Two Decades of Robot Soccer, +8400 Matches, and a Record Nobody Can Edit
At six in the morning UTC on November 4, two autonomous soccer teams named Versus Projects and Wayzz played a match on FoxLeague, and Wayzz won five goals to nil. Eleven minutes after kickoff the broadcast ended, and later that day a file of roughly 830 bytes was written to Auto Drive that will still be there, unchanged, long after the tournament, the teams, and the website have gone.
The file names the teams, the score, the competition, the kickoff time, and the exact fingerprint of the 23 megabyte replay, so anyone who downloads that replay from FoxLeague can prove it is the one the match actually produced, without an account, without special software, and without asking Foxsy AI to vouch for anything. The league has now done this for every match it has ever aired.
It started as a graduation project
The work behind FoxLeague is older than most companies in this industry. In July 2002 a student named Sebastian Marian started a team called Oxsy as a graduation project in multi-agent systems in the Department of Computer Science in Sibiu, Romania, and he kept at it after graduating, entering the RoboCup Soccer Simulation League year after year. By the time Oxsy filed its 2023 team description paper, it was preparing for a nineteenth consecutive appearance. RoboCup is the research competition where autonomous systems play soccer against one another, organized around a stated long-term goal of fielding a robot team capable of beating the human World Cup champions, and nothing in it is scripted, so a team wins only because its code reads a situation it has never seen and responds better than the other team’s does.
Foxsy AI grew out of that lab. In the company’s own description:
Foxsy AI is building the evolution engine for robotics. Through FoxLeague, autonomous competitions become a live evolution loop: every match played by the community produces competition data and makes the next generation of intelligent machines smarter. Foxsy AI grew out of Oxsy, a RoboCup team since 2003, and is taking that engine from simulation toward real robots.
The practical form of that is a public league. Developers upload a RoboCup-compatible team binary through the Foxsy AI Developers Hub, run a verification match against a base team, review the replays and logs, and then enter tournaments that are broadcast live while the community watches.
Why a league cannot be its own witness
Every competition keeps two things, the result and the recording that shows how the result happened, and in almost every league both of them sit on servers the organizer owns. The arrangement is fine while nobody is suspicious, and it has no answer at all for the first person who is, because a score in a database can be corrected quietly and a replay file can be re-encoded, trimmed, or swapped without leaving a mark that a viewer could find. An organizer’s assurance that nothing was touched is not evidence, however honest the organizer happens to be.
In a league of human players that is a governance problem, handled well enough by referees and reputations. In a league of autonomous ones it turns into something closer to a scientific problem, because each match records how a set of independently written systems behaved under conditions nobody controlled, and those records feed the design of the systems that come next. Rather than ask anyone to take its word for it, Foxsy AI put the evidence somewhere it cannot reach.
Live in a day, complete in a week
Records ran on a test network on November 3 and went live on the production network the next day, with seventy-five matches written in the first two and a half hours. By November 10, more than 8,400 records covered the complete history of everything FoxLeague had aired, with none pending and none failed.
A record is written only after the live broadcast ends, which keeps provisional results out of the public history. Here is what one holds, using the November 4 match as the example:
.png)
The record itself is public and readable right now at gateway.autonomys.xyz. Every match page on FoxLeague carries a “Permanent record on Autonomys” link beneath the replay.
Checking a replay takes three steps: download the replay log from the match page, compute its SHA-256 fingerprint (shasum -a 256 on macOS or Linux, Get-FileHash on Windows), and compare that fingerprint and the file size against the record. Matching values mean the file in your hands is the one that existed when the broadcast ended. The record behind that check cannot be rewritten, because a file written through Auto Drive is addressed by its own content, so an altered version would resolve to a different address while the original stayed exactly where it was put. Foxsy AI can add new records as matches air, but changing the ones already written is not something it or anyone else can do.
The part that matters in ten years
What exists today is a set of small files describing replays. The next phase stores the matches themselves. From November, the full replay file of every tournament match FoxLeague broadcasts and every 1v1 challenge, close to 300 matches a day, is written to Auto Drive once its broadcast has ended, linked to the match’s permanent record. That includes every Champions FoxLeague match, from the first round on December 16 to the final on December 24. After that Foxsy AI intends to assemble an open match dataset for AI and robotics research out of the history the league has already accumulated, while continuing to move the engine from simulation toward physical robots.
That dataset would be a continuous, time-stamped account of how thousands of independently written autonomous systems behaved in a competitive environment, produced by matches nobody could arrange in advance, with every entry fingerprinted before anyone had a reason to argue about it. Material like that is rare, and it grows rarer as the share of machine-generated content rises, since the cost of producing a convincing fabrication keeps falling while the cost of proving something genuine does not fall with it.
