Air Taxi (Hybrid or Electric) aero Nautical Simulation (ATHENS)

Automated design processes, especially using Machine Learning/AI techniques, require proposed systems to be evaluated across all relevant attributes, requirements, and concerns.  Traditionally, teams create models in a set of engineering tools for design evaluation data. 

This project creates a Design Oracle, where automated designers can submit a system as a graph of components and request a full range of evaluations, composing analysis workflows containing multiple engineering tools across multiple physical domains, fidelity, and scenarios.  Manually created system models are not required.  The system works across cloud architectures, supporting the use of parallel computation.

The design process for Cyber Physical Systems (CPS) spans a wide range of abstractions.  At the most abstract, functionality is mapped to a physical implementation which can then be evaluated.  For the designer, either human or AI, this is where innovation occurs, making key decisions and optimizing physical performance.  Taking these designs, and mapping them to performance estimations, using computer-based simulation and analysis, can be time consuming, complex, and knowledge-intensive due to the variety and difficulty of modern engineering tools.  Isolating and automating the performance estimations allow the AI/human designer to focus on systems and design, without 'polluting' this process with tool specifics.  This is the job of the Design Oracle.

The cost of evaluation matters for AI-based design exploration, as the algorithms learn from many examples.  Prior implementation of the Design Oracle using OpenMETA, while comprehensive in terms of domains and tools supported, encountered significant performance limits, due to the costly model import and conversion steps. 

This project eliminates this cost by operating directly on the AI Symbiotic design models.  Further optimizations are obtained by reusing tool states over multiple design instances, greatly accelerating the evaluation process.  Further tight integration with an enterprise-level open source graph database, workflow manager, and S3 results database promote portability, scalability, and concurrent operations.  

The system leverages technology developed under the DARPA Adaptive Vehicle Make META program, which created a uniform, multi-domain design and component representation, and tools for composition to engineering tool models across multiple tool types, domains, and fidelity levels. 

The Design Oracle has been used within the DARPA Symbiotic Design for Cyber-Physical Systems, across two classes of target designs, Unmanned Air Vehicles and Unmanned Underwater Vehicles.   

Lead PI
Ted Bapty
Jason Scott