (small vs medium vs large)Deep Learning (PyTorch)· Subjective · AI-gradedInterview