A corpus assembled after the fact inherits that problem, because the only thing standing behind it is the word of whoever assembled it, while one whose every entry was committed to a public record on the day it was created does not. That is the quiet value of what is being written now. The records are small, cheap and individually unremarkable, and their worth lies in when they were written and in the fact that nobody can adjust them later.
Why this integration matters to us
Sebastian Marian has been on the same problem since a graduation project in 2002, which is longer than most companies in this industry have existed, and he had permanent records running on the live network for the league’s entire history before anyone asked him for it. Patience measured in decades and delivery measured in days do not often turn up in the same team. Sebastian has been a pleasure to work with throughout, and we are looking forward to the next phase as much as to this one.
Auto Drive’s role is narrow and permanent. It keeps each record exactly as written, for as long as anyone cares to check it, with no pinning, no renewal cycle, and no subscription whose lapse would quietly erase the archive. As Foxsy AI moves from describing replays to storing them, and from stored replays to an open research dataset, every one of those steps depends on the records staying exactly as written.
The robots are still playing in simulation, and by the time they are not, there will be a verifiable record of every match that got them there.
We look forward to sharing more about Foxsy AI and the team behind it as the integration deepens, and we are proud to provide the permanent storage holding the record of their work.
Auto Drive is free to start at ai3.storage.
Auto Drive Now Works With rclone
A surprising amount of the world’s important data moves through a command-line tool called rclone. System administrators rely on it for overnight server backups, research teams use it to archive large datasets across institutions, and developers lean on it to migrate between cloud providers without writing custom code. With support for more than 70 cloud storage products, it has become a standard part of the toolkit for anyone who manages data at scale.
Starting with rclone v1.76.0, Auto Drive, Autonomys Network’s gateway to permanent storage, is one of the destinations rclone can reach.
Who relies on rclone
Users often call rclone the Swiss army knife of cloud storage. You set up a remote once, point a command at it, and rclone handles the differences between storage systems on your behalf. Its core audience is the system administrators, data engineers, and infrastructure leads who decide where an organization’s files live. They choose tools carefully, since a poor choice can mean higher costs, outages, or lost records, and once a tool earns a place in their workflow it tends to stay there.
How the integration works
Auto Drive offers an S3-compatible application programming interface (API), the same kind of interface Amazon S3 uses, so rclone connects to it through its standard S3 settings. Setup takes an API key from ai3.storage and a short entry in your rclone config file. From there, the commands and scripts a team already runs can send files to Auto Drive, where they stay available permanently with no pinning or subscription renewals to manage.
Because nothing stored on Auto Drive can be removed, rclone can add files there but not delete them, so the setup guide recommends using rclone copy rather than rclone sync.
Another way in
rclone joins the ways people already use Auto Drive. Anyone can drag and drop files into ai3.storage much as they would with Google Drive, and developers can build lasting storage into an application with a few lines of code. This integration extends the same access to engineers and infrastructure teams who would rather keep working in the tools they already trust.
A research institution can point rclone at Auto Drive and keep years of scientific data available without a subscription to renew or a vendor that might discontinue its product. A developer moving off a legacy cloud provider can keep their existing tooling, and a team backing up critical records gets verifiable storage through a tool it already knows.
The timing also matters for companies working to meet Article 12 of the EU AI Act, which requires high-risk AI systems to automatically record events over their lifetime. Tamper-evident, permanent storage can help support that kind of record-keeping, and rclone gives teams a simple route for getting logs there.
Upload Once. Access Forever.
That has been the promise behind Auto Drive from the start, and rclone brings it to a much larger group of people. Teams don’t need to understand how Auto Drive works under the hood, since permanence is built into the network itself and applies as soon as files arrive.
Auto Drive is free to start. You can create an account and generate an API key at ai3.storage, and the rclone setup guide walks through the configuration step by step.
Want the full technical story? Our forum deep dive covers the S3 compatibility work behind the integration, from multipart ETags to moving the S3 API off our CDN. It’s written for developers, so expect protocol-level detail.
October 2026 | End-of-Month Report
On October 13, we launched Pay with USDC, so teams can now buy permanent storage on Auto Drive with a dollar-backed stablecoin instead of acquiring AI3 first. That extra step had been a real barrier for many teams, and removing it means buying storage on Auto Drive now works much like buying storage from the cloud services they already use.
The rest of the month was shaped by partners moving from evaluating Auto Drive to building on it. healthpass1 is integrating, work began on the accounts that will let partners see usage from their own platforms, and feedback from teams mid-integration is now guiding our reliability work.
