Show HN: Decentralized robots (and things) orchestration system

docs.p2p.industries

51 points · hannesfur · 3 days ago

Hi HN, we build an open-source operating system extension for orchestrating robot swarms fully decentralized.

This first beta version allows you to create fully decentralized robot swarms. The system will set up a wireless mesh network and run a p2p networking stack on top of it, such that nodes can interact with each other through various abstractions using our SDKs (Rust, Python, TypeScript) or a CLI.

We hope this is a step toward better inter-robot communication (and a fun project if you have a few Raspberry Pis lying around).

Our mesh network is created by B.A.T.M.A.N.-adv and we’ve combined this with optimized decentralized algorithms. To a user, it becomes very easy to write decentralized applications involving several peers since we’ve abstracted away much of the complexity. Our system currently offers several orchestration primitives (Key-Value Store, Pub-Sub, Discovery, Request-Response, Mesh Inspection, Debug Services, etc.)

Internally, everything except the SDKs is written in Rust, building on top of libp2p. We use gRPC to communicate between the SDKs and the CLI, so libraries for other languages are possible, and we welcome contributions (or feedback).

The C++ SDK and a ROS package that should feel natural to roboticists are in the works. Soon we also want to support a collaborative SLAM and a distributed task queue.

We’d love to hear your thoughts! :)

24 comments

accurrent · 19 minutes ago

This looks super cool, Ive been working in a similar space for some time. I work with open-rmf which is semi-decentralized and provides tools for task and traffic deconfliction (we don't handle the network layer at all). Excited to see more similar software coming up.

wngr · 5 hours ago

Great idea combining batman with libp2p! You guys have the heart in the right place :-).

Currently, your project seems to be an opinionated wrapper ontop of libp2p. For this to become a proper distributed toolkit you lack an abstraction to for apps to collaborate over shared state (incl. convergence after partition). Come up with a good abstraction for that, and make it work p2p (e.g. delta state based CRDTs, or op-based CRDTs based on a replicated log; event sourcing ..). Tangentially related, a consensus abstraction might also be handy for some applications.

Also check out [iroh](https://github.com/n0-computer/iroh) as a potential awesome replacement for p2p; as well as [Actyx](https://github.com/Actyx/Actyx) as an inspiration of similar (sadly failed) project using rust-libp2p.

Oh, and you might want to give your docs a grammar review.

Kudos for showing!

Show replies

Animats · 3 hours ago

Read the architecture document here.[1]

The usual problems with these things are discovery and security. Discovery is done via local WiFi broadcast. Not clear how security is done. How do you allow ad-hoc networking yet disallow hostile actors from connecting?

[1] https://docs.p2p.industries/concepts/architecture/

Show replies

NotAnOtter · 6 hours ago

Very fun. Is this primarily a passion project or are you hoping to get corporate sponsorship & adoption?

Can you provide some insight as to why this would be preferred over an orchestration server? In this context - Would a 'mothership'/Wheel-and-spoke drone responsible for controlling the rest of the hive be considered an orchestration server?

This isn't my area of expertise but I think "Hive mind drones" tickles every engineer.

Show replies

jfantl · 5 hours ago

This is awesome stuff, I'm going to look into getting this running my Pis this weekend. How hard would it be to add in custom services? I like to play with decentralized algorithms such as Size Estimation and Clock Synchronization (https://jasonfantl.com/) and have always wanted to get them running on real hardware.

Show replies