RedClara built a continental backbone through federation, not full ownership

Summary

• RedClara is a non-profit association based in Uruguay that unites national research and education networks; it is not a shareholder-owned company nor a carrier that controls the connection of every university.
• Eight core routing nodes and long-term international capacity form a 100G-class backbone, with declared 100, 200 and 300 Gbps links and a smaller segment in Central America.
• ALICE and ALICE2 created the institution and the first network; then BELLA secured long-term capacity on the direct route between Portugal and Brazil and reduced dependence on indirect routes through North America.
• BELLA II’s test is converting project funding into permanent member-supported infrastructure, while strengthening weaker national networks, improving evidence, and preventing supplier or membership bottlenecks.

BELLA changed the network geography

BELLA, or Building the Europe Link to Latin America, combined a submarine segment and terrestrial upgrades. BELLA-S secured long-term optical spectrum on a direct cable between Portugal and Brazil, and BELLA-T bolstered the terrestrial routes that carry capacity from the landing point to regional nodes and national networks.

The EllaLink system, around 6,000 km between Sines and Fortaleza, came into service for research and education traffic in 2021, with two initial 100 Gbps links, one between GÉANT and RedClara and the other for Copernicus. The 35,830 km and Nx100G figure in a 2024 report describes the wider BELLA architecture, including terrestrial and international components.

Lengths and costs measure different scopes. CORDIS records around €12.4447 million for the BELLA-S1 action, including €10 million from the EU, while other figures aggregate terrestrial works, additional funding lines, and national contributions. They should not be blended without explanation.

Programme materials mentioned latency reductions of up to 50% on certain paths. That must be kept path-specific and attributed. The enduring gain is a direct South Atlantic route and rights that can scale as optics and demand evolve.

Forty optical channels do not mean cable ownership

BELLA acquired long-term rights to forty optical channels on EllaLink. A spectrum right gives control over operating and upgrading the capacity, but it does not make RedClara or GÉANT exclusive owners of the commercial cable, landing stations, or supporting facilities.

The chain extends from the fibre pair to the spectrum right, then channels, then 100G or higher transport, then terrestrial networks. The cable operator maintains the submarine system, other parties manage the landing and hosting, RedClara and GÉANT coordinate the academic capacity, suppliers provide the equipment, and the NRENs carry the traffic to the institutions.

Spectrum rights may be more valuable than a fixed circuit because capacity can grow by changing equipment within optical and contractual limits. But they impose shared governance, repair, power, spectrum planning, and interoperable operations. The right is long-term, but it does not operate itself.

This boundary is fundamental. RedClara can describe its regional backbone and transatlantic path because it coordinates and operates the capacity, but it should not be described as owning every line on the map. Its strength comes from the federation and contractual control of strategic capacity, not from complete physical integration.

BELLA and the direct South Atlantic route

The clearest geographical change BELLA made. Before the direct path, traffic between Latin America and Europe often transited North America even when both ends sat on the South Atlantic. The Sines–Fortaleza route created a shorter alternative for many cases and a long-term anchor for the research community.

That does not mean latency halves everywhere. The outcome varies by source, destination, national access, congestion and return path. The “up to 50%” wording must remain attributed to specific routes. The stronger result is path diversity: the European relationship no longer depends exclusively on northern gateways.

This benefits large data transfers, astronomy, Earth observation, distributed computing, academic telehealth and interactive collaboration. It also raises RedClara’s and GÉANT’s value as interconnection platforms that offer more predictable capacity than the public internet.

The direct cable remains exposed to cuts, power, stations, equipment and maintenance. Resilience requires alternative paths, failover tests and sufficient capacity via Miami or elsewhere when EllaLink is disrupted. “Direct” improves architecture but does not remove the need for redundancy.

BELLA was also a long-term institutional relationship between Europe and Latin America. The capacity came as part of a science cooperation programme, not an ordinary commercial transit purchase, giving the community a larger role in use and upgrade while commercial suppliers remained critical.

Spectrum rights, terrestrial access and ownership limits

The phrase “RedClara network” can hide many forms of control. The organisation may own routers, lease racks and power, buy managed service, hold IRUs, use spectrum on someone else’s cable, or exchange capacity. Each form has different duration, maintenance, upgrade rights and risk.

At sea, BELLA secured long-term EllaLink spectrum. On land, BELLA-T and the members assembled fibres, wavelengths and national services to reach points of presence and institutions. The 35,830 km figure in a 2024 report represents a broader programme scope; it does not prove RedClara owns every kilometre.

This avoids two errors: minimising the strategic value of spectrum rights or exaggerating that RedClara single-handedly controls the trenching, landing stations, repair ships, equipment and every national node.

The real resilience lies in the contracts: the right’s duration, who pays maintenance, whether upgrades are possible, what happens if the operator is sold or insolvent, station access, exit rights and alternatives. A line on a map means little without these terms.

The federation model enables broad reach without owning every physical asset, but it increases the need for supplier management and transparency. One must know which dependencies are replaceable and which have become hard-to-reverse points.

A continental network is not a single, operator-controlled network

RedClara is easily misunderstood when described simply as “the Latin American research network”. The phrase suggests a single operator controlling the path from the international backbone to the university’s network port. The reality is a federation. A researcher normally reaches RedClara through their institution’s network and then through the national research and education network, or NREN. RedClara provides the regional and intercontinental layer that makes these independently administered systems work within a wider scientific environment.

