Driverless Car Park Guidance Using Road User Motion Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for autonomous driving within parking lots lack efficiency in managing interactions between driverless and manually operated vehicles, as well as pedestrians and animals, leading to potential collisions and inefficient operation.
Innovation Solution
A method and device that detect road users, predict their movements, and automatically guide driverless vehicles within the parking lot based on these predictions, ensuring a predetermined minimum distance is maintained and optimizing traffic flow by adjusting trajectories and timing to avoid conflicts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If driverless vehicles operate autonomously in parking lots without coordination systems, then individual vehicle operation is simple, but collision risk increases and overall efficiency decreases
Solution Approach 1:
The patent merges multiple detection devices (cameras, radar, ultrasonic sensors) into a unified detection system that monitors the entire parking lot environment. This centralized approach allows coordinated control of multiple driverless vehicles, enabling them to share spatial and temporal information to avoid collisions while maintaining individual vehicle operational simplicity.
Solution Approach 2:
The coordination system acts as an intermediary between individual driverless vehicles, receiving data from detection devices and transmitting control commands to vehicles. This mediator architecture enables collision avoidance by calculating safe trajectories and timing, while isolating the complexity from individual vehicles and centralizing it in the coordination system.
2Reliability
If driverless vehicles maintain large safety distances from road users, then collision risk is reduced, but movement efficiency and parking lot utilization decrease
Solution Approach 1:
The system dynamically adjusts safety distances based on real-time detection of road users and predicted movements. Instead of maintaining fixed large safety margins, the coordination system calculates optimal dynamic distances that adapt to the presence, absence, and predicted behavior of pedestrians and other vehicles, thereby maximizing movement efficiency while maintaining adequate safety.
Solution Approach 2:
The detection devices continuously monitor and predict road user movements in advance. By anticipating pedestrian paths and other vehicle trajectories, the coordination system can plan driverless vehicle routes proactively, maintaining minimal safe distances while avoiding actual collisions through提前 prediction and route optimization.
3Productivity
If multiple driverless vehicles operate simultaneously in the parking lot, then overall throughput increases, but trajectory conflicts and coordination complexity increase
Solution Approach 1:
The coordination system segments the parking lot into multiple spatial zones and time slots for different driverless vehicles. By dividing the operational space and time, the system enables simultaneous operation of multiple vehicles without trajectory conflicts, as each vehicle is assigned specific zones to operate in during specific time periods, reducing overall coordination complexity.
Solution Approach 2:
The system implements continuous feedback loops where detection devices monitor all vehicles and road users, the coordination system processes this information to detect potential trajectory conflicts, and control commands are adjusted in real-time. This feedback mechanism enables high throughput by dynamically resolving conflicts as they arise, allowing multiple vehicles to operate simultaneously while maintaining safety.
4Reliability
If the system continuously monitors and predicts all road user movements, then collision avoidance improves, but computational load and processing time increase
Solution Approach 1:
The detection system applies local quality monitoring by focusing computational resources on areas and objects with higher risk potential. Instead of uniformly monitoring all road users with equal detail, the system intensifies monitoring in zones where driverless vehicles are present or approaching, while reducing monitoring intensity in low-risk areas, thereby improving collision prediction accuracy where needed while reducing overall computational energy consumption.
Data Source
Figure 1~2
Figure 3
AI summary
The invention relates to a method for conducting a driverless motor vehicle in a car park, comprising the following steps: - detecting one or more road users present in the car park, - predicting a respective movement of the one or more road users, and - automatically conducting the driverless motor vehicle in the car park on the basis of the respectively predicted movement. The invention further relates to a corresponding device, to a car park, and to a computer program.