Sign In
Skip to article

Back to research

[ PERSPECTIVE ]03 Nov 202614 min

Unmanned Surface Vehicles & the Black Sea

Ukraine's sea drones made the Black Sea Fleet's harbors unsafe. Of everything that produced that result, the boat travels worst.

Jesse Nuese
Managing Partner · 03 Nov 2026

View the PDF

Contents 14 min

Part 1 of 3. Part 2, Unmanned Surface Vehicles & the Strait of Hormuz, tests the same lessons against a first-rate navy. Part 3, Taiwan's Unmanned Deep Strike Challenge, tests whether Taiwan's unmanned strike components still work together once China's military attacks them.

Every serious conversation about defending Taiwan now reaches for the Black Sea. The US Pacific commander's Hellscape concept,1 which would flood the Strait with drones, leans on it. So does the People's Liberation Army's counter-drone training.2 What each borrows first is the hull: the unmanned surface vehicle, or sea drone, a five-meter explosive boat that sinks warships at a fraction of their price.

That is the wrong import. Four years of fighting show a platform whose results depended on what surrounded it: time to scout, a network to cue it, a supply line behind it, and an opponent that stayed in port. Few of those conditions hold in the Taiwan Strait. What does hold is less comfortable. Any port within reach of a cheap boat is now a target, and that applies to the side building the boats as much as to the side facing them.

01Reach, then denial

Missiles and mines won Ukraine its coast. The Neptune strike that sank the cruiser Moskva in April 2022, and the recapture of Snake Island that June, pushed Russian surface ships out of the north-western Black Sea and ended the amphibious threat to Odesa.3 Neither could reach the fleet at anchor in Sevastopol, about 300 kilometers from Odesa, or later in Novorossiysk, nearly 600. Sea drones could.

The record moves in steps. The first raid on Sevastopol,4 on 29 October 2022, damaged a minesweeper and a boom by Russia's own account and sank nothing. By October 2023 Russia had moved at least ten Black Sea Fleet ships from Sevastopol to Novorossiysk. In early 2024 swarms were sinking warships:5 a missile corvette near Lake Donuzlav on 1 February, a large landing ship off Alupka on 14 February, a patrol ship near the Kerch Strait on 5 March. In August 2026 a package of jet drones, missiles and maritime strike systems reached Novorossiysk too.6

The boats changed faster than their targets did. By the end of 2024 they carried infrared missiles, and on 31 December one destroyed a helicopter and damaged another. Ukrainian intelligence says they shot down two Su-30 fighters in May 2025;7 one loss is independently supported. By 2026 they were launching aerial drones and interceptors, and on 12 September the Ukrainian Navy said a gun-armed boat had destroyed a Russian one,8 in what Kyiv calls the first engagement of its kind. That claim has not been independently verified.

The result was sea denial: Russia lost free use of the water without Ukraine taking control of it. Russia still sorties submarines and fires cruise missiles when it accepts the risk, and in late September three damaged Kalibr-carrying ships were still operating. An April Carnegie article cites a Ukrainian naval analyst's estimate that Russia's sea-launched land-attack capacity is down about a quarter and its surface anti-ship capacity nearly 40 percent.9 Those are expert estimates, not counts of missiles or launchers. Most strike tallies come from Ukrainian services. Imagery confirms the major sinkings and the fleet's relocation, which is enough to carry the argument; it does not confirm every claim.

02The ratio holds, and narrows

Against confirmed sea-drone strikes from mid-2023 to July 2026, about $16 million of Ukrainian boats removed roughly $750 million of Russian ships and aircraft: a cost exchange on the order of 47 to 1,10 by PalBerg's unaudited reconstruction. The reconstruction is not itemized, and its target value includes two aircraft. It counts only losses attributed to sea drones, so it excludes the Moskva and everything else missiles sank. The mixed-weapon strike on Novorossiysk in August 2026 falls outside it, because public reporting does not assign its damage to sea drones. It also omits the largest cost, which recurs: layered booms, barges, gun crews and air patrols across two ports, and the loss of secure heavy repair anywhere on the Black Sea.