This distinction defines both the organisation’s value and its limits. RNP in Brazil, REUNA in Chile, RENATA in Colombia, CEDIA in Ecuador, CUDI in Mexico, RedCONARE in Costa Rica and RAU in Uruguay each retain their national governance, local infrastructure, institutional relationships and operational responsibilities. RedClara does not replace them; it connects them, represents their shared requirements, coordinates capacity and services, and opens the path to peer networks such as GÉANT, Internet2 and CANARIE.

The full path may cross the campus network, an institutional firewall, a national access circuit, the NREN backbone, RedClara, an international peer, the destination NREN and then a remote research facility. A 100 Gbps regional link cannot compensate for a bottleneck at the campus edge, a weak national link, poor storage configuration or destination policies. Equally, a single national network cannot negotiate every intercontinental path, identity relationship and collaboration service that modern science needs.

Therefore, the more accurate description is that RedClara is a shared coordination and infrastructure layer. “Shared” means governance, engineering, long-term capacity rights and common services, not ownership of every fibre, router, point of presence or cable segment on the map. This is not a linguistic note; it is the central mechanism by which the organisation transforms multiple national networks and commercial infrastructure contracts into a regional network serving a public purpose.

Why Latin America needed a regional layer

The original problem was fragmentation. In the early 2000s, several countries had national academic networks or nascent ones, but they lacked a permanent continental institution to aggregate demand, operate a regional backbone and negotiate on equal terms with European and North American networks. International circuits were expensive, national markets differed, and commercial routes were designed around telecoms economics rather than the needs of distributed science.

This imposed a practical price. A university might have good national connectivity but still rely on costly, indirect international transit to collaborate with another institution in the region. Europe–Latin America traffic could traverse North America because the cable market and commercial interconnect made that path easier to buy. The result was not just longer delay but limited bargaining power, less route diversity and reliance on services whose priorities lie outside the research community.

A centralised continental operator was not a realistic solution. The region varies in NREN maturity, public funding, regulation, fibre availability, university systems and cable landing points. National institutions had to keep their legal and operational independence. So the viable model was a federation: national roles stay inside, and a regional association looks after the shared layer and the relationship with global partners.

The design also produced political and institutional value. One NREN might struggle to influence an international research programme or justify a specialised path. A regional organisation can aggregate use cases, pool contributions, blend grants with national assets, and act as a single counterpart for the European Commission, GÉANT, development banks, telecoms companies and cable operators. RedClara was created to turn that shared need into an institution and a working network.

Toledo, Valle de Bravo and the legal association

The idea crystallised in 2002. Representatives of Latin American NRENs discussed the feasibility of a regional network at the CAESAR meeting in Toledo, Spain, then continued work in Rio de Janeiro, Buenos Aires, Santiago and Mexico. The question was not only topology; it was about who the entities were, how the regional level would relate to national networks, how decisions would be made, and how a project funded externally could become a permanent organisation.

The legal timeline contains three dates worth preserving. RedClara states that the statutes were signed at Valle de Bravo, Mexico, on 9 June 2003. An Organisation of American States civil-society filing records 10 June as the date of establishment. The organisation also says its legal existence was recognised under Uruguayan law on 23 December 2003. It is best to understand them as related but distinct events: a founding agreement, an administrative filing, then later legal recognition.

The strongest description is that it is an international non-profit civil association recognised under Uruguayan law and based in Montevideo. The phrase “international law organisation” is sometimes used in English, but the records do not show it was created by treaty or enjoys intergovernmental status. It is a membership association with regional aims, not a supranational regulator or a government agency.

The current statutes were approved in November 2019 and published in January 2020. They make the General Assembly the highest authority, define a board of directors, a fiscal oversight committee, a technical committee that pools NREN expertise, and an executive directorate for implementation. Legitimacy comes from membership and cooperation, not from shares or regional authority.

Seven legal members, eleven connected countries and a wider mission

The current public membership directory lists seven NRENs: RNP, RENATA, RedCONARE, REUNA, CEDIA, CUDI and RAU. By contrast, the advanced connectivity page says the network connects eleven countries: Argentina, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Mexico, Paraguay and Uruguay. Older materials mention thirteen countries. These are not interchangeable metrics.

Legal membership determines governance rights, obligations and participation in the association. Operational connectivity may be provided through a partner, a transitional arrangement or a project even if the national network is not a full voting member. Older numbers may reflect previous memberships or historical programme scopes. The distinction between seven current legal members and eleven connected countries is therefore meaningful, and older figures should be dated.

The gap creates a strategic tension. RedClara aims to serve a region larger than the association that formally governs it. A country may benefit from connectivity or from BELLA II without having a matching seat in the General Assembly. In return, the seven members bear a greater share of governance, technical work and possibly recurrent funding. This does not prove failure, but it affects representation, cost-sharing and the project of building a genuinely regional platform.

The 2025–2027 board reflects the active membership base: José Palacios of REUNA as president, Eduardo Grizendi of RNP as vice-president, Juan Pablo Carvallo of CEDIA as treasurer, Carlos Gamboa of RedCONARE as secretary and Eduardo Grampín of RAU as member, while Luis Eliécer Cadenas serves as executive director. The structure gives national networks formal control, but the public record does not provide recent general assembly minutes, voting weights, current fees or a full explanation of how connected non-member countries participate.

