At Aurora Flight Sciences, a Boeing Company, the philosophy of “fly early, fly often” is the driving force behind its autonomy development. This principle is institutionalized through the ATLAS program (Accelerated Testing of Live Autonomy Software), a comprehensive infrastructure designed to transition code from development into a certified mission capability for aircraft with maximum efficiency.
Aurora’s strategy hinges on the belief that robust autonomy software requires frequent, mission-level testing that moves beyond pure simulation. The ATLAS framework facilitates a structured progression through three simulation stages—Software-in-the-Loop (SIL), Processor-in-the-Loop (PIL), and Hardware-in-the-Loop (HIL)—before advancing to live flight tests. This process begins with low-cost, low-risk small Unmanned Aircraft Systems (UAS) and scales up to larger, more complex platforms.
This methodical yet rapid approach, often involving monthly flight tests, allows engineers to iteratively increase mission complexity, build system robustness, and reduce technical risk. The gap between a test aircraft and an operational mission platform is systematically narrowed with each cycle.
A key innovation of ATLAS is its ability to blend physical and virtual assets seamlessly. In a test scenario, a low-cost drone can be equipped with a simulated sensor, while additional virtual aircraft are introduced into the mission. The ground control station interface is identical, making the virtual and real elements indistinguishable to the operator. This creates unparalleled flexibility, allowing testers to scale scenarios dramatically while managing cost and risk.
“We can rapidly modify tests and adjust assets with great flexibility,” said Nick LaBarbera, a program manager at Aurora. “This allows us to find the perfect balance between test cost, risk, and real-world relevance, smoothing the path from simulation to operational capability.”
