Reliable Robotics and the autonomous cargo plane that keeps the pilot on the ground
Reliable Robotics is retrofitting existing cargo planes for fully autonomous flight, with uncrewed Caravan demonstrations already completed.
Reliable Robotics is pursuing a fundamentally different strategy than most aviation startups: instead of building new aircraft, the Mountain View, California company is retrofitting proven cargo planes with autonomous flight systems to operate with zero pilots on board. They’ve already demonstrated fully uncrewed flights of a Cessna 208 Caravan, and they’re targeting commercial cargo operations within the next few years.
Who Is Reliable Robotics?
Founded in 2017 by Robert Rose, former head of SpaceX’s autopilot program, and Juerg Frefel, who came from the self-driving car industry, Reliable Robotics has raised over $100 million in funding. Their core thesis is deceptively simple: the airframe and propulsion problems are already solved. Thousands of capable cargo aircraft exist. The only unsolved problem is the automation layer.
This narrow problem set—software, sensors, and certification—distinguishes Reliable from electric aviation and eVTOL startups that must simultaneously solve airframe design, propulsion, manufacturing, and autonomy.
What Have They Actually Demonstrated?
In September 2023, Reliable Robotics flew a Cessna 208 Caravan completely uncrewed over an active airport under an FAA experimental certificate. The flight included:
- Autonomous taxi
- Takeoff
- Pattern work
- Approach and landing
- Full coordination with air traffic control
No human touched the controls. This wasn’t conducted on a remote test range—it took place in active airspace with standard ATC communication.
How Does the System Work?
Reliable calls their technology a remotely supervised autonomy platform. The aircraft flies itself through every phase of flight from engine start to shutdown. A ground-based remote operator monitors the flight and can intervene if necessary, but under normal operations functions more like an air traffic controller watching a radar blip than a pilot flying by wire.
The system architecture includes triple-redundant flight computers, independent sensor fusion, and the remote operator as an additional backup layer. This creates a fault-tolerant system designed to demonstrate a safety case equivalent to or better than single-pilot operations.
Why Target Cargo First?
The economics of small cargo operations in the United States make this the obvious entry point. Hundreds of routes—FedEx feeder operations, mail runs to small communities, pharmaceutical deliveries—use single-engine turboprops carrying a few thousand pounds of payload.
The limiting factor isn’t the aircraft. It’s finding pilots willing to fly a Caravan at 3 AM for modest pay to rural strips. The airline hiring wave has siphoned qualified aviators from these operations, making staffing the single biggest constraint.
Removing the pilot from cargo aircraft solves three problems simultaneously:
- Eliminates crew scheduling constraints that limit when flights can operate
- Removes the single largest variable cost in small cargo operations
- Eliminates fatigue risk inherent in early-morning departures and short turnarounds
NTSB data for Part 135 cargo operations shows pilot error as the dominant accident factor—not because these pilots lack skill, but because operational pressures, weather, and odd hours create fatigue. An autonomous system flies the procedure identically every time.
What’s the Certification Strategy?
This is where Reliable’s approach shows strategic sophistication. They’re not asking the FAA to certify a new aircraft category. They’re seeking a supplemental type certificate (STC) on an existing airframe. The Cessna 208 Caravan is already one of the most proven utility aircraft on the planet with thousands in service.
The certification question narrows considerably: you’re not proving the airplane can fly. You’re proving the automation meets the same safety standard as a human pilot.
What Are the Major Challenges?
Regulatory framework gaps. The FAA has no established pathway for certifying a large aircraft with zero crew. The agency has shown willingness to engage with Reliable, but engagement and a signed certificate are vastly different things.
Detect and avoid. Uncrewed aircraft sharing national airspace with manned traffic must demonstrate see-and-avoid capability equivalent to a human pilot. Reliable uses ADS-B, radar, and electro-optical sensors, but proving regulatory equivalence remains an unsolved problem.
Communications reliability. The remote supervision model requires continuous data link between aircraft and ground station. Loss-of-comm scenarios must be handled autonomously, including diverting to a suitable airport without human input. Certifying this contingency logic to required safety standards is non-trivial.
Public acceptance. Cargo eliminates passenger anxiety, but these aircraft share airspace with passenger flights and fly over populated areas. The public trust equation still factors into political and regulatory decisions.
What’s the Timeline?
- 2023: Completed uncrewed Caravan flight demonstrations
- Ongoing: Partnership with the U.S. Department of Defense for logistics missions, providing flight hours, data, and credibility
- 2026–2027: Stated goal for commercial cargo operations (realistically, add 1–2 years for FAA certification realities)
Who Are the Competitors?
Xwing has pursued autonomous cargo using the Cessna Grand Caravan. Merlin Labs worked on autonomous flight before pivoting strategy. On the military side, Shield AI is building autonomous systems for defense applications. Reliable Robotics has the most publicly demonstrated progress on the civil cargo side.
What Does This Mean for Pilots?
Autonomous cargo will not replace pilots flying passengers. The regulatory, psychological, and liability barriers to removing pilots from passenger aircraft are orders of magnitude higher than cargo.
What it will change is the broader aviation ecosystem:
- New job categories will emerge: remote operators, autonomy system maintenance technicians, integration specialists
- Safety statistics should improve for the overall fleet, affecting insurance and regulation downstream
- Routes most likely to go autonomous first are those hardest to staff today: short hops, night operations, thin routes to small airports
Pilots currently flying Part 135 cargo should monitor this technology’s progress as it relates to career planning on those specific route types.
Key Takeaways
- Reliable Robotics has demonstrated fully autonomous, uncrewed flight of a Cessna 208 Caravan over an active airport under FAA experimental certificate
- Their retrofit strategy avoids the need for new airframes, propulsion, or manufacturing—narrowing the problem to software, sensors, and certification
- Small cargo operations face acute pilot staffing challenges that create strong economic incentive for automation
- FAA certification remains the critical unknown, with no established pathway for zero-crew large aircraft operations
- Autonomous cargo affects career planning for Part 135 cargo pilots but poses no near-term threat to passenger operations
Radio Hangar. Aviation talk, built by pilots. Listen live | More articles