COST EXCHANGE, SEA DRONES VS RUSSIAN LOSSES

Cost Exchange Ratio View full size (opens in a new tab)
Fig. 1

About $16 million of Ukrainian boats against roughly $750 million of Russian ships and aircraft, confirmed sea-drone strikes, mid-2023 to July 2026.

Source: PalBerg, Intelligent Mass, August 2026; unaudited reconstruction.

That ratio is the right ruler because it prices cost per effect, the measure that fits a weapon built to be expended. Carried to Taiwan, it also shows where the analogy thins. Reported attack groups ran up to ten boats per large target; six hit Ivanovets. A 2023 CSIS wargame projected 113 to 138 Chinese ships lost across its scenarios.11 At about ten boats to overwhelm each ship's defenses, and allowing for a quarter to three quarters of the stock lost early in the fighting, one advocacy-leaning estimate puts Taiwan's need at roughly 1,500 to 5,600 strike boats.12

Public price points for Ukrainian boats run from about $220,000 for a Sea Baby to $250,000 to $275,000 for a Magura. At those prices, that force costs roughly $0.3 to $1.5 billion. Each ship engaged costs about $2.4 million in boats, rising to about $9.6 million if three quarters of the stock is lost early. Hulls built for rough water and produced at low volume push unit cost higher still. The exchange falls from 47 to 1 to somewhere in the low double digits or below. It still favors the defender against amphibious ships and most requisitioned ferries. The arithmetic is illustrative; no costed public estimate exists.

03What the Strait takes away

Distance favors Taiwan. The Strait is about 130 kilometers wide at its narrowest and 180 on average, and CNAS puts the outer defensive layer about 80 kilometers out.13 That allows smaller, cheaper boats in larger numbers. Near shore, shallow water, large tides and mudflats push an invasion fleet into predictable lanes.14

Almost everything else runs the other way.

The sea is rougher. Successful Black Sea strikes are documented in significant wave heights up to about 1.6 meters. A 1988 to 2017 simulation of the Strait puts mean significant wave height at about 1 meter in September and up to 2.8 meters in the January monsoon. Rough water limits small boats, though it limits landing craft too.

The opponent is stronger and has read the same record. Ukraine struck a degraded fleet with thin air defense, mostly at anchor, over years, after months of scouting each target. The PLA would likely push escorted convoys across over days to weeks, and its writers already study networked Ukrainian boats that scout,15 jam and strike. Its fleet defense against small craft should be expected to improve before any conflict.

The rear is gone. Ukraine had a long coast, hidden launch points and a land border to allied supply. Taiwan's west coast is dense and exposed, and one stress test assumes resupply is cut for the first 30 days of a blockade or invasion.16 On that assumption, whatever fleet exists on the first day is the fleet. Commercial satellite links, which carried Ukrainian targeting, cannot be assumed either.

Taiwan's program sits short of the need on any count. A defense ministry official put the Cabinet's June request at 1,320 sea drones over six years,17 though the released proposal gives no unit numbers and other reporting cites 2,000. The opposition-drafted act promulgated on 11 September directs NT$240 billion to unmanned systems over six annual budgets and sets no unit numbers.18 The Cabinet's September supplementary budget proposes a first batch of more than 100 small one-way boats in 2027,19 still subject to legislative approval.

TAIWAN SEA-DRONE NEED VS PLAN

Fleet Need View full size (opens in a new tab)
Fig. 2

Estimated requirement of roughly 1,500 to 5,600 strike boats against a June request of 1,320 and a first batch of 100-plus in 2027.

Source: Serpens Center, 2026, via SOFX, May 2026; Focus Taiwan, June 2026; Executive Yuan, September 2026.

04Where the lesson lands

The Black Sea finding that transfers intact is about harbors. No port within reach of a cheap boat is safe, and the defender pays for barriers, patrols and dispersal indefinitely. Russian booms at Sevastopol seemed fairly effective at protecting ships in harbor, mainly by slowing attackers,20 and a November 2023 storm damaged them at Sevastopol and on the Kerch Bridge.

