Autonomous platforms have created new possibilities for persistent, large-scale ocean observation, but they have also sharpened a longstanding tension in oceanographic instrumentation: the more a sensor demands in power, the less a platform can offer in endurance, sampling resolution, or sensor diversity. For AUVs, ocean gliders, profiling floats, and wave-powered vehicles operating far from ship support, this is not an abstract trade-off — it directly determines what science gets done and what gets left undone.
This paper examines how treating power consumption as a primary design constraint, rather than a secondary specification, shapes instrument architecture and ultimately expands the envelope of autonomous observation. Using RBR's experience developing sensors across the autonomous platform ecosystem — including CTDs for gliders and AUVs, profiling float instruments for the international Argo and BGC Argo programs, and OEM sensor suites for integration into third-party vehicles — we explore the practical consequences of design decisions around conductivity cell architecture, sampling strategies, and multi-sensor power management.
Specific examples include the use of naturally flushed inductive conductivity cells that eliminate pump power and mechanical complexity; adaptive multi-rate sampling that allows high-frequency data acquisition while running auxiliary sensors at reduced duty cycles; and the development of biogeochemical sensors including fluorescence, backscatter, dissolved oxygen, turbidity — within the tight power and form-factor constraints of operational autonomous platforms. Data from Argo float deployments illustrate how energy budgeting at the instrument level translates directly into float longevity and data return across multi-year observation campaigns.
The broader argument is that instrumentation designed around platform constraints — rather than adapted to them after the fact — is a prerequisite for realising the full potential of autonomous ocean observing. As Australian marine science increasingly relies on autonomous systems for sustained, large-area observation, the relationship between sensor engineering and platform capability deserves more explicit attention in how programs are planned and instrumentation is selected.