Loading...
What the Two Production Launches of 2026 Change in UISC's Earlier Assessment
In February, UISC concluded that Russia had been preparing for the loss of Starlink but had failed to create a viable alternative in time. Since then, Bureau 1440 has conducted two production launches. The new orbital data confirm a transition from plans to deployment, but also show uneven progress: by 6 September, only 13 of the 32 launched production spacecraft had joined the observed cluster at roughly 509–514 km, while none of the 16 satellites in the second batch had risen above about 426 km.
Previous article: Roscosmos After Starlink: Russia Was Preparing for Disconnection – But Was Not Ready in Time
On 19 February 2026, UISC published an article on the consequences of disabling unauthorised Russian Starlink terminals. Its central conclusion was that Moscow had prepared in advance for the loss of Western satellite infrastructure, supported Bureau 1440, and announced mass production of terminals, but the production timetable had not matched battlefield requirements. The issue was no longer intent, but pace.
Six months later, this conclusion can be refined. In March and July, Bureau 1440 placed two batches of 16 Rassvet-3 spacecraft into orbit. One satellite from the first batch has been lost. According to orbital data current to 6 September, 13 of the 32 launched production spacecraft had formed a nearly circular cluster at roughly 509–514 km, while none of the second batch had climbed above about 426 km. Two second-batch satellites remained in very low orbits of approximately 273 × 298 km and faced an increased risk of accelerated decay unless manoeuvring resumed. Russia has therefore moved from prototypes to physical deployment, but orbit raising remains slow and uneven.
The earlier article described Rassvet primarily as a programme that had not yet become an orbital capability. That wording now requires revision: production satellites exist, are launched in batches, and manoeuvre independently after separation from the launch vehicle. Yet launch, arrival at an operational orbit, communications testing, entry into service, and service provision are separate stages. Public data confirm only part of this chain.
As of 6 September, it was accurate to speak of 32 launched production spacecraft and 31 that physically remained in orbit, rather than 31 ‘operational’ satellites. RASSVET-3 Nos. 17 and 23 remained in very low orbits of approximately 273 × 298 km, leaving them more exposed to accelerated orbital decay if manoeuvring did not resume; this is not the same as confirmed imminent re-entry. Nor should the six older experimental Rassvet-1 and Rassvet-2 spacecraft be mechanically added to the production constellation. Assessing the threat requires separate monitoring of the orbital group, the ground segment, and verified use of the service.
The chronology shows a transition from experimentation to production deployment. Three experimental Rassvet-1 spacecraft were launched on 27 June 2023, followed by three full-scale Rassvet-2 prototypes on 17 May 2024. The first production batch of 16 Rassvet-3 satellites was launched on 23 March 2026; one was subsequently lost. A second batch of 16 satellites was launched on 19 July, after which the spacecraft began raising their orbits.
The first 16 production satellites were launched from Plesetsk on 23 March 2026 aboard a Soyuz-2.1b into an initial orbit of around 300 km with an inclination of 82.3 degrees. The spacecraft were then expected to raise their own orbits using plasma thrusters. RASSVET-3 No. 4 performed no observable manoeuvre and re-entered the atmosphere on 6 June. This was the first confirmed loss of a production spacecraft in the system.
By 6 September, 13 satellites from the first production batch had formed a nearly circular observed cluster at roughly 509–514 km. RASSVET-3 No. 10, which had been one of the principal laggards in early August, had climbed to approximately 509 × 512 km; two surviving first-batch spacecraft remained outside the cluster. The different orbit-raising profiles may reflect the test programme, orbital phasing, or individual problems, but open data do not establish the precise cause.
The second batch of 16 satellites was launched on 19 July. Orbital data current to 6 September showed that none had risen above about 426 km. RASSVET-3 No. 29 was the highest, at approximately 415 × 426 km, while Nos. 17 and 23 remained at roughly 273 × 298 km. Several spacecraft had paused after an initial period of orbit raising. This is evidence of a slow and uneven campaign, but not proof that all second-batch satellites have failed.
Important caveat. None of this means that the launch vehicle failed to place the second batch in space, or that all 16 spacecraft have been irretrievably lost: some may still resume manoeuvring. The confirmed fact is narrower — by 6 September, no second-batch satellite had joined the observed 509–514 km cluster. Open orbital data establish neither whether this cluster is the final operational shell nor the reason for the second batch’s delay. Claims involving a specific failure of thrusters, software, power systems, or imported components therefore remain speculative.
Bureau 1440's official website still lists 800 km as the system's target altitude. By 6 September, however, 13 first-batch satellites had converged in a nearly circular cluster at roughly 509–514 km. This strengthens the hypothesis that the project is using a lower shell near 511 km, but Bureau 1440 has not publicly confirmed whether it is the final operational altitude rather than an interim test or phasing orbit.
