Redesign a cargo spacecraft

Japan developed HTV-X to carry supplies to the International Space Station and provide an orbiting platform for technology demonstrations after departure. It inherited experience from the original HTV Kounotori program, but changed the spacecraft architecture and cargo services. Studying the successor separately makes those choices visible.

The first vehicle, HTV-X1, launched on October 26, 2025. Its station visit and later free flight formed different phases of one mission. A cargo spacecraft can therefore deliver value after its freight has been unloaded, provided it retains the power, communications, control and other resources needed for that later work.

HTV-X combines navigation, power and propulsion functions in a service module, alongside a pressurized cargo module. Its projecting solar arrays and rechargeable batteries support the electrical system. Thrusters and associated plumbing are concentrated in the service module. The pressurized section holds supplies that crews can access once the vehicle is properly connected.

NASA’s approach photograph shows the projecting array and foil-covered spacecraft body. It illustrates the configuration, but does not reveal internal plumbing or prove the power available to an experiment. Those details need engineering documentation. Visible shape, specified capability and measured performance answer different questions.

JAXA describes three flight computers comparing their results. Redundant computing can help identify a disagreement, but counting computers alone does not prove that every failure is tolerated. A fault shared by multiple channels is different from one independent channel producing an incorrect result.

Consider an original simplified voting exercise: three channels report values 12, 12 and 19. A majority rule selects 12. If all three instead receive the same faulty input and report 19, agreement does not make that input correct. This illustration is not a reconstruction of HTV-X flight software; it explains why redundancy needs assumptions about failure detection and independence.