Ros - Research OS

The number of academic papers published every year since 2021 is over 5mln.

That’s roughly 10 every single minute, and the number is growing.

Yet, we somehow still don’t have flying cars or hoverboards (!!!)

The problem The research process is inefficient and unstructured.

• Data shared in endless emails
• No coordination between researchers
• Phantom data (links that point to nowhere because the server went down years ago)

Let’s give researchers better tools, like ROS - Research OS.

What is ROS - Research OS? It’s a decentralized node-based research management and collaboration system.

Canvas It gives structure to scientific research by letting you create workflows using a node-based canvas, where every step of the research is a node.

IPFS for files and version control Each node can have IPFS files attached onto it - so no file is ever lost - and the nodes themselves can be backed up to IPFS through a version control system (basically GitHub for research!)

Nodes <3 And of course a node can be connected to other nodes to manage dependencies through many-to-many relationships.

Once a node (research step) is marked as complete, the workflow proceeds and next connected nodes get activated, allerting the researcher that is assigned to the node and passing him all the necessary files and information he needs.

Recognition through Hypercerts Lastly, when the research is complete, a Hypercert can be minted, to give the right recognition to everyone involved.

An artifact for the future The final research can now have a way richer artifact than the usual final paper: a node graph detailing the whole process, with all the data attached - so reviewing a research for further study, or to reproduce an experiment, is easier and more informative than ever before.

Everything has been done, zero to working prototype with frontend, auth, smart contracts and IPFS file management, during Eth Global Buenos Aires hackathon by a solo hacker (me! Hi mom!).

Next.js app with shadcn and tailwind. React Flow for the node canvas and Zustand for client-side graph state. Wallet auth runs through RainbowKit/wagmi on Base Sepolia, with project artifacts pinned to IPFS via Pinata. When nodes finish, downstream nodes auto-progress, and each step’s metadata/version history is caputered on IPFS. Once a project wraps, a Hypercert ERC1155 mint is preped via a Hardhat-built contract targeting Base. I used client-side seed data plus Pinata uploads to simulate a full decentralized VCS without running my own DB yet, I’m a one person team after all so a shortcut was needed. Some GPT for coding assistance, but no boilerplates or pieces from any of my past projects - everything created by my typing or by tokens I prompted.