Autonomous Decision Making Process TRIZ Cases

Roadside sensors and V2X links fill blind spots in ports and warehouses, giving autonomous vehicles more complete low-latency environment data.

Waveform peaks and heuristics let autonomous vehicles remove LIDAR returns from exhaust, dust, rain, snow, and fog.

GNSS and onboard sensors replace boundary wires, calibrating mower direction and speed for precise mowing in irregular areas.

Stored driving snapshots let tele-operators validate exception handling faster, cutting intervention time and support demand.

Doppler radar gust detection guides HAPS relay flight control to stabilize attitude, altitude, and route under strong upper-air winds.

Precomputed semantic layers let UAS combine object-specific configuration spaces to find safe routes without heavy 3D image processing.

When yielding stalls too long outside normal traffic causes, the vehicle sends sensor and map context to a teleoperator to restore progress.

Derived signals and causal traces let robotic systems isolate fault sources across many parameters with low-latency runtime diagnostics.

Headway-based control lets connected vehicles indirectly guide human-driven cars, stabilizing mixed platoons and easing congestion.

By decoupling nonlinear thrust-vector dynamics into independent control variables, this case improves multicopter stability and angle control.

Bin transceivers and BLE-tagged items let autonomous vehicles confirm object placement, manage storage, and trigger retrieval actions.

Parallel collision checking on a reconfigurable processor cuts motion planning load and supports fast replanning around dynamic obstacles.

Automated status logs and incident timing replace human observation, enabling detailed fleet performance reporting for self-driving industrial vehicles.

Salient spectrogram patches and clustered GAP features improve audio classification under noise and mixed sounds while adding explainability.

A wearable crossing request signal lets autonomous vehicles confirm pedestrian intent, adjust motion, and provide clear crossing feedback.

By comparing motor input with actual movement, this case detects wheel slip in articulated robotic tools and reduces surface wear.

Pairs of vehicle trajectories are encoded and clustered with unsupervised learning to classify encounter scenarios for autonomous maneuvers.

Semantic map regions are decimated at different levels to cut memory and compute use while preserving localization accuracy in critical driving areas.

Triggered display changes add pilot-focused flight parameters during autonomy handover, improving awareness and smoother human intervention.

A learned reference contour lets vehicle sensors adapt protected-area monitoring to boundary changes while avoiding false safety reactions.

A charging station stores cloud-sent scheduling commands and relays them to the robot later, enabling remote updates without robot availability.

A dedicated memory device detects adverse events and saves selected volatile vehicle data to non-volatile storage before power loss.

Real-time slamming and roll sensing lets an autonomous vessel adjust speed or tack its course to reduce extreme motion and damage risk.

A candidate region based on vehicle position and heading selects a suitable work route before autonomous travel, reducing startup delay.

Weighted heading and distance error control helps autonomous machines stay on guidance paths with GNSS-based correction and no operator input.

Adaptive CNN thresholds use radar, sensor, and platform-state data to improve weather threat detection under changing altitude and temperature.

By confirming consistent boundary sensor outputs over time, the controller filters interference and stabilizes mower position judgment near the wire.

Image or LiDAR sensor data builds a learned elevation mask to reject blocked GNSS satellites and improve positioning in urban canyons.

Multiple radar pulses processed in SAR mode are correlated with map objects to localize autonomous vehicles when GPS is unreliable.

Confidence-score heuristics flag uncertain autonomous vehicle sensor frames for annotation, cutting data storage and annotation effort.

Double-row RFID tags with variable spacing improve industrial vehicle position and direction detection near aisle entry and exit zones.

Partitioned LIDAR height checks detect miscalibrated vehicle sensors accurately while reducing processing time, memory use, and network load.

Low-confidence and ambiguous perception outputs trigger annotation of only the most useful autonomous vehicle sensor data, reducing labeling cost.

Structured schema-based encodings turn unstructured vehicle environment data into reliable scenario classification and hazard recognition.

A normalized controller framework decouples autonomous applications from external system APIs, easing deployment and maintenance.

An upward-facing RFID reader and built-in printer reduce tag misreads and speed mobile packing and shipping workflows.

Pre-generated diverse perturbations speed adversarial training while improving classifier robustness against image and audio attacks.

A retractable harness with rotating arms and electromagnets enables UAV docking, launch, retrieval, and charging from moving autonomous vehicles.

GNSS virtual boundaries replace wire layouts, letting intelligent lawn mowers mow irregular areas with higher path precision and less maintenance.

Satellite positioning replaces slip-prone wheel rotation to calculate actual travel time and improve ground work progress management.

Coordinated status signals let robotic refuse containers swap positions before bins fill or batteries deplete, improving continuous operation.

Roadside sensing, prediction, planning, and control shift automated driving functions off the vehicle to cut onboard complexity and improve stability.

A ground-based movement system lets an autonomous aircraft approach, grasp, and secure objects precisely in unimproved environments.

Variable guide line width lets a line-following vehicle trigger preset actions at route positions without dedicated markers or extra sensors.

A programmable logic state machine shifts servo control to a backup autopilot, preserving UAV control without bulky backplane redundancy.

Using three spaced receiver antennas and PDOA timing, this case improves indoor relative tracking accuracy with low power and less recalibration.

Opposing fisheye views with peripheral overlap let a UAV measure parallax disparity for more precise object distance, speed, and flight control.

A learned verifier checks whether LIDAR ICP alignments are trustworthy before object tracking and trajectory planning use them.

Voxel-based checks detect duplicate surfaces, holes, and uneven regions in environment meshes, improving map accuracy and vehicle navigation.

Filtered vehicle data and comparison conditions expose sensor, steering, trajectory, and battery abnormalities before clear failure.