Autonomous Decision Making Process TRIZ Cases

Coordinated shelf-tote and order-tote carriers enable moving-item picking at a transfer station, improving asynchronous warehouse fulfillment.

A single control unit coordinates machines and unmanned aircraft for inspection, repair, and material transport with less complexity and human intervention.

Ground-truth disengagements and human annotations are compared with driving-log reports to cut evaluator false positives and negatives.

Multiple laser and obstacle sensors let an autonomous forklift locate itself, detect racks and pallets, and unload more safely and accurately.

Multiple sensors are selected by detected surroundings so a mobile object can estimate its position accurately when GPS, odometry, or LiDAR vary in reliability.

Radar-camera fusion data lets AI classify detected objects more stably at night and in bad weather despite radar resolution limits.

Autonomous situation detection lets a remotely controlled vehicle park or exit in forward or reverse without extra user input.

A field robot offsets its EMI sensor from onboard metal and electronics to capture dense soil conductivity data for real-time moisture mapping.

Bounding-box occupancy seeding and convex expansion improve free-space mapping accuracy for robotic path planning in cluttered environments.

When obstacles degrade satellite signals, the control module switches navigation modes and adjusts movement to keep positioning accurate.

Dynamic soft and hard refueling tasks balance vehicle energy levels with service assignments to reduce wasted depot trips and downtime.

Synthetic pre-training and selective layer updates help neural networks reject out-of-domain artifacts while adapting to real-world images.

A POMDP-guided filter architecture detects degraded navigation sensors in near real time and excludes them to preserve vehicle navigation integrity.

A parameterized autonomy architecture separates awareness, planning, consensus, and execution modules to speed vehicle adaptation and integration.

Multiple radar pulses gathered in motion are correlated with roadside map features to localize autonomous vehicles without GPS.

A coordinated autonomous vehicle fleet uses cameras, machine learning, and route recalibration to identify and track suspicious vehicles.

When GPS degrades or fails, internal-sensor self-location and return data help an autonomous moving body navigate back safely.

Multi-sensor mode switching combines GNSS, LiDAR, and kinetic data to keep farm robots on path under crop canopies and recover when stuck.

Only sensor data that triggers model or driver disagreement is sent to the server, cutting storage and network load while preserving edge cases.

Weighted averages from recent manual driving adapt semi-ated acceleration limits, improving stability and efficiency under varying loads.

Dynamic fleet scheduling cuts infeasibility scores in facility transport by adapting routes and task assignments to changing demands.

Uses external camera capture ranges to plan autonomous movement paths that balance monitoring coverage, risk reduction, and travel efficiency.

Dynamic decision-tree hardware and configurable interconnects cut classification latency while preserving retrainable random forest logic.

Risk-based teleoperation lets autonomous vehicles handle low-risk obstructions on their own and request human help only for exceptional cases.

Structured agent, location, motion, distance, and time encodings improve vehicle scenario classification in complex sensor environments.

By predicting fill level along a predefined harvester path, the control system guides haulage vehicles to timely field rendezvous points.

Adjustable wheels expand the support polygon and reduce user kicking, improving stability for seated walking assistance and training.

Electronic data exchange plans, controls, and monitors autonomous farm work so vehicles can operate safely without on-site remote operators.

A swarm assistance network matches autonomous farm machines to requested tasks, reducing reconfiguration effort and improving machine utilization.

Passive metallic dock targets and onboard inductive sensors guide final backing alignment, cutting dock power and improving docking reliability.

Partitioning large SLAM maps into hyper-graph subgraphs cuts optimization load while preserving coverage for autonomous vehicle navigation.

Sensor-built environmental maps and signal checks help UAVs adapt routes in real time for obstacle avoidance and safer navigation.

Onboard computing adjusts vehicle data types and transmission rates by mode or context to improve communication efficiency and reduce resource use.

Historical driving data flags high-risk road segments so autonomous vehicles can request remote manual control before collisions or uncertain maneuvers.

Imaging-based bogie mode detection lets an AGV adjust travel control to prevent wobbling and route deviation while towing through curves.

GPS heading comparison lets automated industrial vehicles detect post-turn path changes and establish a new guidance curve to avoid veering.

Simulated load assignment matches robots or personnel to each transport task using map, capacity, and status data to cut time and energy.

Work-state feedback lets a setting device detect unfinished tasks, assign the right machine, and reduce missed work in autonomous operations.

Sensor-driven cell decomposition lets an automatic mower reroute around unknown obstacles while maintaining coverage of the remaining area.

Route and destination weather data guide UAV medication delivery timing to limit thermal exposure and preserve drug efficacy.

Event-camera data compression and reinforcement-learned control cut latency for agile UAV navigation in cluttered environments.

Balances detachable modules across regions by moving surplus units between autonomous vehicles to cut storage waste and maintain service availability.

Steerable wheels, a U-shaped base, and an extendible forklift let this robotic reachtruck lift pallets in narrow aisles and at multiple rack levels.

Removable battery packs, charging bays, and tilt sensing cut recharge downtime while enabling safer remote lawn mower operation.

Training-time monitoring and property-based datasets help verify ambiguous neural network outputs while reducing test CPU time.

A mobile insect suppression unit uses autonomous movement and environmental control to expand coverage and reduce manual operation.

A mobile insect suppression unit replaces fixed repellents with autonomous movement, expanding lawn and outdoor coverage with less user intervention.

Adaptive selection of GPS, odometry, and 2D/3D scanner data keeps mobile object positioning accurate across indoor and obstructed environments.

By detecting backlighting and changing path or camera angle, the machine avoids recognition errors without heavier image processing.

Real-time path regeneration helps a user-guiding mobile object stay ahead of users who suddenly change direction or move behind obstacles.