Autonomous Decision Making Process TRIZ Cases

Virtual 3D replicas of real geographic areas generate synthetic sensor feeds for autonomous driving training with less risk and faster scenario scaling.

Pre-stop detection lets an elevator robot shift its waiting position before doors open, reducing boarding interference and stop time.

Faster-than-real-time self-simulation lets autonomous agents test action sets online and adapt missions to changing conditions.

Sensor-detected vehicle state changes update a distributed ledger to enforce smart contracts and assign liability objectively.

Consumption data from prior UAV images and video is used to infer preferred flight settings, improving capture quality without manual tuning.

A fixed master GPS receiver broadcasts correction data to mobile units, cutting golf course position errors for navigation and distance measurement.

GPS, roll, and pitch data build 3D lawn maps that improve mowing paths, remote fleet oversight, and utility vehicle scheduling.

A continuous controller guides nonholonomic robots to a final pose without switching, enabling smoother and more precise parking.

Dynamic skip and switch operators let a neural network choose faster execution paths from remaining inference time while preserving useful accuracy.

Multi-source signal keys and causal tracing detect robotic faults in real time, cutting diagnostic latency and processing load for prompt action.

Task selection based on congestion and location conditions helps autonomous warehouse robots cut disruptions, energy waste, and delays.

A two-step TDoA localization approach narrows calculations to anchor sub-units, cutting complexity while improving warehouse collision avoidance.

Real-time path maps mark covered, claimed, and open field tracks so multiple vehicles avoid overlap, missed areas, and collisions.

Autonomous route planning and terrain-based implement control help underground loaders dump safely and consistently in low-visibility mines.

Fleet capability and service-demand evaluation helps choose the right autonomous vehicle mix to cut idle time and computational waste.

Terrain-aware path spacing adjusts overlap on slopes to prevent missed areas and improve full-coverage work by autonomous robots.

Angle sensors and independent screed plate adjustment keep the angle of attack optimal, improving paving quality and reducing machine wear.

Autonomous navigation and a low transom ramp let an unmanned lifeboat reach casualties quickly and recover them without onboard crew.

Two AI models combine visible-object detection with hidden-object inference to help mobile electronics avoid collisions during navigation.

Using timing data from three spatially separated RF antennas, this case improves indoor relative position tracking accuracy with low power use.

A dual-authorization switch gates remote maintenance access to vessel operation systems, reducing unauthorized control risk.

When object perception becomes uncertain, the robot switches from a global path to a local route to avoid conflicts while preserving travel efficiency.

Multi-frame CNN taillight recognition helps autonomous vehicles distinguish brake, turn, and emergency stop signals across vehicle types.

Checkpoint logging traces sensor-to-action execution flow, exposing latency sources and decision data in robotic systems.

Time-multiplexed coherent LiDAR shares photodetector channels to add beams without bulky fiber coupling or higher resource use.

Control-state events are logged to a distributed ledger so smart contracts can automatically enforce transparent liability assignment.

Onboard cameras, LIDAR, and autopilot let a watercraft detect any dock, plan a path, and dock without manual control or dock sensors.

Virtual pillars turn sparse vehicle point clouds into pseudo-images, enabling faster 3D object detection without losing spatial detail.

Location-based virtual measurements filter false sensor readings and improve robot navigation decisions with more consistent environmental maps.

Preselected roadside evaluation units validate vehicle data latency along a planned route, improving transmission reliability for automated driving.

A dual-assurance architecture separates autonomous command generation from certified validation to enable certifiable aircraft control and lower pilot load.

A single detachable sensor maps lawn boundaries and records robot position data, reducing setup complexity while keeping edge guidance precise.

Sensor-triggered ledger transactions automate smart contract liability shifts in autonomous vehicles with transparent, immutable records.

A crossing boundary-wire and charging-loop path guides the mower into precise charging contact alignment with fewer wires and simpler installation.

Expected steering angle ranges from speed and yaw data detect motor deviation and trigger backup steering in autonomous vehicles.

CNN-based parking spot endpoint detection corrects short-length misclassification to improve autonomous parking accuracy in non-standard spaces.

Predefined digging styles and vision-based monitoring let automated work vehicles adapt to changing earthmoving tasks with less operator fatigue.

Selective pruning keeps influential connections for critical driving data, cutting arithmetic load and power use while limiting inference accuracy loss.

Context-aware LiDAR point and pulse dropping relieves channel congestion while preserving timely transmission of useful environmental data.

An IMU-first navigation architecture uses separate filters for GPS and perception data to deliver faster, more reliable vehicle localization.

Onboard flight data trains tail-specific neural models to set cruise speed and cost indices that cut fuel burn and emissions.

A higher-assurance validator checks autonomous flight commands before execution, enabling certifiable aircraft control with lower pilot workload.

Behavior models let a self-driving follower pause, wait, and proceed with human leaders through doors and elevators without obstruction.

Tunnel thrusters draw water beneath a spherical AUV and vector it upward to cut silt disturbance and entanglement in tight wreck spaces.

Radiation-based external scanning measures submerged pipeline density and composition to diagnose plugs faster with less subsea complexity.

Dynamic transport plans and checkpoint interaction let freight vehicles run autonomously, cutting driver fatigue risks and labor costs.

Similarity-based agent grouping shares relevant training data to improve personalized recommendations without losing context-specific accuracy.

Arc-arranged solid-state LIDAR and infrared sensors expand machine vision coverage for power equipment while keeping obstacle detection cost-effective.

Dynamic action reselection lets aircraft automation handle changing mission states without decision trees, reducing pilot intervention.

Inner-side map data creates reference lines for unworked field areas, cutting teaching travel and route calculation time.