A lower orbit can reduce latency and improve the radio link budget, but it also reduces the footprint of each satellite and increases the number of spacecraft required. One expert cited by ComNews estimated that moving from 800 to about 500 km, with other parameters unchanged, could require a constellation roughly 1.8–2.2 times larger to provide comparable continuous coverage.
On 28 August, Foreign Minister Andrii Sybiha said that Ukraine had submitted a note to the ITU Radiocommunication Bureau requesting that Ukrainian territory be excluded from the declared service areas of all Rassvet satellite networks across all frequency bands. Kyiv also called on partners to seek the same exclusion for their own territories. This was the first publicly confirmed international countermeasure against the system, coordinated by the Ministry of Foreign Affairs, the Ministry of Digital Transformation, and the Ukrainian State Centre of Radio Frequencies.
At the same time, an appeal to the ITU does not physically switch off satellites. The Union coordinates the international use of radio frequencies and the registration of satellite networks, while the outcome will depend on the type of Russian filings, the applicable procedures, and the positions of member states. The note creates a formal basis for challenging Rassvet's declared service areas, but its practical effect remains to be assessed.
The federal project 'Internet Network Access Infrastructure' envisages the production of 383 spacecraft in 2025–2030: 292 satellites for the operational constellation and 91 replacements for spacecraft reaching the end of their service lives. The state is also providing subsidised concessional lending and launch services. Published financial estimates indicate that the project is expected to receive RUB 102.8 billion in federal funding and around RUB 329 billion in the company's own investment by 2030.
The first production launch was originally scheduled for 2025 but was postponed until March 2026. The formal schedule required 156 satellites in orbit by the end of 2026 and 292 by the end of 2027. After two launches, the 2026 target was still short by 125 spacecraft. Meeting it would have required seven further launches in five months, carrying about 18 satellites per rocket, although the two demonstrated configurations each carried 16. Experts consider an increased payload for the Soyuz-2.1b technically possible, but describe this timetable as exceptionally demanding because the spacecraft would have to be manufactured, tested, accepted, and brought into service simultaneously.
In February, UISC assessed that several hundred satellites by 2027 could form a regional network sufficient for some Russian military requirements. That threshold remains meaningful, but the current pace does not bring Russia close to it under the formal schedule. By 6 September, 13 of the 32 launched production spacecraft had joined the observed 509–514 km cluster — 40.6 per cent. This is an orbital snapshot, not a measure of operational availability. The headline figure of ‘31 satellites in orbit’ therefore overstates service readiness. For the next few years, the federal plan of 292 operational satellites and 383 spacecraft manufactured in total remains the more reliable benchmark.
Under the manufacturer's concept, the satellites are intended to operate as orbital 5G base stations. Laser inter-satellite links are expected to relay traffic across the orbital network to distant gateways, reducing the number of ground stations required. The stated performance is up to 1 Gbps with latency of up to 70 ms. These figures, however, come from Bureau 1440 itself; no independent measurements of production Rassvet-3 performance are publicly available.
The publicly displayed user terminals use an active phased-array antenna, measure approximately 60 by 60 cm, and weigh up to 15 kg. They are intended for fixed sites, road vehicles, trains, ships, and aircraft. By comparison, the Starlink Mini antenna weighs around 1.1 kg. This does not make the Russian terminal unsuitable for military use, but it substantially narrows the range of viable platforms: the current design is unlikely to suit small FPV drones, although it could be installed on command vehicles, ships, large uncrewed aircraft, ground positions, or transport infrastructure.
The earlier UISC article cited a Russian claim that 200,000 satellite terminals for uncrewed systems would be produced in 2026. As of 7 September, open sources did not confirm production at that scale or serial deliveries to the armed forces. Bureau 1440 says it manufactured enough terminals for its 2026 test programme, but has not disclosed the quantity, price, power consumption, or real-world performance. The publicly displayed terminal, weighing up to 15 kg, should not be automatically equated with an airborne module for an uncrewed aircraft.
At the same time, there are signs that the company is preparing a manufacturing and commercial ecosystem, not only an orbital platform. Bureau 1440 has announced serial production of space-grade solar panels, an agreement with MegaFon to connect the first 500 remote base stations in 2027, and an agreement with Beeline to supply equipment and communications capacity. This does not demonstrate service readiness, but it shows that ground integration is being prepared in advance.
The earlier article identified the fundamental gap in scale between Starlink and Russia's plans. Two launches have not removed it: of the 31 production spacecraft that physically remained in orbit by 6 September, 13 had joined the observed 509–514 km cluster, while none of the second batch had risen above about 426 km. Launch and orbit counts alone do not demonstrate service readiness.