ALICE provided the initial capital for the regional network

The legal creation and the first production network were tied to the ALICE project (America Latina Interconectada Con Europa). The European Commission and DANTE signed the contract in June 2003 with a budget of €12.5 million, €10 million from the Commission and €2.5 million from Latin American partners. It continued, after extensions, until March 2008.

ALICE solved several start-up problems simultaneously: it supplied funding, created joint procurement, gave the new association an immediate operational mission and connected the region to GÉANT. This made it possible to prove the network’s value while the fee model, revenue and the institution itself were still taking shape.

The early infrastructure was modest compared with today’s, but the institutional impact was large. It established points of presence and production links between NRENs, connected the region to European and North American networks, and opened relationships that later extended into identity, middleware, grid computing, open science and joint projects. A working network made the association more than a forum.

ALICE also introduced a hard question: what happens when a temporary grant funds infrastructure that must operate indefinitely? The contracts, equipment, teams and expectations do not vanish with the final report. RedClara’s subsequent history can be read as a search for longer-horizon rights—IRUs, fibres, wavelengths, spectrum, services and co-investment—to sustain the institution between funding cycles.

ALICE2 moved coordination to Latin America

ALICE2 started in December 2008 with an €18 million European Commission contract. The most important institutional shift was that RedClara became the coordinator, rather than simply participating in a Europe-led programme. Operational leadership and project management transferred to the Latin American association.

The programme explicitly targeted sustainability: lower recurrent cost through dark fibre, wavelengths and IRUs; expanded coverage; strengthened NRENs; training; user communities; and more stable financing. Sources differ between an end in 2012 and an administrative closure in January 2013. The safest wording is that operational activity ended before the final administrative close.

ALICE2 also broadened the concept of infrastructure. Grid computing projects linked regional resources to international collaboration; identity and middleware helped access shared services; training recognised that a fibre path yields little if national networks lack staff, governance and applications.

RedClara moved from a young transport infrastructure to a regional e-infrastructure organisation. Dependence on external funding remained, but it gained experience coordinating large projects, managing multi-country technical work and turning transport into a services platform.

From short-term circuits to long-term capacity rights

Network economics depends on how capacity is bought. A short-term managed circuit provides a specific service but exposes the buyer to repricing and limits upgrade control. A long-term right in a fibre, wavelength or spectrum requires more investment and governance, but it allows planning and transport upgrades without buying each increment anew.

RedClara’s strategy moved towards IRUs, dark fibre, wavelengths, optical channels and capacity exchanges wherever possible. It did not become a traditional owner of all cables, but it acquired rights that can outlast a technology generation. 2015 reports linked this to lower variable cost and greater sustainability.

Long rights also change the risk profile. They reduce repeated procurement but increase dependence on the cable, the landing station, the operator, maintenance and the contract. Spectrum is worthless if the system is not repaired, operated and powered. Optical equipment, hosting, terrestrial extensions and skilled teams remain necessary.

The asset base cannot therefore be drawn as a simple ownership map. Some fibre is owned by NRENs, some is leased, contributed in kind, protected by IRUs or represented by optical rights inside a shared cable. The regional platform is made of legal and technical relationships as much as it is made of physical assets.

Before BELLA, reaching Europe often meant transiting North America

The absence of a direct, high-capacity Europe–Latin America path affected both performance and dependence. Even between institutions on opposite sides of the South Atlantic, traffic could pass through commercial systems in North America. That might make sense for carriers, but it added distance, concentrated transit and reduced the community’s control over capacity.

North American connectivity remained essential. Initiatives such as WHREN/LILA strengthened routes through Tijuana–San Diego and Pacific Wave, and Miami became a gateway to Internet2, CANARIE and others. These routes provide reach and diversity.

The problem was over-dependence, not the existence of the routes. A continental network needs alternatives. A direct link to Europe can lower latency for some pairs, diversify failure, carry large scientific flows without forcing them through a single corridor, and match physical geography to the long-standing RedClara–GÉANT collaboration.

With Earth observation imagery, genomics, telescope data, climate models and HPC workloads, the need grew. BELLA turned the direct path from an ambition into a long-term physical and contractual system.

Current topology: eight nodes and multiple capacity types

Current technical documents name eight core routing nodes: Buenos Aires, Fortaleza, Porto Alegre, São Paulo, Santiago, Manta, Panama City and Miami. South American networks connect to one of these in a 100 Gbps or higher ring, while the shared infrastructure in Central America is described as 20 Gbps. Fortaleza represents the primary direct route to GÉANT, Europe and Africa; Miami represents the gateway to Internet2, CANARIE and North America.

The advanced connectivity page also mentions 100, 200 and 300 Gbps links, eleven countries and more than 1,300 institutions. Yet the site still carries legacy language about 10 Gbps. The weight of the evidence supports a 100G-class description; the 10G text should be treated as a documentation debt, not a snapshot of the current state.

The speeds do not mean every path offers the same usable capacity. Line speed differs from protected, reserved or actually delivered member capacity. A logical ring may share commercial ducts and fibres, so not every alternative is physically separate.

