AxisroboticsJP

🏆 Axis Referral Championship [JULY] is LIVE! June was huge — we shipped major product upgrades and validated our data-to-model closed loop. Now it's your turn. 🔁 🗓 Duration: Jul 1st, 12PM UTC – Jul 31st, 12PM UTC, 2026 🎁 Prize Pool Breakdown: 🥇 500 USDT 🥈 300 USDT 🥉 200 USDT 🏅 4th–10th: 50 USDT each Remember: Only users you newly invite during the campaign window count — and only the trajectories they complete within that window add to your rank. Let's grow the loop together (✱,✱)

This is what Physical AI looks like when it's powered by people: a community collaborating to shape the future of Robotics, together. 🤖🇵🇭 The first Philippines Meetup was just the beginning—the journey into Physical AI has only just begun.

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.

On the Web Policy and failure-recovery side, we organized the policy release protocol. We now track success rate, model path, and heatmap status for uploaded policies, making it easier for the web frontend to integrate and verify models consistently. We also exported and uploaded browser policies and success heatmaps for several tasks, which will go live soon after a few final polish passes. At the same time, we continued collecting failure-recovery task data. Previously released failure tasks are now waiting for testing. We are also continuing new model and task data collection. Some results are still pending due to a server interruption, and we are investigating the current bottlenecks on a [redacted] embodiment.