Network Status
The network ran normally throughout October. Block production, storage, and domain operations were stable, with no incidents to report, and no action was required from farmers, operators, or nominators.
Ecosystem & Developer Momentum
Pay with USDC Is Live
Auto Drive users can now purchase storage credits with USDC on Ethereum, alongside the existing Pay with AI3 option. Storage on the network is still settled in AI3 whichever option a user chooses, and the independent farmers who provide that storage continue to earn AI3 rewards. Full details are in More Ways to Access Permanent Storage: Introducing Pay with USDC, and integration instructions are in the developer docs.
Partner Accounts and Usage Attribution
Work began on partner accounts and usage attribution, so that storage used through a partner’s platform is recorded against that partner. The changes coming to the grants program depend on this work.
Consolidation Work
We have started a program of consolidation work focused on Auto Drive’s reliability, guided by the needs of partners who are integrating now. So far, that has meant targeted investigation and fixes for issues we have identified, and the biggest improvement delivered to date is faster downloads for larger files. More fixes are in progress, all aimed at making sure what Auto Drive already offers works the way it should.
Auto Drive Is Now an Official rclone Provider
Auto Drive is now a built-in storage provider in rclone, the widely used command-line tool for syncing and backing up files across cloud storage. Our provider was merged on October 16 (rclone#10093) and ships in rclone v1.76.0. Users pick “Autonomys Auto Drive” from the setup menu, add their API key and sync straight to permanent storage. October’s S3 compatibility work, including a CDN-free S3 hostname and support for large files, took rclone’s integration suite from 22 ignored tests to just 3. Setup details are on rclone.org.
It’s the latest step in making storage on the Autonomys Distributed Storage Network (DSN) easy to use from the tools people already rely on, meeting storage users where they live. We’ll share the full story, including the technical challenges we solved along the way, in a forum post shortly.
Community & Foundation
Guardians of Growth
Season 2 of the Subspace Foundation’s staking program continued through October. Details on the program and how to take part are in Guardians of Growth Season 2 Is Underway.
Community Contributions
October brought the highest volume of community contributions we have ever seen across our repositories, and we are grateful for it. We are also planning changes to the contribution contest so that it reflects how much easier agentic coding tools have made it to open a pull request. The network benefits most from pull requests that fix existing issues, introduce meaningful features, or address bugs, whether already known or newly identified. If you are considering contributing, please make sure your pull request addresses at least one of those areas.
ICYMI: Content Published in October
- September 2026 | End-of-Month Report
- More Ways to Access Permanent Storage: Introducing Pay with USDC
Metrics: October Snapshot
Staking: 49,877,070 AI3 (+2,551,562 from September)
Auto Drive Files Uploaded: 2,593 (October)
Auto Drive Downloads: 295 (October)
Auto Drive Users: 760 (+14 from September)
Total Announced Partnerships: 66
Announced Auto Drive Integrations: 12
Total Grant Applications: On hold
Looking Ahead
November is shaping up to be a busy and exciting month. We look forward to sharing more on rclone, a new partner integration, and several ongoing partnerships as we continue working to make permanent storage easier to adopt and use.
How to Get Involved
The most helpful thing you can do is put Auto Drive to work. It is free to start at ai3.storage, with 20 MB of uploads and 5 GB of downloads each month, and beyond the free tier you can purchase additional storage directly with Pay with AI3 or Pay with USDC. If you are building something where records need to outlast the system that produced them, that is the fastest way to find out whether this fits.
We would also appreciate support through referrals for integrations. If you know a project dealing with expiring links, pinning costs, or an audit trail nobody can verify, send them our way.
More Ways to Access Permanent Storage: Introducing Pay with USDC
We want permanent storage to be easy to get started with, and that idea guided our first payment option in Auto Drive, Pay with AI3. It works well, but until now, anyone who needed more than the free tier had to acquire AI3 before they could buy additional storage, and for many people that extra step was a barrier.
Today we are introducing Pay with USDC, which lets users purchase Auto Drive storage credits with USDC on Ethereum, a stablecoin many people already hold.
Under the hood, storage on the Autonomys Network is still settled in AI3, whether a user pays with AI3 or USDC. The independent farmers who provide the network’s storage continue to get AI3 rewards. For users, the experience now looks much more like the cloud storage tools they already use every day.
AI3 continues to be the token behind the Autonomys Network. Anyone who wants to hold or trade it directly can find it on Uniswap as an ERC-20 token on Ethereum, and on XT.com in its native form.