On 3 September, X Holding chief executive Alexey Shelobkov said that Bureau 1440 had begun internal tests and initial trials with selected users in August, manufactured enough terminals for the 2026 test programme, and installed a terminal on a Russian Railways long-distance train, where testing was expected to begin shortly. These are company claims, not independently verified results. No public data yet demonstrate production-series throughput, session duration, seamless handover between satellites, service availability, or resistance to jamming.
Calculations for the chain of twelve satellites from the first launch indicated up to 3.2 hours of geometric radio visibility per day over the Moscow region, a maximum continuous contact of about 52 minutes, and, in some cases, a wait of up to 10.7 hours for the next session. Published estimates for Ukraine suggested roughly two sufficiently high passes per day, lasting around one to one and a half hours. These are potentially useful communications windows, but they do not provide continuous coverage.
Key evidentiary threshold. Satellite visibility over Ukrainian territory does not prove that the spacecraft is already providing broadband service, that ground gateways and user terminals are operational, or that Russian uncrewed systems are being controlled in combat through this channel.
Rassvet's civilian positioning should not obscure its dual-use character. On 12 June 2026, Russian service personnel asked Vladimir Putin whether Russia could develop heavy drones controlled through a satellite link, following Ukraine's example. Putin referred to Bureau 1440, claimed that the system was 'in no way inferior to Starlink', and expressed the hope that the military would soon 'feel it on the ground'. The question itself shows that the problem had not yet been solved at unit level, while Putin's words confirm the project's political and military priority, not its readiness.
Vadym Skibitskyi, Deputy Chief of Ukraine's Defence Intelligence, said in August that the current number of spacecraft allowed the network to be used only briefly during a satellite pass, and that countermeasures were already being discussed in Ukraine. On 7 September, Defence Intelligence Chief Oleh Ivashchenko likewise described the current system as having low effectiveness and said that 32 satellites were insufficient for stable communications. Together, these assessments point to a potentially useful but intermittent capability, not a persistent communications channel.
Rassvet does not replace GLONASS: the navigation system determines coordinates, while the satellite communications network provides two-way data transmission. A platform equipped with the appropriate terminal could theoretically follow an autonomous route and receive a mission update or transmit telemetry during a short communications window. This makes limited military use possible before continuous coverage is available — particularly for command posts and large uncrewed platforms.
At the same time, there is no public technical evidence that Russian forces are already using Rassvet systematically to control Shahed, Geran, or other strike platforms. A theoretical possibility, Putin's statements, and the geometric visibility of satellites do not amount to a confirmed combat episode. That distinction is essential in an analytical assessment.
Active electronically scanned array antennas and narrow-beam operation in the Ku/Ka bands may complicate interception and jamming, but they do not make the channel inherently invulnerable. Nor does a satellite pass coinciding with a launch or impact prove control through Rassvet: such a conclusion would require confirmed terminal emissions, data on ground-segment activity, and sound statistical analysis.
Laser inter-satellite links reduce dependence on a large number of gateways, but do not eliminate it. The orbital network still requires control centres, ground stations, connections to the internet backbone, and the manufacture and programming of terminals. Unlike gateways serving Ukraine from NATO territory, Russia's infrastructure would have to be located primarily in Russia or allied states. This makes it potentially more accessible to intelligence collection, cyber operations, and physical disruption.
Agreements with mobile operators nevertheless change the picture of the ground segment. If MegaFon does connect the first 500 base stations to the network in 2027, Rassvet could use existing telecommunications infrastructure both to serve remote regions and to back up terrestrial links. The Beeline agreement and partnerships with transport and industrial companies could create a user base even before continuous coverage becomes available. For military analysis, this means monitoring not only dedicated gateways but also integration with civilian operators.
Ground stations or operators' base stations should not, however, be portrayed as a single 'off switch'. A normal architecture will incorporate redundancy and dispersal, while laser links will allow traffic to be rerouted to another gateway. Likewise, the loss of one or several satellites would not destroy a large distributed network. Effective countermeasures must operate at several levels, from component supply and launches to terminals, frequencies, software, and ground infrastructure.
Bureau 1440 claims to have developed its satellites, laser terminals, and ground infrastructure in-house. Yet a study by the Centre for Defence Reforms, reported by RBC-Ukraine, describes a production network of more than 1,300 companies. The researchers identified 21 major suppliers, each of which received more than RUB 100 million in 2025; together, they received approximately RUB 7.4 billion. Only five were under Western sanctions: Sotis, RPC Micran, Elesar Group, Rezonit, and OKB Fakel.
According to the same study, Bureau 1440 made foreign-currency payments for imports worth approximately USD 68 million in 2025. Specialised equipment from Western manufacturers, including Rohde & Schwarz, Keysight, and TOPTICA, reached Russia through Chinese and Turkish suppliers. Particularly sensitive areas include radiation-hardened electronics, attitude-control systems, optics, plasma thrusters, phased-array antennas, and test equipment. These figures do not come from the company's audited accounts and should therefore be treated as the findings of an OSINT investigation, not as an officially confirmed map of all supply chains.
