From idea to implementation: a look into IO's research and development

As blockchain technology evolves into a foundational layer for global infrastructure, rigorous research and development are crucial to ensure robustness, scalability, and relevance. At Input | Output (IO), research is not confined to theory – it is the engine behind secure, high-performance systems like Cardano. 

IO uses evidence-based engineering to construct resilient systems that leverage formal methods and mathematical proofs to support security, correctness, and adaptability, and guarantee that core protocols are correct by design. This scientific approach supports a wide spectrum of use cases – from decentralized identity and governance to cross-chain interoperability and large-scale computation – laying the groundwork for a decentralized future that extends beyond digital finance.

This blog post highlights recent achievements from IO’s research and development (R&D) department. R&D translates foundational research into practical advancements, affirming its commitment to innovation and shaping the future of decentralized systems. 

R&D process and methodology

The R&D department was established in early 2024 with clear objectives: 

• Accelerate the delivery of research results to end users
• Enhance communication with key stakeholders
• Mitigate the risk of failure during implementation. 

The team adapted the Technology Readiness Levels (TRL) framework – originally developed at NASA and now widely adopted across research and engineering domains – from the start to better suit the context of scientific innovation and applied research. In this adapted model, the Software Readiness Levels (SRL) framework describes how ideas mature from early theoretical work (where fundamental models and security proofs are developed) through validation (where feasibility is tested via simulations, prototypes, and specifications) to targeted implementation, where engineering teams build production-ready systems guided by research outputs.

The R&D department is committed to continuous improvement and follows a clear process to explain the rationale behind its work and how research is validated. For example, in Cardano, some R&D work begins with a Cardano Problem Statement (CPS) proposal and, if accepted, leads to a Cardano Improvement Proposal (CIP). The department is also increasingly building in public, with the Leios technology validation work stream serving as a flagship example, featuring an open repository and monthly video /broadcast updates.

Recent accomplishments

Since its inception, the R&D department has rapidly expanded to become a dynamic team of 35 architects, software engineers, formal methods engineers, and applied cryptographers, fostering a collaborative environment that promotes innovation and expertise. The team is proud of several significant milestones they have recently achieved:

• Ouroboros Peras. This stream focuses on validating and recommending the implementation of Ouroboros Peras – a protocol version that brings fast finality to Cardano. Peras enhances Nakamoto-style consensus by introducing a Byzantine Fault Tolerant (BFT) voting mechanism. Participants vote on past blocks, and once a block receives enough votes, it is given a ‘boost’, enabling faster settlement without waiting for multiple confirmations. This approach significantly reduces finality time while maintaining strong security guarantees. Peras can tolerate up to 50% adversarial stake, achieves constant expected settlement time when adversaries hold under 25%, supports dynamic participation, and can recover from short-term adversarial majorities.
• Ouroboros Leios. This stream supports the development of Ouroboros Leios, a major high-throughput upgrade to Cardano’s consensus protocol, with a CIP to be proposed in the coming months. Leios is designed to significantly boost scalability and throughput while preserving decentralization and security. It introduces a concurrent blockchain structure with parallel processing to address communication and data exchange bottlenecks. Using a three-tier block system – input, endorsement, and ranking blocks – Leios enables parallel transaction validation and ordering, greatly improving performance across data, scripts, and transactions. It is tailored to Cardano’s extended UTXO model, resists up to 50% adversarial stake, and includes a fallback mechanism to the current protocol for added resilience.
• The Cardinal protocol. This protocol provides the first trust-minimized bridge between Bitcoin and Cardano, enabling secure wrapping of Bitcoin UTXOs, such as Ordinals, as native-like Cardano assets. These wrapped tokens maintain a one-to-one peg with the original Bitcoin and can be burned to release the asset back to Bitcoin. Cardinal is managed by a group of operators and maintains security as long as at least one behaves honestly. It uses a fixed-stake model with slashing, removing the need for liquidity providers or proportional collateral. The system supports verifiable, secure transfers using BitVMX and MuSig2, with a long-term goal of full trustlessness through recursive state proofs.
• CIP-0118 (nested transactions). This proposal introduces validation zones to support nested transactions, enabling use cases like Babel fees and intent-based services. It allows a transaction to include partially specified sub-transactions, which are completed by later ones, letting users transact without holding ada. The approach simplifies smart contract development and broadens access to Cardano. While it may offer a cleaner solution than some Plutus-based designs, concerns remain around ledger complexity, tooling impact, and possible centralization effects. Community discussion and evaluation are ongoing.
• CIP: Ouroboros Φalanx (avoiding grinding attacks). This upgrade to Cardano’s consensus protocol strengthens resistance to grinding attacks and improves fairness in leader election. Φalanx extends nonce generation across two epochs and introduces a verifiable delay function (VDF) based on class groups, making it much harder for adversaries to bias randomness or manipulate slot leadership. It ensures honest nodes remain efficient by supporting distributed VDF computation and includes a fallback to Praos. Simulations show a significant reduction in trivial attacks and enhanced settlement reliability, especially under high adversarial stake.
• Jolteon. This is a network-adaptive BFT consensus protocol being formalized in Agda for use on Cardano partner chains. The formalization includes liveness proofs, decidability results, a sound-by-construction trace verifier, and conformance testing. The effort includes a sliced semantic view of individual replica behavior, proven sound and complete relative to global semantics, and the extraction of a corresponding sliced trace verifier. Jolteon’s rigorous formal foundation will be recommended as a secure, reliable protocol for partner chains.
• Plutus-Halo2. This initiative demonstrates a breakthrough in on-chain verification of recursive zero-knowledge proofs on Cardano, specifically Halo2. It includes transpilers converting Rust-based Halo2 circuits to Plutus scripts, supporting operations like multiplication and ATMS signatures. Prototypes such as plutus-halo2-poc and plutus-halo2-recursion-splitted validate monolithic and multi-script verifiers within on-chain limits, tested on Cardano pre-production networks. This work expands Cardano’s support for private and scalable computation in DeFi and DApps, with open-source code to be released shortly.
• Partner chains restaking framework. A conceptual framework enabling shared security across partner chains by reusing staked ada and leveraging existing SPO infrastructure, without requiring liquidity pools or new collateral. Introduced in the partner chains toolkit (alpha v1), the model supports mixed validator committees and flexible consensus designs, allowing chains like Midnight to inherit Cardano’s security while integrating rewards and resources from multiple ecosystems. It minimizes trust assumptions (one honest operator suffices), supports slashing for misbehavior, and enables trustless interoperability via bridges like Cardinal. By aligning incentives and reducing setup costs through reusing Cardano’s stake distribution, the model fosters privacy-focused innovation and strengthens ecosystem resilience with fallback to Ouroboros guarantees.

What’s next?

Input | Output Research hosts monthly Cardano R&D sessions in collaboration with the Intersect research working group, held on the first Tuesday of each month. These research-led meetings foster strategic collaboration and engagement with the wider community. 

Stay tuned for updates on ongoing R&D efforts, and consider signing up to take part in the conversation to explore the latest advancements shaping Cardano’s future.