Auto Drive exists to keep important data available for the long term, and we will keep looking for ways to make it easier to use. You can create an account at ai3.storage and start with the free tier, then use Pay with AI3 or USDC whenever you need more storage.
For full release details, visit: https://forum.autonomys.xyz/t/introducing-pay-with-usdc-more-ways-to-access-permanent-storage/5033
Developer Docs:
https://develop.autonomys.xyz/auto_drive/pay_with_usdc
September 2026 | End-of-Month Report
In The OpenAI Incident: When Agents Can Edit Evidence, published this month, we made the case that permanent, immutable records are the layer every other AI safeguard depends on. That argument is not difficult to make anymore. Agents now act with real autonomy, incidents get investigated long after the fact, and regulators are writing record-keeping obligations into law. The harder question is whether the teams who need that infrastructure can adopt it without rearranging how they already work.
Most of our September worked toward answering that second question. Enabling payment for storage in a currency teams already hold, returning unused credits without anyone having to file a claim, and meeting developers inside an SDK they already use. Each one removes a step between deciding to store something permanently and actually doing it. Here is where things stand heading into October.
Network Status
The network ran normally throughout September. Block production, storage, and domain operations were stable, with no incidents to report, and no action was required from farmers, operators, or nominators.
Ecosystem & Developer Momentum
Pay with USDC
Work on Pay with USDC, the feature that lets teams buy Auto Drive storage credits in USDC, continued through September. End-to-end testing is complete on Sepolia, an Ethereum test network, and the feature is now in its final stages of internal testing as we move from Sepolia to Ethereum. We are preparing for a public launch.
Storage Credit Refunds
Storage credits purchased on Auto Drive are valid for three months. Any credits left unused when that window closes are refunded in AI3, automatically, to the wallet address used for the purchase. All unused expired credits have now been refunded, and there is nothing for anyone to claim. Worth stating plainly, since the two are easy to confuse: the credits expire, the data does not. Anything already stored stays permanent.
Partner Integrations
XYO added Auto Drive as a supported Data Lake in the XYO SDK this month. XYO is one of the largest Decentralized Physical Infrastructure Networks (DePIN), with more than 10 million nodes validating real-world data, and its Layer One chain records a compact reference to each piece of data while the data itself lives in a Data Lake. That design puts real weight on where the data is kept. Developers can now point that storage at Auto Drive using the SDK modules they already work with. Full details are in XYO Adds Auto Drive as a Supported Data Lake in the XYO SDK.
Community & Foundation
Guardians of Growth Season 2 Underway
Season 2 of the Subspace Foundation’s staking program began on September 9 and ran its first full month. Details on the program and how to take part are in Guardians of Growth Season 2 Is Underway.
Grants Program Relaunch
The Subspace Foundation’s grants program is close to relaunching. We look forward to sharing details in October.
A Note on Community Contributions
Contributions across our repositories increased noticeably over the past month, and we appreciate the effort and interest behind that. The network benefits most from pull requests that fix existing issues, introduce meaningful features, or address bugs, whether already known or newly identified. If you are considering contributing, please make sure your pull request addresses at least one of those areas.
ICYMI: Content Published in September
- August 2026 | End-of-Month Report
- The OpenAI Incident: When Agents Can Edit Evidence
- Guardians of Growth Season 2 Is Underway
- XYO Adds Auto Drive as a Supported Data Lake in the XYO SDK
Metrics: September Snapshot
Staking: 47,325,508 AI3 (+340,066 from August)
Auto Drive Files Uploaded: 825 (September) / 136,058 (All time)
Auto Drive Downloads: 278 (September) / 19,150 (All time)
Auto Drive Users: 746
Total Announced Partnerships: 66
Announced Auto Drive Integrations: 12
Total Grant Applications: 110 (+5 from August)
Looking Ahead
In October, Pay with USDC moves toward public launch, Guardians of Growth Season 2 continues, and work carries on across the integrations now in progress.
How to Get Involved
The most helpful thing you can do is put Auto Drive to work. It is free to start at ai3.storage, with 20 MB of uploads and 5 GB of downloads each month. If you are building something where records need to outlast the system that produced them, that is the fastest way to find out whether this fits.
We would also appreciate support through referrals for integrations. If you know a project wrestling with expiring links, pinning costs, or an audit trail nobody can verify, send them our way. Teams needing larger allocations can apply through the Subspace Foundation Grants Program.
XYO Adds Auto Drive as a Supported Data Lake in the XYO SDK
The data behind a proof on XYO Layer One lives outside the chain, in a Data Lake. Developers building on XYO now have a storage option for that data that cannot expire, be deleted, or be altered after the fact.