Manta anchors the Pacific side, Fortaleza ties the South Atlantic, while Panama and Miami serve Central America, the Caribbean and North America. The map is shaped by landing stations, national markets, NREN partnerships and available fibres, not by a symmetrical engineering ring.

The “more than 1,300 institutions” is a reach indicator without a published methodology. It does not show how many are active, at what speed, or using which services. Maturity should be measured by capacity, traffic, availability and recurrent usage, not by one aggregate number.

AS27750, peering and traffic class separation

AS27750 is RedClara’s public autonomous system number. The organisation uses it to carry or exchange member routes, research and education networks and international services. Public databases show relationships with GÉANT, ESnet, Internet2, CANARIE, RNP, REUNA, RENATA, RedCONARE, CEDIA and TENET, but they are snapshots, not a complete contractual list.

“Advanced internet” does not simply mean a faster public internet. Eligible research traffic can take dedicated or private paths to scientific networks and facilities, avoiding commercial transit where a direct relationship exists. This improves cost, predictability and preserves capacity for research uses.

RedClara also provides separation of commercial internet, peering and global connectivity. A university needs both academic resources and the public internet, and policies, suppliers and billing may differ between the two categories. The research network is not physically isolated; it is distinguished by eligibility, routing, capacity and governance.

The selective peering policy allows direct connection where there is mutual interest, but it does not grant every network automatic acceptance. Even when present in the same facility, ports, cross-connects, filters, policies and ongoing operations are still needed.

The organisation does not publish a complete, up-to-date picture of RPKI validation status, ROA coverage, IRR filtering, route-leak protection, BGP communities and per-peer availability. More curated public evidence could make the regional infrastructure’s resilience assessable without revealing sensitive security details.

Layer‑2 circuits, operations and the limits of regional responsibility

Services include dedicated layer‑2 virtual circuits. A VLAN can connect laboratories, data centres or scientific instruments in different countries as though they were on the same Ethernet segment, which is useful for large data, distributed experiments and needs that standard routing does not serve well.

The circuit crosses several administrative domains. The institution provides the local hand‑off; the NREN carries the traffic to the regional network; RedClara coordinates the international segment; and the process repeats at the destination. VLAN, MTU, redundancy, capacity, addressing, security and change windows must all be aligned.

Operational responsibility is distributed accordingly. A fault may lie in the campus, the NREN, RedClara, a commercial carrier or the far‑end institution. The regional operations centre needs clear circuit identification, demarcation points, 24/7 contacts and an escalation path; otherwise the user does not know who can fix it.

RedClara describes engineering groups, a NOC and a security function, but it does not publish full SLA methodology, historical repair times or an independent incident database. Older pages mention a monthly average availability of 100% and annual traffic of 7.7 PB without scope or method; they should not be treated as an audit.

In the federation model, RedClara coordinates the regional piece and works with the national networks, but it does not guarantee every component alone. Professional service requires clear responsibility boundaries and a shared support chain rather than one that shifts blame.

BELLA II: expansion through co‑investment, not a completed map

RedClara and the European Commission signed BELLA II on 14 December 2022. The programme was designed for 48 months within a €28 million plan, with the Commission contributing €13 million and RedClara expected to mobilise at least €15 million from governments, development agencies, companies, banks and other partners.

Objectives include connecting or strengthening Peru, Costa Rica, El Salvador, Guatemala and Honduras, with possible extension to Mexico, the Caribbean, Paraguay, Uruguay, Bolivia and Belize. Those names represent targets and dialogues; they do not prove that every path was in production on 3 August 2026.

The 2025 competitive dialogue called for a fibre system with a minimum fifteen‑year lifespan, initial capacity of at least 20 Gbps in Central America and 100 Gbps in Peru, with scalability. Offers can combine existing fibre, new build, capacity exchange or co‑investment.

Flexibility attracts investment but complicates accountability. One must identify who builds, owns and maintains, what capacity is reserved for research, how a country joins, and who funds operations after the grant. Fifteen‑year rights are not sustainable unless power, equipment, repair and upgrades are funded.

BELLA II should be assessed by accepted, operational paths, enforceable rights, live traffic and post‑project arrangements. Memoranda and dialogues are real progress, but they are not a service in operation.

From a transport network to a regional research platform

The first generation proved that national research networks could exchange traffic over a regional backbone. The organisation then tried to answer a broader question: what shared capabilities could be built on top of it? The portfolio expanded to identity, file transfer, collaboration environments, training, HPC access, open science, security and thematic communities.

This is not a departure from networking; it raises its value. A fast link loses much of its impact if a researcher cannot log in, find a supercomputer, move data, find partners or obtain support. The application and community layers turn raw capacity into usable research infrastructure.

BELLA II embodies this logic through innovation hubs, testbeds, open calls, training, dialogues and mechanisms for partners and funders. The 2024 Ideathon on food and agriculture brought together about 125 experts from 28 countries, and early HPC platforms linked training and prototyping to shared resources.

A testbed is not a production service. It often serves a selected group with limited capacity, specialised support and an experimental goal. It reveals barriers before permanent investment, but it can mislead if a successful experiment is presented as general availability.

