PORABOLA(✱,✱)

web3 & AI, content creator
@axisrobotics - contributor

[PORABOLA(✱,✱) @Porabolla]

https://t.co/dkfy4wK1zB

[郡主Christine (✱,✱) @0xsexybanana]

base拍视频的人是天才吧，@donofbase

给我tm拍的腿长一米八。

Axis robotics致力于解决物理AI的数据鸿沟，任何人都可以在Axis 上贡献机器人数据。像玩游戏一样简单。

七月axis robotics将会有非常多大新闻，我们正在构建具身智能的数据飞轮。

我们将数据从传统离线、串行的采集方法，换成全设备集群和并行的采集。

对于数据采集者来说，不要任何门槛和硬件。对客户公司来说：完全软硬件解耦。极大地提升了采集的accessibility和scalability。

同时我们随机化任务生成引擎、丰富的embodiment支持、海量的资产库确保了最重要的东西：数据的多样性。

真是非常bullish axis非常bullish base

base is for AI, based is for robotics and Base js for everyone

Let’s build on Base lalala

[PORABOLA(✱,✱) @Porabolla]

@antonina8622 @axisrobotics 🥰ай молодец

[PORABOLA(✱,✱) @Porabolla]

Не успел наложить субтитры 😁 пусть будет эксклюзивно для CIS региона 😎

@axisrobotics обучение для заработка, бесплатное участие.

Немного информации о железном двуруком пареньке...

[PORABOLA(✱,✱) @Porabolla]

@deedmemories @axisrobotics 😁

[PORABOLA(✱,✱) @Porabolla]

July promises to be hot and the @axisrobotics team has a lot of interesting things in store for us...

But what could it be??? We're waiting 😎

[PORABOLA(✱,✱) @Porabolla]

@antonina8622 @axisrobotics top 100 sooooon 🤓

[PORABOLA(✱,✱) @Porabolla]

@antonina8622 @axisrobotics 😎😎😎

[Axis Robotics @axisrobotics]

Axis Weekly

This week, we consolidated our June progress into a clearer data-loop direction: moving beyond standard short-horizon single-arm demonstrations toward complex-task data, correction data, and continuous model iteration.

Key updates:

- Teleoperation UX: We improved direct gripper dragging, object selection, and bimanual control to skip low-information actions and preserve high-value demonstration segments.

- Data quality: We strengthened verification and checker logic against new cheating patterns, extending stricter validation to bimanual tasks and DAgger collection.

- Model iteration: The automated task-to-policy loop is now largely connected, and we are shifting toward DAgger-style correction data to better distinguish human intervention from policy rollouts.

- TaskGen: Articulated-object support expanded beyond six categories, using a coding agent for asset generation and a semantic LLM agent with DINO for better asset retrieval.

- Real-world validation: Dataset v2 long-horizon data collection is underway, with early real-world results suggesting AXIS + DROID co-training preserves useful learned priors.

Details below 🧵

[PORABOLA(✱,✱) @Porabolla]

@antonina8622 @plpiaoliang @axisrobotics thank you 🥰

[PORABOLA(✱,✱) @Porabolla]

today i achieved the result had been working hard toward. 😤

I want to thank @plpiaoliang and entire @axisrobotics team for the trust they placed in me.

we aren't stopping - we’re moving straight ahead, as there is still a great deal of work to be done!

[PORABOLA(✱,✱) @Porabolla]

@deedmemories плохих не делаю 😁спасибо

[PORABOLA(✱,✱) @Porabolla]

https://t.co/ZZsvKtytlh

[PORABOLA(✱,✱) @Porabolla]

@axisrobotics great upgrade especially for a two-armed robot 👍

[Axis Robotics @axisrobotics]

Double-Click Auto-Navigate & Save/Rewind are LIVE!

Following last week's click-and-drag gripper control, here's what's new:

Double-click any object and the nearest arm automatically moves into position.
- Each movement is randomized, so every approach adds training diversity.
- Works for both single-arm and dual-arm tasks.

We also shipped Save & Rewind.
- Press N to save a checkpoint at any critical moment during a task. The checkpoint indicator in the top-right corner will light up to confirm.
- Press B to instantly roll back to your saved state if anything goes wrong.

That means you can now drag arms into place, double-click to auto-navigate to any object, and save checkpoints on the fly — all directly from your browser.

Smoothest simulation teleop UX, only on Axis. Try it out now!

[PORABOLA(✱,✱) @Porabolla]

@deedmemories большое спасибо 🦾

[PORABOLA(✱,✱) @Porabolla]

https://t.co/3oPLMuE0zK

[PORABOLA(✱,✱) @Porabolla]

@antonina8622 🥰

[Axis Robotics @axisrobotics]

A few weeks ago, we shared our progress on articulated objects and long-horizon tasks.

Here are two representative examples:

- We've been steadily expanding our asset library to cover more articulated objects.

Articulated objects have always been a challenging asset class to handle in simulation. Interacting with them requires robots to master atomic skills such as pushing, pulling, opening, and closing, and to understand part structure, interaction constraints, and how the object moves.

- Long-horizon tasks can now be generated at scale.

Long-horizon tasks are the other hard category: they require chaining multiple sub-goals in sequence. A failure early in the task can cascade and make the rest unrecoverable.

Axis is scaling along three dimensions at once: data volume, data quality, and task difficulty.

[PORABOLA(✱,✱) @Porabolla]

@axisrobotics 3 star difficulty... i'm ready for it 😎

[PORABOLA(✱,✱) @Porabolla]

@antonina8622 @axisrobotics impressive result 😎