HANBIT-Nano: connect an observed failure to corrective evidence

HANBIT-Nano attempted its first commercial orbital mission, SPACEWARD, from Alcântara on 22 December 2025 by Brazilian local time. INNOSPACE reported termination about thirty-three seconds into flight. The Brazilian Space Agency lists the carried Brazilian satellite missions as launch failures. Leaving the pad therefore did not mean the payloads reached their intended orbit.

This distinction matters when a launch has several participants. A ground team may complete its assigned operations while the vehicle fails later; payload builders may produce working hardware that never reaches space. An account should identify the mission outcome and the evidence for each subsystem without assigning a cause to a customer satellite simply because it was aboard the failed launch.

In March 2026, INNOSPACE reported that its joint investigation with Brazil’s CENIPA connected combustion-chamber rupture to gas leakage at the forward chamber assembly. The account associated the sealing problem with on-site reassembly. In June, the company reported completion of the official investigation and described component, assembly-process and quality-management improvements. These are the operator’s published accounts of that investigation.

A failure investigation needs a time sequence. Telemetry can show when a parameter changed; video can show visible events; recovered hardware can reveal physical damage. None of those sources alone automatically identifies the initiating cause. A crack observed after impact might be a consequence rather than the original failure. Investigators compare independent records and examine whether the proposed explanation accounts for their order.

Different evidence has different time resolution

Changing a component is only one part of a corrective action. The team must document what changed, why it addresses the identified mechanism and which test will evaluate it. Assembly records also matter: a component tested in one state may behave differently after transport, replacement or reassembly. Verification should remain connected to the actual configuration — open definition rather than to an earlier unit bearing the same name.

For an invented example, a sensor records pressure every twenty milliseconds while a video captures one frame every fifty milliseconds. Those streams have different time resolution and must be aligned before comparing events. A one-frame difference cannot establish an exact millisecond sequence. The practical lesson is to preserve timestamps, calibration and uncertainty along with the observations. A completed investigation and ground qualification support a return-to-flight decision; they do not guarantee the next flight’s result.