ocean · environment · family: the solution exists but nobody can afford it
underwater batteries nobodycan easily replace
Underwater sensor networks run on batteries that are difficult and costly to replace
Problem statement
Underwater Internet of Things (IoT) sensor networks are essential for continuous ocean monitoring — tracking water quality, currents, temperature, and marine life across coastal and deep-sea environments. But these networks are fundamentally limited by energy: underwater sensor nodes are battery-powered, battery replacement is difficult and costly, and, as Xu et al. put it, "a longer-range communication consumes more energy." Maintenance access is more difficult for submerged hardware because of submersion and environmental wear. Of the roughly forty systems Xu et al. reviewed, most operated on the water surface and only a few were deployed under water, though the authors note that underwater wireless sensor networks were not the focus of their review.
Why this matters
The 2025 Global Ocean Observing System (GOOS) Status Report found that the world's ocean observation infrastructure remains "subcritical" — with uneven coverage, declining drifting-buoy and ship-based observations, and heavy reliance on a small number of contributors — putting at risk the services it underpins: climate projections, weather and extreme-event forecasts, early warnings, and biodiversity conservation. The report names the Southern Ocean as the most under-sampled region, because of its remoteness and harsh conditions. Both papers name energy supply as a challenge: Xu et al. call battery replacement in marine monitoring systems difficult and costly, and Li and Shafieezadeh say power constraints remain a challenge for underwater sensor grids.
What’s been tried and why it hasn’t worked
Researchers have explored energy harvesting from solar, wind, wave, and ocean current sources, but only solar energy has been used in practice, in about a third of the systems Xu et al. reviewed. On the consumption side, Xu et al. point to optimized network topologies and routing protocols as the ways to reduce energy use. For submerged nodes, Li and Shafieezadeh note that power constraints are often addressed with underwater docking or energy harvesting systems, but that maintenance access is more difficult because of submersion and environmental wear. Most completed IoT marine monitoring systems operate on the water surface; only a few have been deployed under water. The harsh environment compounds the problem: Xu et al. cite currents, waves, tides, and typhoons and the need for very high water resistance, and Li and Shafieezadeh name storms, high salinity, and biofouling as harsh conditions a network must keep working under, and fluctuating currents, salinity, biofouling, and sensor drift as uncertain field conditions.
What would unlock progress
A breakthrough likely requires one of three approaches or their combination: (1) dramatic reduction in acoustic communication energy through novel modulation schemes or hybrid acoustic-optical-RF communication that matches the medium to the data rate needed; (2) reliable, persistent energy harvesting from ocean currents or thermal gradients at a scale that keeps a node running, robust against biofouling; or (3) a fundamentally different network architecture that minimizes per-node communication requirements, such as delay-tolerant networking where mobile platforms (AUVs, gliders) physically visit nodes to collect stored data, reducing the energy spent on long-range acoustic transmission. The hybrid glider–satellite architecture that Li and Shafieezadeh's analysis of thirty expert surveys ranked highest works on this pattern: gliders collect data and surface periodically to relay it by satellite.
Entry points for student teams
A student team could design and bench-test a marine energy harvesting unit that combines piezoelectric wave energy and thermoelectric current/gradient energy in a single housing, characterizing power output under simulated ocean conditions (wave tank with temperature differential). This is a feasible mechanical/electrical engineering project. Alternatively, a team could prototype a delay-tolerant data collection architecture where a small autonomous surface vehicle visits simulated underwater nodes (represented by Bluetooth or acoustic beacons in a pool), comparing the energy budget against conventional always-on acoustic networking.
Genes
Structural cousins: stuck the same way, in other fields
Sources
"Internet of Things in Marine Environment Monitoring: A Review," Xu et al., *Sensors*, MDPI, 19(7):1711, 2019. (accessed 2026-02-10). Supplemented with GOOS Status Report 2025 (UNESCO IOC) and "Smart sensor architecture selection for coastal marine monitoring," *Applied Water Science*, Springer, 2025. go to source ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
- GOOS Status Report 2025: https://www.ioc.unesco.org/en/global-ocean-observing-system-2025-status-report-reveals-both-progress-and-gaps-worlds-critical
- An earlier version of this brief gave a 125-times transmission-to-reception power ratio for underwater acoustic communication. None of the sources named here gives such a figure, so it has been removed; Xu et al. say only that longer-range communication consumes more energy.
- Underwater acoustic communication security (vulnerability to DoS, spoofing, and eavesdropping; the lack of lightweight encryption for resource-constrained hardware) is a related problem that none of the sources named here covers. It could be a separate brief with its own sources.
- A 2024 SCOR working-group proposal (SAFE-uIoT, "Global environmental risks of underwater acoustic data signalling: recommendations for a biofriendly underwater Internet of Things," https://scor-int.org/wp-content/uploads/2024/05/SAFE-uIoT_Proposal_2024.pdf) has flagged that scaling underwater acoustic IoT networks may itself pose environmental risks to marine biota through noise pollution — a constraint on any solution that relies on increased acoustic communication. The proposal was on the agenda of the SCOR 2024 Annual Meeting; it does not appear on SCOR's current working-group list.
- Cross-domain connection: the energy harvesting challenge shares structure with remote terrestrial sensor networks (e.g., glacier monitoring, volcanic monitoring) where similar power-communication tradeoffs exist but in less hostile environments.
- Argo float program provides context for the scale of the challenge: the GOOS report says recent growth in observations has been driven primarily by autonomous networks such as Argo profiling floats, and Deep Argo as measuring temperature and salinity down to 6 kilometres. The report does not discuss float battery budgets.