The first gap concerns the core of the project. X Holding has been under US sanctions since February 2023, but Bureau 1440 LLC and its chief executive, Alexey Shelobkov, were not directly listed in the Western registers reviewed by the Centre for Defence Reforms as of August 2026. The production entities FAB Rassvet LLC and Labs 1440 JSC should also be treated as priority candidates.
The second gap concerns twelve major suppliers for which the study identifies no Western restrictions: Pallada LLC, Belarusian Aerospace Instruments CJSC, RPE TEKS LLC, Aquamash Production Association JSC, Kazakhstan's Alternative Solutions LLP, Radio Gigabit LLC, Belarusian Belrobot CJSC, Newton Steel LLC, Macro Solutions LLC, SPC Small Spacecraft LLC, Labs 1440 JSC, and El-Optics LLC. These are not incidental links: each company received more than RUB 100 million from Bureau 1440 in 2025.
A further four suppliers are already under Ukrainian sanctions but, according to the study, were not covered by corresponding Western listings: Infostera LLC, MicroEM Component LLC, Yue Tech JSC, and Keldysh Research Centre JSC. Restrictions on these entities should be aligned primarily with the United States, the European Union, the United Kingdom, and Canada. The EU's 21st sanctions package strengthened export controls targeting Russia's military-industrial complex and third-country suppliers, but the Council of the EU's official announcement does not identify Bureau 1440 as a separate sanctions target.
Materik LLC warrants separate scrutiny. According to the Centre for Defence Reforms, Chinese SMA-D-6 attitude-control reaction wheels were supplied through the company. At this stage, it is more accurate to treat Materik as a candidate for enhanced due diligence, export controls, or sanctions for facilitating circumvention than to present its status as established. Likewise, the appearance of equipment made by Rohde & Schwarz, Keysight, or TOPTICA in a Russian supply chain does not prove the manufacturers' knowing involvement: sanctions should focus on intermediaries, importers, and re-export channels.
The diplomatic sanctions track is also moving. On 2 September, Foreign Minister Andrii Sybiha urged EU counterparts to impose sanctions on the Rassvet system and Roscosmos. Ukraine subsequently supplied partners with a proposal covering nine legal entities linked to Rassvet, according to reporting based on official sources. The names have not been made public and cannot be assumed to coincide with UISC's candidate list above; the government package should therefore be treated separately pending disclosure.
The coming months will reveal more about the project's actual viability than high-profile statements by Russia's leadership. The key indicators will be:
· the interval between subsequent launches, the actual number of spacecraft per rocket, and compliance with the state schedule;
· the proportion of second-batch satellites that resume orbit raising towards the observed 509–514 km cluster — or any officially confirmed target shell — and the number entering sustained orbital decay;
· the ITU Radiocommunication Bureau's response to Ukraine's note, the positions taken by partners, and changes to Russian frequency filings;
· independently published results from production-series communications tests, including throughput, session duration, seamless satellite handover, service availability, and resistance to interference;
· mass production of mobile and airborne terminals, the connection of 500 MegaFon base stations, and integration with other operators;
· the appearance of terminals or characteristic emissions within Russian military units;
· the publication and implementation of Ukraine's proposed sanctions package covering nine legal entities, and procurement of critical components through China, Türkiye, Belarus, and Kazakhstan.
UISC's February conclusion has largely survived scrutiny: Russia had prepared for the loss of Starlink, but by 7 September it had not independently demonstrated a stable replacement. Rassvet is no longer merely a plan — two production launches have taken place, the company reports that service testing has begun, and component manufacturing and integration with telecommunications operators are under way. Yet the orbital results of the second batch show that scaling the system is more difficult than simply launching new spacecraft.
One production spacecraft has been lost. As of 6 September, none of the 16 satellites in the second batch had risen above about 426 km, two remained in very low orbits, and only 13 of the 32 launched spacecraft had joined the observed 509–514 km cluster. Against this background, the state schedule for 2026 looks increasingly unrealistic. Company announcements show a transition to testing, but no independently published results yet demonstrate production-series broadband performance, seamless handovers, mass production of mobile or airborne terminals, or systematic combat use. Claims of technological parity with Starlink therefore remain unsupported.
For some military tasks, however, Russia may need far less than a global Starlink: a few predictable communications sessions, a limited number of terminals, and appropriately selected platforms. Ukraine has already challenged Rassvet's declared service areas through the ITU and proposed sanctions against entities linked to the system. These steps should be supplemented by orbital monitoring, signals intelligence, efforts to locate ground infrastructure, targeted export controls, and disruption of critical component supplies. Russia has not yet created its own Starlink, but it is building a system that cannot be assessed solely against the criterion of global coverage.
English
Українська