Low-altitude aviation is moving from isolated drone missions toward networked, data-driven operations. The FAA’s proposed Part 108 framework is important because it points toward routine BVLOS flights, more automated aircraft, and a new operating model where safety depends not only on the pilot, but also on software, data services and real-time coordination.
From pilot-centric flight to system-centric flight
In traditional aviation, responsibility is built around a pilot seeing, deciding and acting. BVLOS drone operations change that model. The aircraft may be far from the operator, flying repetitive routes for inspection, delivery, surveying, emergency response or infrastructure monitoring. In this environment, AI becomes useful as a support layer for route planning, anomaly detection, conformance monitoring, fleet prioritisation and decision support.
UTM is the digital traffic layer
UAS Traffic Management, or UTM, can be understood as a collaborative digital ecosystem for managing many drone operations at low altitude. Instead of relying only on voice communication and visual separation, UTM uses shared flight intent, authorisations, surveillance, conflict management and operational data. This is where AI can add value: identifying traffic conflicts early, ranking risk, predicting route congestion and helping operators respond before a small deviation becomes a safety event.
Why Part 108 matters
The proposed FAA Part 108 rule is not just another drone rule. It signals a shift from waiver-based BVLOS approvals toward a more repeatable regulatory structure. It also connects drone operators with automated data service providers, including UTM services. That creates a practical question for the industry: which data is trusted, which services are certified, and how quickly can automated systems detect a risk and recommend a safe response?
AI use cases in a Part 108-style UTM stack
AI can support strategic deconfliction before take-off, tactical alerting during flight, route optimisation around constraints, weather-aware mission planning, battery and maintenance prediction, and post-flight compliance analytics. The strongest systems will probably combine rule-based aviation safety logic with machine-learning models trained on operational data.
The next challenge: trust
The hardest part is not building dashboards. It is proving that automated services are reliable, explainable and resilient. Aviation needs audit trails, cybersecurity, predictable system behaviour and clear responsibility when automation recommends an action. For drone operators, this means AI should not be presented as magic. It should be presented as monitored infrastructure: useful, bounded and continuously validated.
For panka.pl readers, the takeaway is simple: the future of drone aviation will not be defined only by aircraft performance. It will be defined by the quality of the digital airspace layer around the aircraft. FAA Part 108, UTM and AI all point in the same direction — aviation becoming more connected, automated and data-governed.