Reconciliation 2026-09-30: Fact repair after reading the whole write-up against every source it names, with a second reader going over the repairs the same day. Xu et al. 2019 was read in full from the PubMed Central copy (https://pmc.ncbi.nlm.nih.gov/articles/PMC6479338/ and the Europe PMC full-text XML for PMC6479338), because the MDPI page returns 403 to scripts; its citation record was checked at https://api.crossref.org/works/10.3390/s19071711. The GOOS Status Report 2025 was read from the IOC news page named above and from the interactive report at https://www.ocean-ops.org/goosreport/ (its full text bundle searched). The Li and Shafieezadeh paper (Applied Water Science 15:302, 2025, https://doi.org/10.1007/s13201-025-02664-2) was read in full from the publisher's PDF, saved earlier the same day (Springer answered later requests with a browser challenge); its record was checked at Crossref, DOAJ and OpenAlex. Earlier the same day a first pass had removed figures that none of the sources contains (a 125-times transmit-to-receive power ratio, systems that "die within months", Deep Argo at 18% of target, microwatt-to-milliwatt harvesting outputs, inductive charging, pressure and corrosion, and a claim that no lightweight encryption exists) and replaced them with what the sources say; both later readers confirmed that none of those figures appears in any source. Changes that stand after the second reading: (1) The long title said underwater sensor networks "die faster than they can be serviced," and the display title said their batteries "don't last the season." No source gives a lifetime or a servicing interval; Xu et al. say that "battery replacement is difficult and costly." The long title is now "Underwater Sensor Networks Run on Batteries That Are Difficult and Costly to Replace" and the display title "Underwater Batteries Nobody Can Easily Replace." (2) The problem statement said batteries "cannot be easily replaced or recharged at depth" and that maintenance access is harder for submerged hardware "than for surface platforms." No source discusses recharging at depth (Li and Shafieezadeh say power constraints are often addressed with underwater docking), and they say only that maintenance access for underwater grids "is more difficult due to submersion and environmental wear," noting separately that buoys may need more maintenance because of surface exposure; the sentence now follows the papers. (3) "Both review sources name energy supply and battery replacement as a central obstacle to sustained underwater deployment, which limits how far monitoring can scale" overstated both papers (and Li and Shafieezadeh is a decision study, not a review). The sentence now gives each paper's own wording: battery replacement in marine monitoring systems is difficult and costly (Xu et al.) and power constraints remain a challenge for underwater sensor grids (Li and Shafieezadeh). (4) Smaller wording fixes by the second reader: Li and Shafieezadeh list sensor drift among uncertain field conditions, not among the harsh conditions a network must withstand; their ranking came from a Fuzzy TOPSIS analysis of thirty expert surveys, not an expert panel's vote; the GOOS report credits growth in observations to "autonomous networks such as Argo profiling floats," not to Argo alone; the Source Note on the transmit-to-receive power ratio no longer says the asymmetry "is often described" as a driver, since no source named here describes it; and the problem statement no longer says that the small number of underwater systems in Xu et al.'s review is "the result" of the energy limits, because the paper draws no such link and says that underwater wireless sensor networks were "not the focus of this paper." (5) The Source Note attributed the acoustic-noise caveat to "the SCOR Working Group (2024)." The first reader could not find it on SCOR's working-group list (https://scor-int.org/work/groups/) and marked it unsourced; the second reader located the document: a 2024 SCOR working-group proposal, SAFE-uIoT (https://scor-int.org/wp-content/uploads/2024/05/SAFE-uIoT_Proposal_2024.pdf, read in full), which was on the agenda of the SCOR 2024 Annual Meeting (https://scor-int.org/wp-content/uploads/2024/09/Annotated-Agenda-SCOR-2024.pdf) and says that global deployment of an underwater Internet of Things "will entail a new input of sound into an already noisy underwater environment, leading to degradation and possible loss of habitat to marine animals." The original claim is restored with the attribution made precise (a working-group proposal, not an established working group) and the link added. Confirmed in the sources and left as written: the Source line for Xu et al. (Sensors 19(7):1711, 2019) and for the Applied Water Science paper; the quotation "a longer-range communication consumes more energy"; about forty systems reviewed, most on the water surface and only a few under water; only solar energy used in practice, in about a third of the systems; optimized topologies and routing protocols as the ways to cut energy use; currents, waves, tides, typhoons and very high water resistance; underwater docking or energy harvesting for power constraints; storms, high salinity and biofouling; gliders surfacing to relay data by satellite; "subcritical", uneven coverage, declining drifting-buoy and ship-based observations, reliance on few contributors, and the services listed; the Southern Ocean as the most under-sampled region because of remoteness and harsh conditions; Deep Argo to 6 kilometres; and that the report's text does not discuss float battery budgets. Not checked, because they are the write-up's own proposals rather than claims about a source: the three approaches in "What Would Unlock Progress" and the two entry points. Genome tags untouched. An accounting check the same day compared every changed sentence with the original and found that the first pass had also removed statements this note did not name: that the few systems that persist need ship-based servicing visits costing more than the data is worth and that wide-area underwater monitoring remains economically impractical; that the observing gap is most severe in polar regions, coastal developing nations and the deep sea, and that fisheries management is among the services at risk; that solar harvesting has been deployed only on surface buoys; and that on-node data processing saves battery at the price of hungrier processors. In "What Would Unlock Progress" the first pass also dropped a "milliwatt-to-watt" harvesting scale and a reference to thermoelectric generators from hydrothermal-vent research, and changed "the energy budget dominated by acoustic transmission" to "the energy spent on long-range acoustic transmission." Xu et al., Li and Shafieezadeh, the IOC page and the GOOS report text were fetched again and searched for each of these; none of them says any of it, with one exception: the GOOS report does mention fisheries in several of its case studies and network pages, but the IOC page's own list of services at risk, which the sentence now follows, does not include fisheries management. The removals stand.