Merlin Labs and the autonomous flight software retrofitting into aircraft that already exist

Merlin Labs, a Boston-based company, is taking a fundamentally different approach to autonomous aviation. Instead of designing new aircraft, Merlin has built a software and hardware system that installs into existing airframes—like the Beechcraft King Air—and gives them full autonomous capability from takeoff to landing. The company has already demonstrated a completely unmanned King Air flight in the National Airspace System, and secured a U.S. Air Force contract to develop autonomous capability for large military cargo aircraft.

What Is the Merlin Pilot System?

The system Merlin calls the Merlin Pilot is not a traditional autopilot. A conventional autopilot is a control system: it holds a heading, maintains an altitude, follows a glideslope. When something unexpected happens, it disconnects and hands the problem back to the human.

Merlin’s system operates in a fundamentally different category. It is a decision-making system that uses machine perception, sensor fusion, and what the company describes as a reasoning engine capable of handling novel situations. It can assess weather, reroute around traffic, manage emergencies, and handle the full scope of flight operations from engine start to shutdown—not just pre-programmed GPS routes.

Why Retrofit Existing Aircraft Instead of Building New Ones?

Most autonomous flight companies are designing clean-sheet aircraft with new airframes and new propulsion. Merlin’s argument is simpler: tens of thousands of aircraft are already flying today with decades of service life remaining. Making them autonomous is faster and cheaper than replacing them.

The certification math reinforces this strategy. Building a brand-new aircraft requires a type certificate—a years-long, billion-dollar process. Merlin is pursuing a supplemental type certificate (STC) for their autonomous system. The airframe is already proven. The engines are proven. The flight characteristics are known. Merlin only needs to certify the brain, not the body.

What Does the Military Contract Involve?

The U.S. Air Force operates hundreds of cargo aircraft—C-130s, KC-135s—with airframes that have decades of life left but growing difficulty finding crews to fly them. Pilot retention is a chronic problem across every military branch.

Merlin signed a contract through the AFWERX program, the Air Force’s technology accelerator, to develop autonomous capability for large fixed-wing military aircraft. The prospect: taking a Hercules, an airplane flying since the 1950s, and enabling it to complete missions without a crew on board.

What Are the Biggest Obstacles?

Regulatory gaps. The FAA has no established pathway to certify a fully autonomous system for passenger-carrying operations. 14 CFR Parts 91, 121, and 135 all assume a certificated human pilot in command making real-time decisions. Merlin can fly the airplane, but getting the FAA to authorize autonomous passenger flights is a separate problem that will take years to resolve.

Sensor limitations. A human pilot processes enormous amounts of visual information instantly—cloud formations, terrain, traffic, birds. Merlin’s system uses cameras, radar, and other sensors to replicate that capability, but questions about sensor redundancy and failure modes remain unresolved across the industry. A human pilot can fly by feel when instruments degrade. Whether an algorithm can match that adaptability is still being tested.

Public trust. The technology may arrive before society accepts it. Self-driving cars demonstrated technical capability years before the public was comfortable using them. That psychological barrier intensifies at 30,000 feet. Passengers want to know someone in the cockpit is personally invested in the outcome—and that concern is not irrational.

Where Does Merlin Fit in the Market Right Now?

The near-term applications are military and cargo. Unmanned cargo flights do not require passenger acceptance—they require the FAA and Department of Defense to agree on operational standards, and the military has far more flexibility to test autonomous systems than the civil side.

The cargo airline market is the natural bridge to civil operations. Single-pilot Caravans flying canceled checks and freight at 3 a.m. are hard to staff. The pilot shortage hits cargo operators harder than airlines because pay and lifestyle are less competitive. An autonomous Caravan flying overnight freight from Louisville to regional hubs is a realistic near-term use case that sidesteps the passenger trust problem entirely.

What Is the Realistic Timeline?

Full autonomous passenger flight in civil airspace is not happening in the next five years, and probably not in the next ten. The technology may be ready, but regulatory and social infrastructure will lag—just as it has with every transformational aviation technology from pressurized cabins to fly-by-wire.

The more likely path is graduated autonomy:

Unmanned cargo in military and restricted airspace
Autonomous cargo in civil airspace under special conditions
Reduced-crew operations (one pilot plus the autonomous system)
Fully autonomous passenger flight, potentially decades away

Why Should Airline Pilots Pay Attention?

That third step—reduced-crew operations—is already an active regulatory discussion. Both EASA and the FAA are studying single-pilot operations for commercial aircraft. Airbus has been conducting extended minimum-crew operations research for years. Merlin’s system, or something like it, could be the technology that makes single-pilot airliners feasible.

What Does This Mean for General Aviation?

For GA pilots, the implications are less about replacement and more about safety augmentation. A system sophisticated enough to intervene when it detects a developing stall, recognizes spatial disorientation, or takes over if the pilot becomes incapacitated goes well beyond what current autopilots offer.

Garmin’s Autoland is a first step in that direction, executing a single emergency procedure. Merlin’s vision extends further: a system that flies the airplane the way a competent pilot would across a full range of scenarios.

How Does the Role of the Pilot Evolve?

Autonomy does not eliminate pilots—it changes what pilots do. Glass cockpits did not remove the need for humans in the cockpit. FMS, autopilot, and autothrottle shifted the job from hand-flying to systems management. Autonomous capability is the next step on that same continuum.

Merlin Labs has raised over $100 million in funding, assembled a team from Lockheed, Boeing, and defense research labs, and has both flying hardware and government customers. Their bet is that the fastest path to autonomous flight runs through the aircraft we already have, not the ones we have not built yet.

Key Takeaways

Merlin Labs retrofits existing aircraft with autonomous flight capability rather than designing new airframes, pursuing an STC instead of a full type certificate
The King Air has already flown autonomously in the National Airspace System with no safety pilot on the controls
Military cargo and commercial freight are the realistic near-term applications; autonomous passenger flight is likely decades away
Reduced-crew operations (one pilot plus autonomous system) are under active regulatory study by both the FAA and EASA
Regulatory frameworks, sensor reliability, and public trust remain the three major barriers—not the underlying technology