The opportunity is for RedClara to become a federation of services and experimentation, not only of transport. But identity, storage, computing, notebooks and sensitive data add security, privacy, support and lifecycle responsibilities. NRENs, SCALAC, universities, clouds and partners continue to run the resources, while RedClara handles integration and the regional gateway.

BIO+IA and the discipline of early‑adopter evidence

The BIO+IA Early Adopters programme is a clear example of the new role. Opened in 2026 under BELLA II, it selected five projects from six countries for training, support and access to an AI and bioinformatics environment. Projects addressed breast cancer biology, molecular screening, antimicrobial or mercury‑resistance resistance, liver‑disease drug repurposing and translational medicine.

At the July 2026 demo day, teams said the environment helped them search, integrate and interpret biological information. This is important evidence of facilitated research workflows and of the network’s ability to combine connectivity, tools and expertise at a regional level.

It does not prove clinical efficacy. Finding a candidate molecule does not prove a treatment works; accelerating an analysis does not replace laboratory work, validation, regulation and health‑system application. The initiative served a limited cohort, and the results were entity and organiser reports, not an independent clinical trial.

These limits do not diminish the programme; they define the correct unit of achievement. RedClara supported a multi‑country experiment, lowered barriers to tools and knowledge, and created relationships between research teams.

The next test is continuity. An experiment becomes infrastructure when teams can return, new groups join, data policies become clear, computing and storage funding continues, and successful paths move into national or institutional systems. Repeated use, publications, independent evaluation and post‑experiment funding are the critical signals.

The wider lesson is to measure innovation programmes by the distance between participation and permanent capability. Ideathons can count entities; testbeds can count trials; but infrastructure must also count environments that stay maintained, returning users, governed data, trained operators and projects that continue after the grant or event ends.

Identity as infrastructure: FIEL, eduGAIN and eduroam

A network link is useless if a researcher cannot prove their identity to a remote service. FIEL supports national identity federations; eduGAIN allows metadata exchange between federations and the use of an institutional identity to access international services; eduroam applies a similar logic to network access roaming.

These systems do not create a regional password database. In federated login, the service redirects the user to their home institution, which authenticates them and sends a signed assertion or selected attributes; the service then applies its access policy. RedClara strengthens federations and links them to the international trust fabric; it normally does not receive passwords.

Trust and risk are distributed. The university retains responsibility for its accounts; the researcher uses one identity across repositories, platforms and tools. But assurance levels vary, metadata can be outdated, attribute meanings may differ, and a compromise of an identity provider or signing key can affect many services.

eduGAIN solves federation discovery and metadata exchange, not every legal question. The service chooses which institutions and assurance levels it accepts; the institution protects keys, manages the account lifecycle and responds to incidents. Cross‑border use raises questions of consent, attribute minimisation, privacy and legal basis.

In eduroam, the visited institution provides connectivity and sends the authentication request through a RADIUS chain to the home institution. Security depends on device configuration, certificate validation, the source identity system and local network controls.

Identity is part of the infrastructure continuity. A supercomputer may be IP‑reachable but inaccessible to a user if the trust path fails. By coordinating identity and connectivity together, RedClara addresses a failure that a pure transport network ignores.

MiLab, eNVIO and the research collaboration layer

MiLab assembles open tools behind federated access: a data environment built on Dataverse, development through GitLab, communication through Mattermost, and computing notebooks linked to JupyterHub, Python and R. It can support a team from communication and coding to analysis, data organisation and publishing.

Distributed teams normally assemble these functions from separate services with different accounts and support policies. A shared regional space reduces that fragmentation and offers one environment to multiple institutions and countries. RedClara has documented research uses involving geographically distributed data.

MiLab should not be described as a public regional cloud. It does not automatically provide large‑scale HPC, unlimited storage, certification for all sensitive data or permanent legal preservation. Capacity, service level and hosting can change, and they must be matched to project requirements.

eNVIO handles large file transfer over the academic infrastructure. Much scientific data does not fit email or consumer services. The value depends on size limits, retention, encryption, support and integration—indicators that are not fully published.

RedClara also runs training platforms and a learning system with CEDIA. Training turns knowledge into courses and certificates, but it is not a substitute for permanent posts, operational experience and institutional responsibility.

This layer sums up the federation model: RedClara integrates tools and a portal, NRENs bring users, institutions provide identity and context, and researchers do the work. The system works when responsibilities are clear, and it cannot compensate for every local shortfall in staff, data governance and support.

Ventanilla Abierta: navigating a fragmented landscape

Research infrastructure can be hard to discover even when it exists. A researcher may know that HPC, storage, federated identity or a specialised network is available somewhere, but may not know who operates it, whether their project qualifies, or how to start a request. Ventanilla Abierta was created as a single advisory gateway for this problem.

The service does not own every resource to which it refers. It translates a research need into a path: connectivity through an NREN, computing through SCALAC or another centre, data and storage services, federated identity, an international partner’s resource, or a project and funding route. Its value lies in coordination and discoverability.

This makes it navigation infrastructure. Capacity can remain unused because institutional interfaces are fragmented, not because demand is absent. If a researcher must understand several separate networks, portals and teams, they may abandon the regional option. A trusted gateway reduces the transaction cost.

It also addresses the need for calls, complementary institutions and funding. Connecting a technical requirement to a partner or funder can turn a latent idea into a project that can use the network.