In the Western Pacific that finding cuts in every direction. An invasion's weakest point may be its logistics tail,21 more than the landing itself, including the at least 63 civilian roll-on/roll-off ferries the Pentagon judges suitable for military use.22 The embarkation ports in Fujian and Zhejiang are within reach. So are Taiwan's own ports and gas terminals, the bases in Okinawa and the Ryukyus, and Philippine sites open to US forces.

The other side can build boats too. China has the world's largest shipbuilding base and a commercial drone industry. Russia, with far less, has already shown the path. In August 2025 it said its sea drone sank the Ukrainian reconnaissance ship Simferopol at the mouth of the Danube;23 Kyiv confirmed a strike and two crew deaths but not the sinking. Open-source analysts count 21 boathouses at what may be a new drone base in Crimea.24

The Ukrainian design is already moving east. In April, US and Philippine special forces struck a target vessel in the Luzon Strait during exercises; press reports say a Magura sank it,25 which US records do not state. In July, the manufacturer signed a memorandum with Oregon boatbuilder ReconCraft to pursue US production of the same boat family;26 no order or quantity has been announced. Where those boats sit matters as much as how many exist. Chinese defense writers argue that drones shipped forward during a crisis could be destroyed before use.27 That critique is fair to any concept built on moving the drones late.

05Judgments

Sea drones are one layer of Taiwan's denial system, alongside mobile anti-ship missiles, mines, underwater drones and aerial drones. They raise the cost of a crossing. On their own they would not stop a determined one. Moderate confidence.

Taiwan's binding constraints are scale, funding stability and the survival of the stock after a first strike. The technology exists. The public plan shows a first batch of 100-plus boats in 2027 and no schedule beyond it. We assess that a meaningful strike-boat stockpile is unlikely before 2028 to 2030, and an act with no unit numbers, reviewed each budget year, leaves room for delay. Moderate confidence.

The PLA will arrive having studied the small-craft problem for years. Per-boat effectiveness in the Strait will be lower than in the Black Sea, and a boat-on-boat fight should be planned for from the first day. It took four years to be reported off Crimea. Moderate confidence.

The most durable lesson for the United States and Japan is defensive. Their bases, repair yards and fuel terminals in the region need harbor defense, dispersal and repair plans now. Moderate to high confidence.

Five indicators would move these judgments:

  • Taiwanese contracts carrying unit numbers and delivery dates, at more than about 300 boats a year
  • Exercises that launch boats from dispersed, unprepared sites under jamming
  • Purchases of long-range, heavy-payload boats over short-range quadcopters
  • PLA escort screens drilled against sea drones, or PLA boats fielded in numbers near the Strait
  • US small-drone squadrons based in Japan or the Philippines before a crisis

The gap is contested testing. No public evidence shows a Taiwanese sea-drone package completing a strike after representative loss of navigation, links or cueing. That test, applied to the whole mission chain, is the subject of Part 3. Ukraine had four years to learn what its boats could do. Taiwan will find out on the first day.

