Reliable Robotics and the autonomous Cessna Caravan hauling cargo with an empty cockpit

Reliable Robotics has flown a Cessna 208 Caravan with an empty cockpit through the National Airspace System, completing what it calls the first fully autonomous flight of a large cargo aircraft. The Mountain View, California company is not building a new aircraft — it is retrofitting one of aviation’s most proven workhorses to fly itself gate-to-gate, targeting the thin-margin cargo routes that are disappearing as the pilot shortage deepens.

Who Is Behind Reliable Robotics?

Robert Rose and Juerg Frefel founded the company in 2017. Rose came from SpaceX, where he developed the flight software that lands Falcon 9 boosters on drone ships. Frefel came from the self-driving car industry. Their founding question was not “how do we build a new aircraft?” but “how do we make existing aircraft fly themselves?”

That distinction shapes everything about the company’s approach. Reliable is not designing an eVTOL or a novel airframe. It is taking a certified, in-service aircraft and adding autonomy as a retrofit.

What Does Full Autonomy Actually Mean?

The system handles engine start, taxi, takeoff, cruise, approach, landing, taxi back, and shutdown — the complete gate-to-gate cycle with no pilot on board. In late 2024, a Caravan took off from one airport, flew a route, and landed at another with an empty cockpit. A remote operator monitored the flight from the ground but did not touch the controls.

The aircraft is fitted with redundant flight computers, navigation sensors, communication links, and a detect-and-avoid sensor suite. Critically, the software is built to DO-178C standards — the FAA’s framework for certifiable aerospace software. This means deterministic code that can prove what it will do in every situation, not the probabilistic machine-learning approach used in self-driving cars.

Why the Cessna Caravan?

The 208 is the backbone of rural cargo operations across North America. FedEx feeders, small package carriers, and regional operators fly Caravans into strips where the runway is 3,000 feet of cracked asphalt and the nearest instrument approach may be ten miles away.

These operations are losing pilots — not because the flying is difficult, but because the economics cannot compete. A regional airline pays more for a CRJ right-seat than a cargo outfit can offer for single-pilot Caravan runs into small towns at five in the morning. Reliable is targeting exactly this gap: short-haul, thin-margin cargo routes that the pilot pipeline can no longer sustain.

How Does the Remote Operator Model Work?

Think of it as air traffic control in reverse. Instead of one controller managing many aircraft, a single pilot on the ground monitors one — and eventually perhaps several — aircraft. The operator has a full cockpit display, sees what the aircraft sees, and can intervene at any time.

In normal operations, the system flies itself. The human handles edge cases: a thunderstorm on the route, a runway closure mid-flight, any situation requiring judgment rather than execution.

Why Cargo Has a Regulatory Advantage

Cargo does not carry passengers, and that single fact reshapes the entire certification conversation. The FAA has historically been more open to new operational concepts when human lives are not in the cabin. Part 135 cargo operations already permit single-pilot flights in instrument conditions. The regulatory distance from one pilot to zero pilots with a remote operator standing by is genuinely shorter than the leap required to certificate an entirely new aircraft type carrying paying passengers.

Reliable is working with the FAA under a type-specific special condition process, operating within existing Part 23 and Part 135 rules and asking the FAA to define additional requirements for the autonomous system. This pragmatic path avoids the need to lobby for an entirely new regulatory category.

What Are the Real Obstacles?

Trust, not technology, is the biggest barrier. The cargo operators who stand to benefit most are among the most conservative businesses in aviation. They run on razor-thin margins, cannot absorb an accident, and cannot afford the insurance spike that comes with being first adopters. Early customers will likely be larger operators who can handle the upfront retrofit cost and higher initial insurance premiums.

Ground infrastructure presents a second challenge. Autonomous taxi requires a cooperative airport environment — predictable gate positions, ground vehicles that stay clear, and communication with ATC without a human voice in the cockpit. Reliable is developing digital data link protocols for ATC interaction, but most airports lack this infrastructure today.

Weather decision-making is the third hurdle. An experienced cargo pilot makes dozens of judgment calls per week — crosswind assessments, icing evaluations, visibility calls — based on a blend of data and instinct. Replicating that in software requires sensor systems and decision logic well beyond any current autopilot.

When Will This Reach Commercial Operations?

Reliable is targeting initial commercial operations in the 2027–2028 window, starting with specific routes and operators under close FAA oversight. That timeline aligns with the technology they have demonstrated, though aviation certification programs have a long history of running longer than projected.

What This Means for Pilots

Autonomous cargo Caravans do not eliminate pilot jobs — they change them. The remote operator role requires the same aeronautical knowledge, decision-making skills, and certificates as conventional flying. The difference is working at a desk rather than in a cockpit, potentially monitoring multiple flights per shift.

For general aviation, the downstream implications are significant. The sensors, flight computers, and autonomous landing systems being built for cargo will eventually filter into safety systems for piloted aircraft — an incapacitation-recovery autoland for a Caravan or King Air is a logical next step from the technology being flight-tested now.

How Does Reliable Compare to Competitors?

Xwing (now rebranded) has also pursued autonomous Caravan operations. Merlin Labs is working on autonomous flight for military and cargo applications. Reliable distinguishes itself through focus — one aircraft type, one use case — until certification is achieved. In aerospace startups, that discipline typically separates companies that certify products from those that run out of funding.

Key Takeaways

• Reliable Robotics completed a fully autonomous Cessna Caravan cargo flight in late 2024, with no pilot on board and a remote operator monitoring from the ground.
• The company targets short-haul cargo routes losing service due to pilot economics, not passenger air taxi or airline operations.
• Cargo-only operations face a lower regulatory bar than passenger-carrying autonomous flight, giving Reliable a faster path through FAA certification.
• Commercial operations are targeted for 2027–2028, starting with specific routes under close FAA oversight.
• The remote operator role preserves the need for pilot knowledge and certificates, shifting where and how the work is performed rather than eliminating it.