Autonomous Decision Making Process TRIZ Cases

Human driving trajectories flag edge cases where vehicle path planning diverges, enabling targeted retraining for safer, more compliant behavior.

Co-locating lidar and cameras in separated housing sections cuts occlusion and improves close-range object classification for autonomous vehicles.

Redundant pose measurement systems are quality-checked in real time so an autonomous vehicle can switch to the best pose source when failures occur.

Tracking representative vehicles over time lets an autonomous vehicle pick the fastest lane while avoiding congestion from frequent lane weaving.

When a local failure is detected, the vehicle selects a safe pullover and pickup area, then requests another vehicle to continue the trip.

Server-built behavior models use sensor, world-view, and action data to improve surrounding-object prediction for autonomous vehicles.

Pre-event and post-event vehicle data sharing helps identify fault automatically and trigger faster response after crashes or near misses.

A secondary control system uses redundant optical and inertial sensing to maintain lateral velocity estimation for safe autonomous controlled stops.

Pixel-region exposure timing uses distance data to balance backlit and low-light vehicle ToF images without global shutter tradeoffs.

Spectral error modeling sets vehicle trajectory safety margins before road testing, reducing collision risk from controller and actuator delays.

Route and vehicle data are used to preselect and push relevant playback content, improving passenger engagement without complex in-trip selection.

Path polygons, agent trajectories, and time-space overlap analysis help vehicles avoid collisions without unnecessary yielding or traffic delays.

When EV charge drops too low, nearby home chargers are ranked and a recharge rendezvous is scheduled to extend travel and ease range anxiety.

Route landmarks and fleet assistance enable in-motion sensor recalibration and trailer pose tracking for safer autonomous truck operation.

Adaptive sensor angles and ranges cut blind areas during turning and slope changes in double-axle cab-less mining vehicles.

Zone-based sensor failure risk and dynamic adjustment define real-time capability boundaries for safer Level 4 autonomous driving.

Independent steering and speed command modules help autonomous vehicles follow trajectories, adapt to obstacles, and stay within lateral and speed limits.

Virtual blocked regions are stored in the map and handled like detected obstacles, improving robot navigation safety without physical markers.

Route-based landmark selection enables frequent sensor recalibration in autonomous vehicles while reducing computational load and drift.

Pre-stored yellow-light durations let autonomous vehicles predict red-light timing from signal state, reducing abrupt braking and red-light running.

Operation amount and duration thresholds let the vehicle shift between autonomous, cooperative, and manual driving to reduce switch burden.

Sensors predict whether cyclists or vehicles will enter the door exit zone, enabling warnings or door locking to prevent unsafe passenger exit.

Track-based rail agent prediction filters probable paths so autonomous vehicles can plan yielding or stopping with smoother, safer navigation.

Dual coordinate frames project estimated vehicle state onto the planned path to smooth autonomous trajectory correction and cut computation time.

Pre-collected static reflection maps enable automatic lidar calibration by minimizing map differences, improving obstacle detection with less calibration time.

Operation and duration thresholds shift driving among autonomous, cooperative, and manual modes to ease temporary driver intervention.

A vehicle API and remote mediator coordinate convoy actions from authorized third-party commands while preserving reliable fleet response.

Gradient-based trajectory planning predicts moving obstacles and cuts path planning delay for safer autonomous vehicle navigation.

Mixed-integer coordination across conflict zones plans vehicle timing and speeds in real time to improve mixed-traffic safety and travel efficiency.

Decision trees infer explainable vehicle trajectory standards from labeled driving data, helping autonomous vehicles choose safe, comfortable paths.

A dual-actuation wheel brake cylinder switches from hydraulic pressure to electromechanical braking to maintain autonomous stopping after leaks or failures.

Sensor fusion and an empathy processing layer let an autonomous vehicle detect accidents or disabled vehicles and coordinate help.

Leader driving states and track patterns are learned so a follower vehicle can quickly rejoin group driving after traffic or road-condition deviation.

Duplex monitoring combines onboard sensors with remote camera review to stop safely, reduce false stops, and restart vehicles during link interruptions.

By separating fixed roadside objects from true avoidance targets, the vehicle maintains speed between fields without unnecessary stopping.

Cumulative trust scoring across time sources helps automotive systems update system time accurately while detecting drift, compromise, and attacks.

Tiling and binary masks let sparse CNNs skip empty LIDAR regions, cutting real-time perception latency with minimal accuracy loss.

When onboard driving logic stalls, remote operators use vehicle visual data to confirm decisions and restore safe autonomous operation.

Predicting nearby vehicle responses lets autonomous trajectory planning rescore candidate paths and avoid dynamic-obstacle collisions.

Secondary alerts are triggered only when user input suggests an unintended autonomous disengage, reducing confusion and cognitive load.

Equivalent sensors back up driving assistance and autonomous control while cutting redundant components, space, weight, and power use.

Authenticated server-issued commands are prioritized against in-vehicle inputs using driving conditions to keep vehicle control safe and consistent.

Seat motion is coordinated with driver readiness so autonomous mode does not hand over control until manual driving posture is ready.

Visible light cues and traffic control context help classify stopped vehicles as short- or long-term, enabling safer lane or route changes.

Scenario simulation with updated internal maps helps autonomous vehicles compare candidate actions and improve functional safety in changing environments.

Predicted regional dose monitoring lets a laser interlock cut or attenuate scanning output before exposure exceeds safety thresholds.

A centralized positioning surveillance system uses 5G transceivers, tags, and triangulation to prevent non-line-of-sight collisions in shared spaces.

Blockchain-backed state transition records protect autonomous vehicle data exchange from tampering while limiting shared data to relevant participants.

An archived state-and-trajectory model helps self-driving agents retrace promising paths, avoid deceptive rewards, and explore sparse environments efficiently.

When the driver is away, connected vehicle monitoring uses onboard sensors and wireless alerts to detect interior anomalies and warn the driver.