Merlin Advances Autonomous Flight Software Toward New Zealand Certification
Aviation technology developer Merlin has successfully cleared the third Stage of Involvement review with New Zealand regulators for its autonomous flight system. This milestone specifically evaluates the software verification for both the platform's flight control computer and its artificial intelligence-driven automated communication system, which can interpret and respond to air traffic control. The process is being conducted alongside the FAA through a bilateral safety agreement. For working pilots, this progress signals a rapidly approaching reality where fully autonomous Part 23 aircraft share the airspace and communicate seamlessly with standard ATC facilities.
Read the full take
The push for fully autonomous flight just took a significant step forward with Merlin's latest software verification milestone. While much of the industry's focus on automation revolves around pilot assistance tools, Merlin is actively pursuing a Part 23 certification designed to manage an aircraft from the takeoff roll all the way through touchdown. The technology relies on a flight control computer working in tandem with a machine-learning communication system. This communication component is particularly noteworthy because it uses natural language processing to understand verbal air traffic control instructions regarding altitude, heading, and airspeed, and then physically translates those clearances into flight control inputs. New Zealand's Civil Aviation Authority is leading the certification effort, but the FAA is closely involved through a bilateral safety agreement, meaning this framework is being built with eventual United States airspace integration in mind. Reaching the third Stage of Involvement confirms that the company's software testing aligns with the strict regulatory requirements established earlier in the development cycle. Additionally, the regulators and developers have officially agreed on exactly how the artificial intelligence elements will be evaluated, closing a major issue paper that could have stalled progress. Although final certification remains on the horizon, resolving how to legally certify machine learning in critical flight systems is a massive breakthrough. For student pilots and certified flight instructors monitoring the industry, this development highlights a shift in how future aircraft might operate. Instead of simply building better autopilots, engineers are effectively creating synthetic crew members capable of handling complex radio communications and executing clearances without human intervention.