Vehicle Sensor Fusion Control for Real-Time Obstacle Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving technologies struggle to achieve Level 6 autonomy, which requires ultra-high performance in real-time decision-making and obstacle avoidance with high accuracy, due to limitations in computational power and sensor data processing.
Innovation Solution
An information processing apparatus utilizing a central brain with advanced sensors and deep learning to process sensor data at nanosecond intervals, calculating control variables for wheel speed, inclination, and suspension adjustments, enabling precise obstacle avoidance and adaptive driving strategies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor data is processed at ultra-high speed (nanosecond intervals) to achieve Level 6 autonomous driving, then real-time decision-making accuracy is improved, but computational power requirements increase 1000 times compared to Level 5
Solution Approach 1:
The patent segments the autonomous driving system into multiple specialized processing units: acquisition units for sensor data collection, calculation units for multivariate analysis and deep learning processing, control units for vehicle behavior control, and transmission units for inter-vehicle communication. This segmentation allows parallel processing of different data streams at nanosecond intervals, achieving Level 6 accuracy without requiring a single monolithic computational system to handle all tasks simultaneously.
Solution Approach 2:
The patent implements preliminary action by continuously acquiring and pre-processing sensor data before obstacles require immediate response. The system maintains a continuous stream of processed information about surrounding vehicles, road conditions, and potential hazards, so that when emergency avoidance is needed, the calculation unit can immediately generate control variables without delayed computation. This pre-positioning of processed information enables ultra-fast response times.
2Measurement precision
If multiple sensors detect situation around the vehicle at second cycle shorter than first cycle, then obstacle detection accuracy is improved, but device complexity increases
Solution Approach 1:
The patent applies universality by designing the acquisition unit to handle multiple sensor types (cameras, LIDAR, radar, ultrasonic sensors) through a unified interface and processing pipeline. The same acquisition unit acquires data from different sensors at different cycles, and the calculation unit processes all this diverse information through multivariate analysis and deep learning models, eliminating the need for separate processing systems for each sensor type and reducing overall device complexity.
Solution Approach 2:
The patent merges multiple sensor data streams and processing functions into integrated units. The acquisition unit combines data collection from various sensors, the calculation unit combines multivariate analysis with deep learning processing, and the control unit combines multiple control variables into unified vehicle commands. This merging approach reduces the number of separate components needed while maintaining high detection accuracy.
3Reliability
If control variables are calculated and transmitted to other vehicles for emergency avoidance, then collision prevention capability is improved, but information transmission requirements increase
Solution Approach 1:
The patent extracts only the essential information needed for emergency avoidance and transmits it to other vehicles. Instead of transmitting complete sensor data or processed images, the transmission unit sends specifically extracted control variables (such as recommended acceleration, steering angle, or braking commands) and emergency avoidance signals. This extraction approach enables other vehicles to immediately implement safety measures without receiving or processing large volumes of raw data.
Solution Approach 2:
The patent applies local quality by transmitting different types of information at different quality levels to different vehicles based on their specific situations. The transmission unit sends detailed control variables to vehicles that are at high risk of collision, while sending simpler emergency signals to vehicles at lower risk. This localized information quality optimization reduces overall transmission volume while maintaining collision prevention capability where it is most needed.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing apparatus includes: an acquisition unit that acquires plural pieces of information related to a vehicle from a detection unit including a sensor that detects a situation around the vehicle including an obstacle at a second cycle shorter than a first cycle at which surroundings of the vehicle are imaged, as a cycle for detecting the situation around the vehicle; a calculation unit that includes a setting unit that sets a traveling strategy for the vehicle to avoid the obstacle based on the plural pieces of acquired information, and calculates a control variable for controlling behavior of the vehicle based on the plural pieces of acquired information and the set traveling strategy; and a control unit that controls the behavior of the vehicle based on the calculated control variable.