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

Reliable Robotics, a Mountain View, California company, has flown a Cessna 208 Caravan through every phase of flight—taxi, takeoff, cruise, approach, landing, and taxi to parking—with no one on board. This is not a concept video or a simulation. The airplane flew itself, and a human pilot monitored the entire flight from a ground control station. The company is now working toward FAA certification to operate autonomous cargo flights commercially, targeting one of aviation’s most persistent problems: the Part 135 cargo pilot shortage.

Who Is Behind Reliable Robotics?

The company was founded in 2017 by Robert Rose, a former SpaceX engineer who led development of the autonomous flight termination system on the Falcon 9, and Juerg Frefel, who came from Tesla’s Autopilot team. Both have deep experience building safety-critical autonomous systems in industries where failure is catastrophic. This is not a startup running on pitch decks—they have hardware flying.

Why Retrofit an Existing Airplane Instead of Building a New One?

This is where Reliable Robotics diverges from most autonomy companies. Rather than designing a novel airframe with exotic propulsion or lift systems, they chose to retrofit the Cessna 208 Caravan, one of the most widely operated single-engine turboprops in the world. FedEx flies hundreds of them. Regional cargo operators depend on them daily. The Caravan is the pickup truck of air cargo.

By building autonomy into an aircraft with a massive service history, the company sidesteps the certification nightmare of a novel aircraft type. They are not asking the FAA to approve a new airplane. They are asking the FAA to approve a new way of flying an airplane the FAA already knows thoroughly.

The parallel is the glass cockpit revolution. The Garmin G1000 did not change the Cessna 172. It changed what the Cessna 172 could do. Reliable Robotics is applying that same logic to autonomy.

How Does the Autonomous System Work?

The Caravan is equipped with redundant computers, a full sensor suite, and actuators controlling every flight surface, the throttle, brakes, and nosewheel steering. The system handles the complete flight envelope from engine start to engine shutdown.

The critical design element: a human pilot is always in the loop, just not in the cockpit. A remote pilot at a ground control station monitors the flight through a data link and can intervene at any point. If the data link is lost, the airplane follows pre-programmed contingency procedures—holding, diverting, or continuing to its destination depending on flight phase and failure type. Multiple layers of redundancy and fallback modes are built into the architecture.

Why Cargo Aviation Desperately Needs This

The pilot shortage hits Part 135 cargo operations harder than almost any other segment. Operators flying Caravans, King Airs, and Metroliners into small airports at 2 a.m. with medical supplies and overnight packages cannot find enough pilots. The lifestyle is brutal: single-pilot operations, night flying, uncontrolled fields, marginal weather. Pilots build time and leave for the regionals as fast as they can.

The safety data reinforces the urgency. Single-pilot cargo flying at night has historically been one of the most dangerous segments of commercial aviation. Controlled flight into terrain, spatial disorientation, and fatigue-related errors dominate the accident record. These are human factors problems. An autonomous system does not get tired at 3 a.m., does not become disoriented in clouds, and does not press into deteriorating weather to get home.

The risk calculus is also fundamentally different from passenger operations. On these flights, the only life at stake in the cockpit is nobody’s.

What Are the Obstacles?

Certification remains the biggest hurdle. Reliable Robotics has been working with the FAA and pursuing a supplemental type certificate, but certifying an autonomous system to operate in the national airspace system alongside manned aircraft is unprecedented at this scale. The FAA is building the regulatory framework as it goes.

Public perception is a real factor. “Pilotless airplane” triggers worst-case thinking, even when the aircraft carries only cargo. It shares airspace with passenger flights, and public concern will influence the regulatory timeline.

Infrastructure scaling presents challenges. Ground control stations require reliable, high-bandwidth, low-latency communication links. That works in a test environment but scaling across hundreds of routes—many in rural areas with marginal connectivity—is a significant engineering problem.

The economics must close. Retrofitting a Caravan with sensors, computers, actuators, redundancy systems, and ground station infrastructure is expensive. The per-flight cost of the autonomous system has to beat the fully loaded cost of a human pilot—salary, training, benefits, scheduling, crew rest, and attrition. The math likely works, but it has not yet been proven at scale.

Who Else Is Working on Autonomous Cargo?

Xwing, also Bay Area-based, has pursued autonomous cargo operations with the Caravan. Merlin Labs, before its acquisition, focused on autonomy for military and cargo applications. The military is further ahead—the Air Force has flown autonomous Black Hawks and modified cargo aircraft through programs like ALIAS (Aircrew Labor In-Cockpit Automation System). The technology is proven. The civilian regulatory framework has not caught up.

When Will Autonomous Cargo Flights Operate Commercially?

The optimistic timeline points to late 2027 or 2028 for limited operations under tightly controlled conditions—specific routes, specific airports, specific weather minimums, with heavy FAA oversight. Scaled operations more realistically arrive in 2029 or 2030.

The trajectory, however, is clear. The pilot shortage is worsening. The economics of human crews on low-margin cargo routes are deteriorating. The technology works. The safety case is strong. The regulatory pathway is being built. An airplane has already flown itself from takeoff to touchdown with an empty cockpit.

What This Means for General Aviation Pilots

This technology is not coming for recreational flying. No one is automating the Saturday morning breakfast run. Reliable Robotics is solving a specific, difficult problem in cargo logistics. If they succeed, the result is safer airspace, a more resilient supply chain, and economically viable air cargo at small airports where service has been declining for years.

Key Takeaways

• Reliable Robotics has already flown a Cessna Caravan autonomously through every phase of flight with an empty cockpit, monitored by a remote pilot on the ground.
• Retrofitting a proven, type-certificated airframe avoids the novel-aircraft certification trap and leverages decades of Caravan operational history.
• The Part 135 cargo pilot shortage is the driving force—single-pilot night cargo operations are dangerous, hard to staff, and economically fragile.
• FAA certification is the primary bottleneck, with limited commercial operations possible by late 2027–2028 and scaled operations likely by 2029–2030.
• This targets cargo, not passengers—general aviation pilots are not affected, but the airspace and supply chain could benefit significantly.