IoT Data Fusion for Autonomous Vehicle Deep Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing IoT solutions face challenges such as large numbers of connections, security, and new standards in 5G networks, particularly in enhancing deep learning algorithms for autonomous vehicles, where determining the prioritized hierarchy of data detection, collection, assessment, and decision-making is crucial for smart environments.
Innovation Solution
An IoT network utilizing distributed IoT devices for data monitoring and processing, combined with a navigation and protection system (NPS) that integrates information from moving and stationary objects, employing deep learning algorithms to enhance autonomous vehicle navigation and protection, with synchronization techniques like IEEE1588 PTP and GPS for precise time coordination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If distributed IoT devices are used for data monitoring and processing in autonomous vehicles, then the navigation and protection capabilities are enhanced, but the device complexity and number of connections increase
Solution Approach 1:
The system segments the autonomous vehicle into multiple functional units (sensing units, processing units, actuation units) that can independently monitor and process data from specific IoT devices. This segmentation reduces the complexity of managing all connections centrally while maintaining comprehensive monitoring coverage through distributed functional modules.
2Reliability
If synchronized data from multiple IoT devices is integrated for decision-making, then the accuracy and reliability of autonomous navigation improve, but the time required for data collection and processing increases
Solution Approach 1:
The system performs preliminary actions by pre-synchronizing clocks of IoT devices using IEEE 1588 PTP and GPS before data collection begins. This preliminary time synchronization eliminates the need for real-time time calibration during data processing, reducing overall processing time while maintaining the accuracy benefits of synchronized multi-device data integration.
3Loss of information
If deep learning algorithms are enhanced with data from moving and stationary IoT objects, then the autonomous vehicle's environmental awareness improves, but the computational load and energy consumption increase
Solution Approach 1:
The system extracts and processes only the most relevant features from the vast amount of IoT device data before feeding them to deep learning algorithms. By taking out and focusing on critical environmental information (such as obstacle detection, traffic signal status, pedestrian presence) while filtering out redundant data, the system maintains comprehensive environmental awareness while significantly reducing computational load and energy consumption.
Data Source
AI summary
This application discloses use of moving and stationary IoT objects to enhance deep learning algorithms used for autonomous vehicles. The autonomous vehicle acts as an IoT device and exchange information data with moving or stationary IoT devices in its vicinity. The moving and stationary objects share their information data using their broadcast, Ethernet, or proprietary packets with the autonomous vehicle through its IoT device. The shared information data is used by autonomous vehicle navigation and protection system (NPS) where the deep learning algorithm resides. The shared information data includes specification, video and images of the stationary device and moving object. When only selected stationary devices are active, then the active stationary device broadcast the information data that belongs to stationary devices in its vicinity along the road and freeway that are not active.