XYO has added Auto Drive as a supported S3-style Data Lake in the XYO SDK. Developers building on XYO Layer One can now store the data behind their transactions on Auto Drive, the permanent storage gateway from Autonomys Network, using the same SDK modules they already use today. XYO has published a working sample project that runs the complete flow end to end.
Who XYO is

XYO is the original and one of the largest Decentralized Physical Infrastructure Networks (DePIN), with more than 10 million nodes collecting and validating real-world data. Its Proof of Location and Proof of Origin technologies validate that data at the source, which makes XYO a trusted input for AI, geolocation, real-world asset tracking, gaming, and other sectors that depend on knowing what actually happened in the physical world.
XYO has since expanded its ecosystem with XYO Layer One, a blockchain purpose-built for the data demands of DePIN, real-world assets, and AI. XYO Layer One runs on a dual-token model. The XYO token supports the DePIN ecosystem, rewards, staking, and governance. The XL1 token handles transactions and fees inside XYO Layer One itself.
The problem every high-volume chain has to solve
A network that verifies real-world data at XYO's scale produces far more data than any chain should carry. XYO solves this with an architecture first described in its original 2018 white paper: Archivists. The modern, lighter-weight version built for XYO Layer One is called a Data Lake.
The design is simple. XYO Layer One records a compact, verifiable reference to each piece of data. The data itself lives in a Data Lake. Because the reference is derived from the content, anyone who retrieves the data can confirm it is exactly what was recorded, without having to trust whoever is storing it. The chain stays lightweight. The data stays verifiable.
That design puts real weight on the Data Lake. If the stored data disappears, the reference on XYO Layer One still exists, but it points at nothing. If the data is altered, the reference no longer matches, and the record is broken. For applications built on proofs of location and origin, that is not an edge case but the failure that matters most.
What XYO released

The XYO SDK supports several storage backends for Archivists and Data Lakes. S3-compatible object storage is handled by the @xyo-network/archivist-s3 package, and that package now includes a dedicated Auto Drive module. A developer imports createAutoDriveArchivist, supplies an Auto Drive API key, a bucket, and a key prefix, and the SDK handles the rest. There is no custom integration to write and no separate client to maintain.
The published sample project shows the full loop in a single command-line tool. It takes a message, stores it on Auto Drive through the XYO adapter, reads it back to confirm the content and its hash match, signs a transaction with the AriesTools CLI wallet, anchors the reference on XYO Layer One, waits for the transaction to be confirmed as final, and prints a link to it in XYO Explore. Storage is verified before anything is broadcast, and verified again after inclusion.
What this means for developers building on XYO

Auto Drive is built for the job a Data Lake needs done.
The data does not expire. Auto Drive has no pinning, no renewal, and no ongoing fees for data already stored. Once a payload is written, it stays retrievable. There is nothing to renew and nothing to let lapse.
The data cannot be altered or deleted. Auto Drive's S3 interface returns an error on any delete request by design. Re-uploading the same key creates a new object rather than overwriting the old one. A reference recorded on XYO Layer One will always resolve to the same bytes it was created from.
The verification model matches. XYO Data Lakes derive keys from content so that readers can check what they receive. Auto Drive stores every object under a content identifier and returns that identifier with each response. The two systems make the same promise at different layers.
It fits the existing S3 workflow. Auto Drive's S3-compatible layer supports the operations the XYO adapter relies on, including standard and multipart uploads, object retrieval, and listing. Teams already using S3 tooling with XYO do not need to change how they work.
It is free to start. Every Auto Drive account receives 20MB of upload capacity and 5GB of download capacity per month at signup. Beyond that, teams purchase additional storage directly. Companies bringing paying customers to permanent storage can also apply to the Subspace Foundation Grants Program.
Try it
- Sign up at ai3.storage and create an API key in the Developers section.
- Clone the autodrive-datalake-sample repository and follow the README.
- For a first run that needs no XL1 test tokens, pnpm test:live writes a real payload to Auto Drive through the XYO adapter using a disposable local chain and wallet.
Full details on the Auto Drive S3 layer are in the Autonomys Developer Hub. Details on XYO Data Lakes and the XYO SDK are in the XYO documentation and the XYO AI SDK skills repository.
Why this matters to us
XYO has spent years solving one of the hardest problems in infrastructure: proving that real-world data is what it claims to be. That work only holds if the data behind each proof is still there, unchanged, when someone needs to check it. Auto Drive exists to make that guarantee at the storage layer. Seeing it added to the XYO SDK, in released code with a public sample, is exactly the kind of integration Autonomys Network was built for.
We look forward to seeing what XYO developers build on it.
Auto Drive is free to start at ai3.storage.
