DAUST Appoints New Scholars to Advance Aerospace and AI Research
The Diamond-Abrams University of Science and Technology (DAUST) has appointed several international scholars to its faculty. The new hires are intended to bolster the university's research and teaching capabilities in the fields of AI, robotics, and aerospace engineering.
- The use of Model-Based Systems Engineering (MBSE) is growing in aerospace to manage increasing product complexity and shorten development time by using digital models as a primary means of information exchange, rather than document-based approaches. - For high-performance embedded computing in aerospace, designers often choose between FPGAs and GPUs. FPGAs offer lower latency and are more deterministic, while GPUs excel at floating-point operations and are supported by a wider range of open development tools. - The open-source RISC-V instruction set architecture (ISA) is gaining traction in aerospace for its scalability and customizability, allowing for workload-specific processors for applications ranging from low-power sensors to high-performance computing for AI workloads. NASA has selected a processor based on RISC-V for its next-generation space computing project. - Real-Time Operating Systems (RTOS) are essential for safety-critical aerospace software, providing deterministic task scheduling and resource management to ensure that critical operations are executed predictably within strict time constraints. - Edge AI is being increasingly utilized in the aerospace industry to enable onboard data processing, which supports real-time decision-making for applications like autonomous navigation, anomaly detection, and predictive maintenance, thereby reducing reliance on ground systems. - The certification of AI-based software in aerospace systems presents new challenges for compliance with standards like DO-178C, which was traditionally applied to deterministic software. The FAA is actively seeking recommendations for updating software compliance standards to address the unique aspects of AI and machine learning. - Model-Based Systems Engineering (MBSE) is being employed to streamline the design and analysis of complex avionics systems, including for Failure, Detection, Isolation, and Recovery (FDIR) processes, by using models as a central reference for various engineering disciplines. - The development of safety-critical systems often relies on specialized RTOSs like SAFERTOS®, which is designed for high-reliability applications and has been ported to RISC-V-based CPUs implemented on FPGAs.