The authority is limited, however. A referral is not a reservation; a call is not a grant; an introduction is not a federation agreement. The service should be assessed by successful transitions and ongoing projects, not by the number of enquiries alone.

LA Referencia, SCALAC and the shared knowledge and computing chain

RedClara’s relationships with open science and HPC show how transport becomes an ecosystem. LA Referencia aggregates metadata and outputs from national repositories; SCALAC connects advanced computing institutions. Both are independent alliances, not subsidiaries, but the regional network and coordination enable cross‑border operation.

The open‑science chain starts at the institution or national repository. Researchers deposit publications or data according to local policies; national systems organise metadata and access; LA Referencia adds regional discovery and interoperability; and relationships such as OpenAIRE open the international dimension. RedClara provides the connectivity and coordination; it does not own every repository or result.

Open science needs more than bandwidth. Metadata quality, persistent identifiers, licences, preservation, governance and incentives determine whether outputs are findable and reusable. A fast network moves the file, but it does not make undocumented data reproducible.

SCALAC addresses a similar computing problem. HPC resources are spread across universities and national centres. The alliance makes them visible, coordinates expertise and enables testing and training for projects that are hard to support in a single country.

An HPC federation does not mean a single supercomputer. Schedulers, software, access policies, queues, storage and accounting differ; data must be staged, users authenticated and quotas governed; sensitive workloads may face cross‑border constraints. The network removes one bottleneck and reveals others.

The benefit is choice: a country that cannot fund every category of computing can share; a facility with spare capacity can attract collaboration. The risk is that access stays episodic, project‑based and dependent on personal relationships rather than a predictable service with published rules.

Health, Earth observation and the communities that give the network purpose

RUTE-ALC brings together university telehealth and digital health communities. It can support specialist consultation, medical education, image exchange, standards, emergency collaboration and digital health governance.

The research network provides trusted paths, identity and collaboration tools, but it does not become a healthcare provider. Clinical privacy, device regulation, professional liability and patient consent remain with the institutions and authorities. A high‑speed link is an enabler of telemedicine, not a complete clinical service.

Copernicus Academy forms another bridge. Earth observation data can support agriculture, disasters, climate, oceans, water, forests, cities and energy. RedClara helps move the data and organise training, but the constraint is often analytical capacity, not raw imagery access alone.

TICAL brings together higher‑education technology leaders and technical communities. The 2024 Rio de Janeiro event recorded more than 200 entities. Meetings are not fibre, but they help align procurement, security, cloud, identity and services among institutions that will need to interoperate.

These communities explain why a research network is not just a discounted telecoms company. Repeated relationships build trust that speeds up incidents and projects. But attendance does not prove implementation; a workshop does not automatically build national capability; connectivity does not prove a health or climate outcome. RedClara enables and coordinates the conditions; it does not own every result.

EduLACSeg and the regional cybersecurity trust network

Academic institutions face similar campaigns across borders, so security coordination is a natural extension of the federation. eduLACSeg provides a trust framework for sharing indicators, context and response expertise, so that a campaign is seen once and multiple countries are warned before each discovers it independently.

In July 2026, threat intelligence sourced from CERN was passed through RedClara to RNP, REUNA, RENATA, RedCONARE, RAU, CUDI and CEDIA. This is concrete evidence of regional security intelligence transfer. Describing it as the largest academic security ecosystem in Latin America is the organisation’s own characterisation, as no independent comparative category is available.

Indicators can be false, outdated or context‑dependent; attribution may be uncertain; data and confidentiality laws differ; and capabilities vary between a mature security team and a small group with limited tools. An alert demonstrates distribution, not a uniform response.

A 2024 report records training with LAC4 for 85 entities from about 60 organisations in 18 countries, 30% of whom were women, plus a Security Baseline Bootcamp with RNP and GÉANT at TICAL2024. These are capacity‑building signals, not evidence that all members have applied the same controls.

RedClara’s own routing sessions, identity data, platforms and testbeds must also be protected. There is no complete, current public picture of RPKI, ROA, IRR filtering, incident disclosure and response times. Trust requires rules for classification, redistribution, data protection, attribution and post‑incident learning, as well as measurable technical controls.

SUBMERSE: the network as a scientific instrument

SUBMERSE explored how operational telecommunications fibres can be turned into sensors. Distributed acoustic sensing techniques send light pulses and analyse backscatter to detect tiny disturbances along the fibre linked to earthquakes, ocean movements, traffic or other phenomena.

The idea exploits existing infrastructure that crosses areas where deploying dedicated sensors is expensive. Research networks can connect scientists with cable owners, landing stations, data and computing.

But the fibre does not automatically become an observatory. It requires instruments, access, calibration, time synchronisation, storage, processing and scientific validation. The measurement must not interfere with production traffic, and the data may reveal sensitive activity, requiring governance.

RedClara was a member of a larger consortium, not the owner of all fibre or the sole method owner. The main funded phase ended with closing events in 2026. SUBMERSE therefore represents a proof‑of‑concept and a direction for collaboration, not a permanent, operational regional service.

If later phases secure rights, standards, funding and users, fibre‑based sensing could give RedClara a distinctive role in geophysics, oceanography and resilience. The network would not just carry data; it would generate data from its own infrastructure.

