System-Level Safety Engineering of Unmanned Aircraft Operations: Incident-Derived Requirements for Control, Feedback, and Recovery

Why It Matters

By exposing how control and feedback failures drive accidents, the study informs regulators and manufacturers about critical safety gaps, prompting stronger standards for beyond‑visual‑line‑of‑sight and people‑centric UAS missions.

Key Takeaways

• Study analyzes 147 UAS incidents across official and voluntary reports.
• Severe outcomes linked to control authority loss and feedback gaps.
• Framework translates incidents into system safety requirements for UAS.
• Highlights need for robust recovery protocols in beyond‑visual‑line‑of‑sight flights.
• Provides reproducible method for incident‑derived safety engineering.

Pulse Analysis

The commercial and public‑safety sectors are rapidly expanding the use of unmanned aircraft systems, pushing them into controlled airspace, densely populated areas, and beyond‑visual‑line‑of‑sight missions. Traditional safety metrics that focus solely on aircraft reliability or pilot skill are insufficient for this new operating envelope. Stakeholders need a holistic view that captures how system‑level interactions—such as control authority, feedback loops, and recovery pathways—affect overall risk. This shift mirrors broader trends in aerospace where complex autonomy demands rigorous safety engineering.

To address the data gap, the authors assembled a cross‑case dataset of 147 UAS occurrences, drawing from official investigation reports and NASA’s voluntary safety reporting system. Each case was coded for variables like unsafe control actions, process‑model flaws, and degraded assurance scores. The analysis uncovered a strong correlation between severe incidents and degraded control authority, feedback loss, and inadequate containment strategies. Notably, technical failures alone accounted for a minority of high‑severity outcomes, underscoring the importance of robust feedback mechanisms and clear recovery protocols in autonomous flight.

The study’s reproducible framework offers a practical pathway for regulators, manufacturers, and operators to translate incident evidence into actionable safety requirements. By quantifying control and feedback deficiencies, the model supports the development of standards for UAS certification, especially for beyond‑visual‑line‑of‑sight and urban operations. Adoption of these incident‑derived requirements could reduce accident rates, accelerate public acceptance, and guide future research into resilient autonomous control architectures. As the UAS market is projected to exceed $100 billion in the next decade, integrating such system‑level safety engineering will be pivotal for sustainable growth.