FASTER-AI: Autonomous Realization of Embedded AI
FASTER-AI aims to integrate Machine Learning into telecom hardware and safety-critical systems for airborne vehicles, providing a sustainable path towards digitization of critical infrastructure.
FASTER-AI aims to integrate Machine Learning (ML) into telecom hardware and time- and safety-critical tasks for airborne systems and vehicles.
Unique Approach
Unlike current AI development environments tailored for cloud providers, our approach focuses on built-in systems’ needs. We streamline ML integration in three core areas:
- Neural Architecture Search: Identifying suitable neural architectures for domain-specific hardware
- Cross-Compilation: Providing a multi-stage cross-compiler for combining traditional logic & ML
- Application-Tailored Software: Offering software tailored to ML-driven applications’ requirements
Future-Proof Design
Our method is efficient for existing hardware and adaptable to future architectures and releases, ensuring accurate and time-critical decisions in various industries.
Impact
We are convinced that the FASTER-AI method is the most sustainable and future-proof path towards digitization and value creation for existing critical infrastructure.