An ecosystem of partners, not a group of subsidiaries

RedClara’s relationships include NRENs, the European Commission, GÉANT, Internet2, CANARIE, development banks, EllaLink, telecoms companies, SCALAC, LA Referencia, RUTE-ALC, LNet/LACChain, universities and communities. These relationships do not share the same legal or operational nature.

A NREN may be a member and an operational partner. GÉANT and Internet2 are peer networks. The Commission is a funder and programme partner. EllaLink is commercial infrastructure carrying capacity rights. SCALAC, LA Referencia and RUTE-ALC have their own governance. A university may be a user, host, beneficiary or expertise provider.

Labelling all of this “RedClara” overstates central control and blurs responsibility. The association’s value lies in coordinating assets and institutions it does not own. One must distinguish membership, funding, capacity, interoperability, execution, collaboration and project participation.

The ecosystem prevents single‑owner dependence, but it can also scatter accountability. When something breaks, the user should know whether the cause lies in the institution, the NREN, RedClara, a carrier, a platform or an international partner. A federation is strongest when its boundaries are visible.

Funding is built on membership, projects and capacity rights

RedClara has no documented shares, market capitalisation or distributable profits. The statutes allow membership fees and other resources; resources also come from consultancy, services, European projects, development agencies, international cooperation, partner contributions, co‑investment, capacity exchanges and in‑kind contributions.

The headline numbers are programme budgets, not organisational revenue. ALICE was about €12.5 million, ALICE2 about €18 million, BELLA had multiple actions, and BELLA II was planned at €28 million with €13 million from the EU and at least €15 million to mobilise. They must not be summed and labelled as revenue.

Significant value sits in spectrum rights, IRUs, national fibre, hosting, equipment and people. It is measured by duration, maintenance, upgrade rights and restrictions, not solely by the nominal project value.

The most recent annual report provides programme and staffing detail, but it does not give a complete, current set of audited accounts with balance sheet, reserves, fee concentration and supplier exposure, and no precise public headcount is available. That limits sustainability analysis.

The core risk is time misalignment: a grant lasts for years, while the network, identity and platforms must operate without a defined end. Long‑term rights preserve part of the value, but power, equipment, support and people need recurring income. Sustainability requires fees, cost recovery, national support and partnerships, without treating a grant as permanent income.

The last mile is national, institutional and uneven

The continental backbone may be advanced while benefits are distributed unevenly. Countries differ in local fibre, NREN funding, government support, competition, landing stations, universities and skills. A central university may have excellent connectivity, while a remote institution relies on a weak link.

RedClara can aggregate demand, negotiate, train and demonstrate the value of a national network, but it cannot substitute for national investment, remove every regulatory obstacle or operate every access link. The last mile remains with institutions, NRENs and national providers.

A “connected country” therefore does not imply uniform quality for all universities. The 1,300‑institution figure indicates reach without revealing speed, availability or usage. Mature metrics should separate registered institutions, active institutions, connection speeds and users of regional services.

BELLA II can extend reach while also creating obligations in places with weak operational capacity. Sustainable participation needs local engineering, budget, support, security and identity. An international link can become a high‑speed island if a strong NREN is absent.

Documentation debt obscures operational maturity

The public record is rich in history and projects, but it mixes seven members, eleven countries and older language about thirteen, modern 100G descriptions alongside 10G text, and institution, traffic and availability figures without consistent methodology.

The most likely explanation is that pages from different eras remain published. That does not invalidate the modern network, but it demands source, date and definition checks. In infrastructure, an outdated document can become an operational risk, not just an editorial one.

A map should state the date, contracted capacity, usable capacity, protection, operator, operational status and the type of right. Traffic needs a period, direction and aggregation method; availability needs a scope and formula; membership must separate governance, connectivity and project status.

Consistent publication helps members, funders and users, and prevents memoranda and proposed paths from being read as live connections. Transparency does not require revealing secrets; it requires defining what is being described precisely.

RedClara’s achievement is federation at infrastructure scale

The central achievement is not a cable, a router or a project; it is a permanent regional mechanism that enables independent national networks to aggregate demand, operate interconnectivity, secure long‑term rights, participate globally and build shared services while retaining their national governance.

ALICE turned the idea into a network; ALICE2 moved coordination to the region and sought sustainability; BELLA changed the geography with the direct path and spectrum; BELLA II is attempting to expand access and turn transport into identity, open science, computing, security and experiments.

The federation also defines the limits. RedClara does not own every asset, does not control every NREN, does not guarantee every last mile and does not have unlimited resources. Its strength is coordination: strong when the incentives of members, funders, peer networks and suppliers align, and strained when responsibility is ambiguous or national capacity is thin.

Future significance will depend less on a new speed number and more on the quality of the institution around the capacity: wider membership, long‑term contracts, financial transparency, measurable security, service lifecycles and a clear path from trial to production. The backbone is the foundation; the federal institution is the strategic asset.

Indicators that BELLA II is turning into permanent infrastructure

The first indicator is actual operation. Memoranda, dialogues and procurement requirements must become named suppliers, accepted paths, production traffic, fault procedures and dates. The initial minimum must be separated from line speed, protected and usable capacity. A fifteen‑year ambition does not help if maintenance, upgrades and operations are not sustained.