Notes

  1. Center for a New American Security, Stacie Pettyjohn and Molly Campbell, Hellscape for Taiwan: Rethinking Asymmetric Defense, 26 February 2026, https://www.cnas.org/publications/reports/hellscape-for-taiwan. ↑
  2. Institute for the Study of War, China-Taiwan Update, 31 July 2026, https://understandingwar.org/research/china-taiwan/china-taiwan-update-july-31-2026/. ↑
  3. Eurasia Review, How Drones, Mines and Missiles Contained Russia's Black Sea Fleet, 21 September 2026, https://www.eurasiareview.com/21092026-how-drones-mines-and-missiles-contained-russias-black-sea-fleet-analysis/. ↑
  4. Rabdan Security and Defence Institute, Leonardo Jacopo Maria Mazzucco, How Ukraine's Unmanned Surface Vessels Have Reshaped Modern Naval Warfare in the Black Sea, 15 October 2025, https://rsdi.ae/en/publications/how-ukraines-unmanned-surface-vessels-have-reshaped-modern-naval-warfare-in-the-black-sea; Ukrainska Pravda, Roman Petrenko, Explosions in Sevastopol: Russians allege "UAV attack" by Ukraine that damaged ship, 29 October 2022, https://www.pravda.com.ua/eng/news/2022/10/29/7374072/. ↑
  5. Defence Intelligence of Ukraine, Naval Drones of the Defence Intelligence of Ukraine Dominate the Black Sea, undated, https://gur.gov.ua/en/content/morski-drony-hur-dominuiut-u-chornomu-mori; The Maritime Executive, Video: Ukraine Sinks Another Russian Warship Using Drone Boats, 1 February 2024, https://maritime-executive.com/article/video-ukraine-sinks-another-russian-warship-using-drone-boats; RBC-Ukraine, Danylo Kramarenko, Magura vs. Sea Baby: Closer look at Ukrainian drone warfare against Russian ships, 7 March 2024, https://newsukraine.rbc.ua/news/magura-vs-sea-baby-closer-look-at-ukrainian-1709831518.html. ↑
  6. Escudo Digital, Juan Carlos De la Torre, Ukraine Takes the Drone War to the Russian Fleet's Last Safe Haven in the Black Sea, 12 August 2026, https://www.escudodigital.com/en/defense/russia/ukraine-takes-the-drone-war-to-the-russian-fleets-last-safe-haven-in-the-black-sea.html; Ukrinform, Zelensky announces unique operation by Defense Forces to strike naval base in Novorossiysk, 12 August 2026, https://www.ukrinform.net/rubric-ato/4153471-zelensky-announces-unique-operation-by-defense-forces-to-strike-naval-base-in-novorossiysk.html. ↑
  7. Finabel, Sebastian Morabito, Info Flash, July 2025, https://finabel.org/wp-content/uploads/2025/07/IF-Sebastian-Morabito-July-2025.pdf. ↑
  8. Kyiv Independent, Ukrainian Naval Drone Destroys Russian Unmanned Boat in Black Sea, Navy Says, September 2026, https://kyivindependent.com/first-ever-battle-ukrainian-naval-drone-destroys-russian-unmanned-boat-in-black-sea-navy-says/. ↑
  9. Carnegie Endowment for International Peace, The Changing Military Balance in the Black Sea: A Ukrainian Perspective, April 2026, https://carnegieendowment.org/research/2026/04/the-changing-military-balance-in-the-black-sea-a-ukrainian-perspective. ↑
  10. Palisade Bergschrund, Intelligent Mass, August 2026, https://pal-berg.com/research/intelligent-mass. ↑
  11. Mark F. Cancian, Matthew Cancian and Eric Heginbotham, The First Battle of the Next War: Wargaming a Chinese Invasion of Taiwan, CSIS, 9 January 2023, https://csis-website-prod.s3.amazonaws.com/s3fs-public/publication/230109_Cancian_FirstBattle_NextWar.pdf. ↑
  12. Serpens Center (formerly the Snake Island Institute), The Black Sea Blueprint for Taiwan: Ukraine's Naval Lessons in Asymmetric Warfare, 2026, as summarized by SOFX, 25 May 2026, https://www.sofx.com/ukraines-naval-drone-warfare-playbook-applied-to-the-taiwan-strait/. ↑
  13. Center for a New American Security, Stacie Pettyjohn and Molly Campbell, Hellscape for Taiwan: Rethinking Asymmetric Defense, 26 February 2026, https://www.cnas.org/publications/reports/hellscape-for-taiwan. ↑
