Pipe Coding: a research direction for learning how intelligent machines might work in constrained spaces.

How might a machine sense, move, communicate and return useful evidence inside a constrained pipe or enclosed infrastructure environment? The question is intentionally broader than any single proposed device.

The material can be organised around locomotion, aerial or mobile access, sensing, localisation, communication, power, inspection evidence, recovery and human oversight. Each subsystem has different assumptions and risks.

Unless supported by dated test evidence, the project does not claim operational navigation, defect detection accuracy, endurance, safe deployment, environmental tolerance or regulatory acceptance.