The second is membership. Seven members are matched by eleven connected countries. One should track whether target countries become members, permanent partners or institutional entities, and how their status affects voting, fees, rights and obligations. Expanding the network without expanding the governance base increases the burden and keeps the concentration.

The third is operational evidence: dated topology and traffic, SLA methodology, incident disclosure, routing policy, RPKI, and measured diversity between EllaLink, Miami and alternatives. A cable cut or a route leak will test whether the logical ring is physically and politically resilient.

The fourth is service evolution. BIO+IA, HPC, MiLab, identity, eduLACSeg and open science must be measured by returning users, production quotas, support, data policies and post‑grant funding, not by attendance counts. An experiment becomes infrastructure when a user can plan around it.

Finally, the financial picture matters. Audited accounts, reserves, fee concentration, supplier concentration and the ratio of unrestricted to restricted funds improve credibility. The crucial question is the ability to preserve capacity between European funding cycles without leaving weaker networks as temporary beneficiaries.

Evidence‑supported scenarios

In a successful expansion scenario, Peru and the Central American networks enter long‑term, scalable paths, and the governance base widens. The European path, northern gateways and Pacific access support larger instruments and computing.

In a service‑federation scenario, identity, security, data, HPC and testbeds become everyday products, and the connectivity itself becomes less visible. Value rises, and so do software, privacy and support obligations.

In a concentrated‑governance scenario, membership stays at seven while projects expand. The organisation is regionally operational but depends on a small core for representation, cost and succession.

In a project‑cycle scenario, innovations start with grants and falter after closure. The backbone remains while the upper layers fluctuate. That is the greatest risk in the absence of a permanent cost model.

Co‑investment with telecoms companies could improve sustainability through exchange, build and long‑term rights, but it increases dependence on a few suppliers and requires assiduous maintenance, exit and upgrade provisions.

And in the network‑as‑instrument scenario, SUBMERSE lines extend to more fibre and produce environmental and geophysical services. That remains prospective, but it could give RedClara a role that commercial capacity alone does not offer.

Who controls and who bears the risk

Control is distributed. The members govern the association and the board; the directorate and technical teams execute; the national networks control the local infrastructure; the peer networks control their own systems; the funders influence programmes; and the telecoms, cable, hosting and equipment companies hold the physical dependencies. Universities create demand but do not always have a direct regional governance right.

That prevents a single owner, but it separates decision from outcome. A board may approve a programme for which an NREN pays the last‑mile cost; a funder sets a four‑year horizon while members inherit decades of maintenance; a carrier may own the only viable path; and a user may not know who can restore service. Those interfaces are the core control surface.

Incentives should reward permanent contributions: fees, fibre, staff secondment, service operation, data stewardship and security response. Supported countries need a path to national responsibility and recurring commitments; stronger members need assurance that solidarity does not become an unlimited cost transfer without accountability.

Supplier contracts must preserve route diversity, upgrade rights, audit, exit and continuity when ownership or strategy changes. Long‑term rights reduce price risk, but they can become stranded if stations, equipment or maintenance fail. The most important decision is not how much capacity, but which dependencies become hard to reverse.

Decisions that cannot wait for the next grant

First, RedClara must decide whether it will remain primarily a connectivity federation or become a permanent operator of shared platforms. MiLab, identity, security and AI testbeds create value but add data, software, staff and liability. Permanent services need an owner, lifecycle funding, response and orderly deletion.

Second, the gap between legal membership and operational scope should be resolved. Connected countries need status, voice and a path to membership or another responsible arrangement. Informality facilitates expansion but leaves governance, payment and exit ambiguous.

Third, financial transparency must become a governance tool. Accounts, reserves, fees, restricted funds and supplier concentration determine the network’s ability to withstand a funding or cable shock.

Fourth, RedClara needs a published continuity architecture covering route diversity, incident escalation, backups, identity and platform restoration, cable alternatives, and the succession of project‑funded services. Core functions must survive the failure of a grant, a supplier, a path or an institution.

The irreversible risk is accumulating services and dependencies faster than authority, funding and operational evidence are built. The opportunity is to turn BELLA rights, NREN governance and partnerships into a regional commons that no single country can build alone. The distributed nature must be preserved while making accountability explicit, so that shared ownership does not mean shared ambiguity.

The long‑term significance of a shared regional backbone

RedClara illustrates a form of infrastructure independence built on cooperation, not isolation. Latin American science gains more capacity because national networks jointly control a regional layer, hold strategic capacity and participate globally on more balanced terms, not because the organisation owns every fibre. Commercial dependence remains, but the mission, routing and upgrade decisions are not all left to the public internet market.

The model goes beyond research. It shows how smaller or disparate institutions can aggregate demand without surrendering national independence, and how grants can turn into rights and institutions that outlive the project. It also shows the limits of a federation: coordination cannot indefinitely compensate for weak national networks or opaque finance and accountability.

The decisive choice is whether to expand the map only or to mature the institution. Country counts and speed numbers matter, but the permanent asset is the governance, operational, identity, data, security and collaboration ecosystem. If it becomes transparent, financially robust and genuinely regional, RedClara will remain one of Latin America’s most important shared digital infrastructures. If not, the backbone may continue to carry traffic while its higher ambitions remain dependent on the next external programme.