What a survey looks like from the surface
From the deck it looks like nothing at all. The vehicle goes over the side and comes back hours later, and in between it makes one acoustic decision per ping interval without asking anyone.
The count
Independent waveform decisions in a single 8-hour survey, at a 2-second ping interval. There is no version of this where an operator retunes between pings. That is the entire reason the decision has to live on the payload.
The run
- 01 ReferenceBrief Deck
Survey area, required range and line spacing are set on deck. This is the last point a human chooses anything acoustic.
- 02 ReferenceLaunch Vehicle
The vehicle goes over the side. The radio link is dropped and the payload is on its own.
- 03 ModelledDescend and settle Payload
The payload samples the water column on the way down and has a working picture of the channel before the first line starts.
- 04 ModelledRun the lines Payload
One decision per ping interval, for hours. Every one of them is the payload re-solving against conditions it just measured.
- 05 ModelledCross a plume Payload
Sediment rises, the link budget stops closing, the centre frequency drops and the pulse lengthens. Nobody is told, because nobody needs to be.
- 06 ImplementedRecover Vehicle
The vehicle surfaces. The log carries every ping, the conditions it was fired in, and why those parameters were chosen.
- 07 ImplementedRead the survey Deck
On deck, the adaptation events are visible in the log as a record of where the water changed.
The deck sets the brief and reads the log. Everything acoustic between those two moments belongs to the payload — which is why "the operator can just change the setting" is not an answer to the problem this project exists for.