  14. SOFX, Ukraine's Naval Drone Warfare Playbook Applied to the Taiwan Strait, 25 May 2026, https://www.sofx.com/ukraines-naval-drone-warfare-playbook-applied-to-the-taiwan-strait/. ↑
  15. Foundation for Defense of Democracies, China's War Wolves: From Commercial Tech to Combat Power, 3 May 2026, https://www.fdd.org/analysis/2026/05/03/chinas-war-wolves-from-commercial-tech-to-combat-power/. ↑
  16. Serpens Center (formerly the Snake Island Institute), The Black Sea Blueprint for Taiwan: Ukraine's Naval Lessons in Asymmetric Warfare, 2026, as summarized by SOFX, 25 May 2026, https://www.sofx.com/ukraines-naval-drone-warfare-playbook-applied-to-the-taiwan-strait/. ↑
  17. Focus Taiwan, Sean Lin, Cabinet Proposes US$6.6 Billion Special Budget for Domestic Drones, 18 June 2026, https://focustaiwan.tw/politics/202606180012. ↑
  18. Office of the President, Republic of China (Taiwan), Presidential Office Gazette No. 7881, Act for Strengthening Defense Autonomy and Developing the Unmanned Vehicle Industry (強化國防自主暨無人載具產業發展條例), 11 September 2026, https://www.president.gov.tw/File/Doc/9cd77974-7763-4bd5-a5e2-2bc0bc19fb89. ↑
  19. Executive Yuan, Republic of China (Taiwan), Cabinet Approves NT$607.6 Billion Supplementary Central Government Budget for 2026 (政院通過115年度中央政府總預算追加預算案6,076億元), 3 September 2026, https://www.ey.gov.tw/Page/9277F759E41CCD91/e6b90fcf-0d42-40e3-b001-1c790c076657. ↑
  20. Royal United Services Institute, Russia Deploys Boom Defenses Around Crimea Bridge To Block Drone Strikes, 31 July 2023, https://rusi.org/news-and-comment/in-the-news/russia-deploys-boom-defenses-around-crimea-bridge-block-drone-strikes; Howard Altman, Russia's Harbor Defenses In Sevastopol In Disarray After Storm, The War Zone, 28 November 2023, https://www.twz.com/russias-harbor-defenses-in-sevastopol-seen-battered-by-storm; Defense Express, How Russia Is Hunting Ukrainian Sea Drones with Be-12 Search Aircraft and Ka-29 Helicopters, and What It Means, 24 December 2023, https://en.defence-ua.com/analysis/how_russia_is_hunting_ukrainian_sea_drones_with_be_12_search_aircraft_and_ka_29_helicopters_and_what_it_means-8972.html. ↑
  21. Defense.info, The Narrow Sea: Why a Taiwan Invasion Fleet's Real Vulnerability Is Logistics, Not the Landing, August 2026, https://defense.info/maritime-dynamics/2026/08/the-narrow-sea-why-a-taiwan-invasion-fleets-real-vulnerability-is-logistics-not-the-landing/. ↑
  22. US Department of Defense, Military and Security Developments Involving the People's Republic of China 2023, 19 October 2023, https://media.defense.gov/2023/Oct/19/2003323409/-1/-1/1/2023-MILITARY-AND-SECURITY-DEVELOPMENTS-INVOLVING-THE-PEOPLES-REPUBLIC-OF-CHINA.PDF. ↑
  23. The War Zone, Russia Just Made Its First Successful Drone Boat Attack on a Ukrainian Naval Ship, August 2025, https://www.twz.com/sea/russia-just-made-its-first-successful-drone-boat-attack-on-a-ukrainian-naval-ship. ↑
  24. The Defense Post, Rojoef Manuel, Russia Builds Crimea Base to Launch Naval Drones at Scale, 4 September 2026, https://thedefensepost.com/2026/09/04/russia-crimea-naval-drones/. ↑
  25. Militarnyi, Ukrainian Magura Sank a Ship Target During the US-Philippine Balikatan 2026 Exercises, 9 June 2026, https://militarnyi.com/en/news/ukrainian-magura-sank-ship-balikatan-2026/; SOFX, Green Berets Sink Ship With Explosive Drone Boat in Philippines Exercise, 5 May 2026, https://archive.sofx.com/green-berets-sink-ship-with-explosive-drone-boat-in-philippines-exercise/. ↑
  26. UForce, UForce Partners with ReconCraft to Build Combat-Tested Autonomous Maritime Drones in the US, July 2026, https://www.uforce.com/news/uforce-partners-with-reconcraft-to-build-combat-tested-autonomous-maritime-drones-in-the-u-s. ↑
  27. Taiwan News, Kelvin Chen, Easton Says China Doubtful of US "Hellscape" Concept in Taiwan Strait, 7 September 2026, https://www.taiwannews.com.tw/news/6435023. ↑
Download the PDF ↓

