What Certificate Level Part 108 Software Should Look Like

Certificate-level Part 108 software has one job, and it is easy to build the wrong thing. The obvious product is a logbook with compliance reports. Store flight records, attach documents, and export a polished PDF when someone asks. That may be useful, but it is not the center of the problem.

The right product is a gate. The most important moment is the second before dispatch, when the system looks at the whole operation and says yes or no. Aircraft configuration accepted. Crew trained and current. Operational area authorized. Required procedures in force. Risk cleared. No open safety blockers. If one of those facts is wrong, the flight should not launch.

The certificate is about system trust

The proposed certificate path changes who carries the trust. Under the familiar Part 107 model, the certified remote pilot is central. Under a certificate-style BVLOS operation, the operator organization and its system become the focus. The FAA needs to believe the operator can repeatedly conduct authorized flights without one-off improvisation.

That means the software has to do more than remember what happened. It has to make the release decision reproducible. The same inputs must produce the same result. Every finding should cite the rule or internal control behind it. Every override should be captured inside the record. Every correction should supersede, not erase.

The five records need one spine

The five core facets are aircraft and configuration, personnel qualification, operational area, risk and SMS, and the flight record. Many tools can store each facet separately. Certificate-level software connects them around a mission. An auditor should be able to start with one flight and traverse to the exact aircraft state, the assigned people, their training, the approved area, the governing procedure, the risk decision, and the sealed evidence.

That graph matters because the real audit question is not whether the operator owns documents. The question is whether this flight was authorized and whether the operator can prove why.

What belongs on top of the gate

A real safety system has to be alive. Hazards should be logged, owned, analyzed, mitigated, and closed. Incidents should link back to the exact flight, configuration, and crew. Open safety findings should be able to block related dispatch. Evidence packs should assemble from sealed source records in minutes. Auditors should be able to inspect a scoped, read-only view without changing anything.

The discipline is knowing what the product is not. It is not the autopilot, not mapping software, not air traffic control, and not a generic analytics dashboard. It sits above execution as the layer that decides and the layer that proves. Build the gate. Make it hard to fool. Make every answer traceable.