Autonomous Vehicle Passenger Detection via Multi-Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately and quickly locating and identifying intended passengers due to the limitations of existing GPS technologies and the inability to recognize passengers through facial images or gestures, leading to delayed ride requests in ridesharing scenarios.
Innovation Solution
The system employs a sensor suite with gesture detection, signal detection, and facial recognition units, leveraging advanced sensing and computational power to analyze video and sensor data, compare it to predetermined gestures and signals, and use pre-negotiated signals or body frame detection to pinpoint and authenticate passengers, thereby reducing pickup time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS technology is used to locate passengers, then the autonomous vehicle can obtain location information, but the precision and accuracy are insufficient to pinpoint exact passenger location
Solution Approach 1:
The patent combines multiple detection technologies (GPS, facial recognition, gesture detection, signal detection) into an integrated passenger detection system. This merging allows the system to overcome the limitations of GPS alone by cross-referencing multiple data sources to pinpoint the exact passenger location with high precision.
Solution Approach 2:
The patent introduces intermediary detection mechanisms such as facial recognition units and gesture detection units that act as mediators between the GPS location data and the final passenger identification. These intermediaries refine the location precision by providing additional verification layers.
2Reliability
If the autonomous vehicle waits for express identification input from the passenger, then it can ensure accurate passenger identification, but the pickup time is significantly delayed
Solution Approach 1:
The system performs preliminary detection actions by continuously monitoring for facial images, gestures, and signals before the passenger needs to provide express identification. This allows the vehicle to pre-identify the intended passenger, reducing the time required at the actual pickup moment while maintaining reliability through multiple verification methods.
Solution Approach 2:
The patent implements feedback mechanisms where the system continuously receives and processes signals from passengers (such as hand waves, facial expressions, or device signals). This real-time feedback allows the system to dynamically adjust its detection and identification process, ensuring accurate identification without requiring lengthy explicit input from the passenger.
3Measurement precision
If the autonomous vehicle uses multiple detection units (gesture, signal, facial recognition), then passenger identification accuracy improves, but the device complexity increases
Solution Approach 1:
The patent designs the detection units to serve multiple functions. For example, the facial recognition unit not only identifies passengers but also detects their presence and orientation. The gesture detection unit can simultaneously track multiple passengers and distinguish between different types of gestures. This multi-functionality reduces the need for separate specialized sensors for each detection type.
Solution Approach 2:
The complex detection system is segmented into modular units (gesture detection unit, signal detection unit, facial recognition unit), each handling a specific aspect of passenger detection. This segmentation allows for independent optimization and maintenance of each unit while working together as an integrated system, making the overall complexity more manageable.
4Speed
If the autonomous vehicle processes video and sensor data in real-time, then passenger detection speed improves, but the computational power required increases
Solution Approach 1:
The system applies partial processing by focusing computational resources on the most relevant data at any given moment. Instead of processing all video and sensor data equally, the system prioritizes processing based on detected cues (such as motion detection triggering facial recognition, or signal detection triggering gesture analysis), reducing overall computational energy consumption while maintaining detection speed.
Data Source
AI summary
A system and method for locating and identifying a potential passenger by an autonomous vehicle includes: receiving a pickup request for the potential passenger, where the pickup request includes a geographic location for locating and picking up the potential passenger by the autonomous vehicle; locating the potential passenger based on the geographic location and using data from one or more sensors of the autonomous vehicle; and in response to locating the potential passenger, controlling the autonomous vehicle to a position to enable pickup of the potential passenger.