Jesse Nuese

Managing Partner

General Disclaimer. This page, and Palisade Bergschrund Corporation’s (“Pal-Berg’s”) research, analysis, validation, testing, and advisory materials, are provided for general informational and evaluation purposes only, are subject to change, and do not constitute legal, regulatory, tax, accounting, investment banking, broker-dealer, placement agent, investment advisory, procurement, technical, operational, or military advice. This page is not a legal opinion, regulatory assessment, procurement advice, financial advice, or recommendation regarding any particular market-entry, transaction, teaming, licensing, investment, export-control, CFIUS, FOCI, facility-clearance, funding, or government-contracting strategy. Pal-Berg does not guarantee the accuracy, completeness, timeliness, future applicability, operational effectiveness, regulatory approval, procurement approval, financing outcome, investment outcome, transaction outcome, product performance, fitness for a particular purpose, or commercial success of any technology, product, company, market, or opportunity discussed. Pal-Berg makes no warranties, express or implied, regarding the accuracy, completeness, reliability, or fitness for a particular purpose of its research. Pal-Berg shall not be held liable for any direct, indirect, incidental, consequential, special, or punitive damages arising from the use or reliance upon this Research, including but not limited to loss of life, personal injury, property damage, or mission failure. Any reliance on Pal-Berg materials is at the recipient’s own risk and should be made only in connection with the applicable agreement, statement of work, and the recipient’s own independent technical, legal, regulatory, operational, and business judgment and advice from qualified counsel and other professional advisors.

Legal, Regulatory, Procurement, and Export-Control Matters. Statements regarding U.S. defense procurement, domestic-preference requirements, FOCI, CFIUS, export controls, cybersecurity requirements, funding programs, facility clearances, teaming, subcontracting, licensing, or other regulatory matters are provided for general informational purposes only. These areas are fact-specific, change frequently, and may depend on the recipient’s jurisdiction, ownership, product classification, customer, transaction structure, end use, end user, funding source, and other facts. Recipients should consult qualified legal, regulatory, export-control, procurement, tax, accounting, and other professional advisors before taking action or making decisions based on this page.

No Product Warranty; Third-Party Products. Pal-Berg does not manufacture, design, own, sell, supply, maintain, warrant, or guarantee any third-party or customer products, systems, hardware, software, equipment, components, or technologies reviewed, tested, validated, researched, or discussed. Pal-Berg disclaims any warranty, express or implied, regarding such products, systems, hardware, software, equipment, components, or technologies, including any implied warranty of workmanship, merchantability, fitness for a particular purpose, safety, operational suitability, durability, or performance. Product-level claims, specifications, warranties, defects, failures, and compliance obligations remain solely the responsibility of the applicable manufacturer, supplier, owner, or customer, and Pal-Berg shall not be responsible for any losses, claims, damages, injuries, failures, or liabilities arising from or relating to such products, systems, hardware, software, equipment, components, or technologies, except to the extent expressly agreed in a written agreement signed by Pal-Berg.

Operational Decisions. Pal-Berg’s services shall not be construed as operational, tactical, or strategic military advice. All operational decisions remain the sole responsibility of the commanding authority and designated decision-makers. Pal-Berg’s research, analysis, validation, testing, and advisory materials are few inputs among many that should inform, but not solely determine, operational decisions.

No Government or Third-Party Endorsement. This page is not prepared, approved, endorsed, or sponsored by the U.S. Army, the U.S. Department of Defense, or any other government agency, military organization, prime contractor, or third-party organization referenced herein. References to third parties are for informational purposes only.

pal-berg.com · ascend@pal-berg.com
 © 2026 Palisade Bergschrund Corporation. All rights reserved.

Legal disclaimers page

Notices · Editorial standards v1.140°44′34″ N · 73°59′15″ W