Reliable Robotics and the autonomous Cessna Caravan that wants to fly cargo without a pilot on board

Reliable Robotics is developing an autonomous version of the Cessna 208B Grand Caravan, targeting the middle-mile cargo market with a supplemental type certificate (STC) approach rather than building a new airframe. Backed by direct investment from FedEx, the Mountain View, California company has already demonstrated fully autonomous flights — including taxi, takeoff, and landing with no pilot on board — and is working with the FAA to establish the certification framework for large fixed-wing autonomous cargo operations.

Why Retrofit an Existing Airplane Instead of Building New?

Most companies pursuing autonomous flight are designing entirely new airframes. Reliable Robotics took the opposite approach, and the reasoning is strategic.

The Cessna Caravan has decades of service data. The FAA knows the airplane thoroughly. Maintenance crews are already trained on it. Operators already own fleets.

By pursuing an STC on a Part 23 certified airframe, Reliable is asking the FAA to certify only the autonomy system on a known aircraft — not a new vehicle and a new autonomy system simultaneously. That is a significantly narrower certification problem, and in aviation, narrower certification problems get solved faster.

What Has Reliable Robotics Actually Demonstrated?

The company has flown the Caravan fully autonomously with no pilot on board, including taxi, takeoff, pattern work, approach, and landing. These flights were conducted at a test range under FAA oversight and have been repeated multiple times.

Reliable has also demonstrated remote piloting, where a ground-based operator monitors the aircraft through a ground control station and can intervene, issue commands, or trigger a go-around at any time.

The system architecture uses dissimilar redundancy — backup systems that are architecturally different from primary systems so a single software bug or hardware failure cannot cascade through both channels. This is the same design philosophy used in fly-by-wire systems on commercial airliners.

Communications between the aircraft and ground station run through multiple independent channels, including satellite. If all communication is lost, the aircraft executes a pre-programmed safe landing at a designated divert field.

Who Is the Remote Pilot and Who Is in Command?

In Reliable’s model, a human is always in the loop. The concept does not eliminate the pilot — it relocates the pilot to the ground.

Instead of one pilot per airplane, a single remote operator could monitor multiple aircraft. That operator has full authority to take control at any point. This answers the immediate command-authority question while addressing the economic constraints of single-pilot cargo operations.

Why Cargo First — and Why FedEx?

The initial target is the middle-mile cargo market: Caravans flying feeder routes between small regional airports and major sorting hubs. These are often single-pilot operations flown at night, in weather, into austere airports.

The pilot shortage hits this segment hard. Cargo feeder operators compete for pilots against regionals, majors, and corporate flight departments — and often lose on pay and lifestyle. The result is Caravans sitting on ramps because operators cannot staff the routes.

An autonomous Caravan could fly more hours per day, requires no crew rest, and cannot be hired away by a major airline.

FedEx has invested directly in Reliable Robotics. FedEx operates one of the largest Caravan feeder networks in the world. When the operator who would actually use the technology puts capital into the company building it, that carries weight beyond a typical venture investment.

What Are the Biggest Obstacles?

Certification timeline. The FAA has never certified a large, fixed-wing autonomous cargo aircraft for routine operations in the national airspace. No established certification pathway exists. Reliable is working with the FAA to build the framework in parallel with development. The regulatory timeline is the single largest variable in the program.

Detect and avoid. This remains the hardest unsolved problem in autonomous aviation. An autonomous system needs sensors and algorithms that reliably identify uncooperative traffic in all conditions — night, haze, rain, ground clutter. The certification standard for detect-and-avoid performance is still being defined by the FAA and industry groups like RTCA.

Weather judgment. Human pilots make dozens of subjective calls per flight about ice, turbulence, visibility, and wind shear. The Caravan operates in some of the most demanding weather environments in general aviation — Alaskan bush strips, Midwestern icing, Gulf Coast convective weather. Encoding that judgment into software is extraordinarily difficult.

Public and pilot acceptance. Even in a cargo role, an autonomous aircraft shares airspace, runways, and traffic patterns with piloted aircraft. How pilots and airport communities respond to sharing approaches with an unpiloted airplane is a political and social question, not an engineering one.

How Does Reliable Compare to Competitors?

Xwing (now Daedalus) has pursued autonomous Caravan operations but has had a less consistent trajectory. Merlin Labs was working on autonomous flight for military and cargo applications before pivoting.

Reliable Robotics appears to have the most consistent funding, the most visible FAA engagement, and the clearest commercial customer in FedEx. CEO Robert Rose, formerly of SpaceX, has shaped an engineering culture focused on robust systems, relentless testing, and a specific achievable mission.

Realistic Timeline for Autonomous Cargo Flights

Expecting autonomous Caravans hauling overnight packages by 2028 is likely optimistic. A more realistic timeline for routine commercial operations is 2030 or beyond, driven primarily by the pace of FAA certification rather than technology readiness.

What This Means for Pilots Long Term

Autonomous cargo is widely viewed as the first domino. Once the FAA certifies an autonomous system for cargo on a Caravan, the conversation about autonomous passenger operations accelerates significantly. The technology that flies boxes today is the technology that flies people tomorrow.

Commercial aviation will have two pilots in the cockpit for a long time. But the role will evolve. For pilots early in their careers, understanding autonomous systems — how they work, how they fail, and how to supervise them — is becoming a valuable and differentiating skill set.

Key Takeaways

• Reliable Robotics is retrofitting the Cessna 208B Grand Caravan with autonomous capability via an STC, avoiding the risk of certifying both a new airframe and new autonomy system simultaneously.
• FedEx is a direct investor, signaling serious commercial intent for middle-mile cargo feeder routes.
• Fully autonomous flights with no pilot on board have been demonstrated repeatedly under FAA oversight, including taxi, takeoff, and landing.
• FAA certification pathway and detect-and-avoid standards remain the biggest unknowns, making a 2030-or-later timeline more realistic than near-term expectations.
• Autonomous cargo is likely the precursor to broader autonomous passenger operations, making familiarity with these systems increasingly important for working pilots.