Automotive
NexaSphere PoC 2: AI-Driven Data Optimization for Automotive Applications
NexaSphere PoC 2 addresses the challenge of data hoarding, which could be exacerbated by the advent of 6G networks, by optimizing the volume and quality of data exchanged within automotive systems. The use case focuses on defining, implementing, and validating AI-driven strategies for efficient data management in safety-critical and non-critical automotive applications.
Key Features & Innovations:
- AI-based Data Optimization: Utilizes advanced AI techniques to determine which data is essential for different automotive use cases, thereby reducing unnecessary data transmission. This includes developing novel dimensionality reduction techniques to optimize the data exchange process.
- Bi-directional Data Exchange: Leverages edge-offloading to perform data aggregation or computation on edge devices, ensuring that only relevant aggregated data is transmitted over cellular and non-terrestrial networks (TNs and NTNs).
- 3D Network Integration: Integrates TN and NTN for seamless geographical coverage, allowing for continuous, uninterrupted data exchange even in remote areas with limited or no 4G/5G coverage.
Safety-Critical & Time-Critical Applications
ADAS & ADFs: Advanced Driver Assistance Systems (ADAS) and Automated Driving Functions (ADFs), which rely on accurate, real-time data from connected vehicles, benefit from the 3D network’s ability to enhance situational awareness. This allows vehicles to operate more autonomously and safely by accessing additional data from nearby vehicles and infrastructure.
- Data Redundancy: Improves the accuracy of onboard sensor data by integrating readings from other connected vehicles and infrastructure.
- Data Extension: Provides additional contextual information beyond the range of onboard sensors, critical for decision-making (e.g., accident ahead or hazardous road conditions).
- Vehicle Automation ODDs: The network enables the exchange of data for strategic decisions, such as rerouting a vehicle to avoid a traffic incident several kilometers ahead.
Time-Non-Critical Applications
- FOTA (Firmware Over-the-Air) Updates: Enables remote software or firmware updates for vehicles without the need for physical access. By leveraging the low-latency 3D network, FOTA updates are delivered faster and more efficiently.
- Predictive Maintenance: Uses machine learning and data analytics to predict and prevent potential vehicle failures, ensuring that vehicles remain operational and reduce downtime.
- EV Range Prediction: Improves the prediction of an electric vehicle’s range by considering factors like battery health, driving behavior, and environmental conditions, ensuring more accurate and timely estimations.
Benefits & Outcomes
- Data Reduction: The project aims to reduce data exchange, storage, and management by at least 30% compared to current strategies, reducing computational resources and energy consumption across vehicles, clouds, and edge nodes.
- Faster Communication: Reduced communication latencies and pervasive 3D connectivity will significantly enhance the effectiveness of automotive applications, including faster delivery of FOTA updates and more responsive predictive maintenance.
- Seamless Connectivity: By ensuring that the 3D network operates effectively across all environments, vehicles can maintain full situational awareness and efficient operation, even in areas traditionally underserved by 4G/5G networks.
