BadMaru

[Scott Manley @DJSnM]

Sound synchronized from @NASASpaceflight footage

[phil beisel @pbeisel]

3 trackers for Artemis

1. https://t.co/25SS7Pv6rr (official)
2. https://t.co/ymFrnf8bhW
3. https://t.co/z4W2AGPrB8

[BadMaru @badmaru777]

@elonmusk Problem with payload of starship?😂

[Jack Vial @jackvial89]

much smoother after applying a low pass Butterworth filter with a 3hz cutoff. this filters out high frequency small movements (aka noise) in the trajectory, it also naturally attenuates the signal so makes the robot move slower, I’ve added a bit of gain back to compensate and speed it up again after the filtering. still a bit jiggly, mainly in the shoulder pan but it seems to be mostly mechanical at this point

[Robots Digest 🤖 @robotsdigest]

A factory that fits on a desk.
Runs 24/7.
Costs ~$5k.
Swaps tools.
Builds electronics on its own.
This isn’t a robot arm.
It’s a manufacturing primitive.

[Ilir Aliu @IlirAliu_]

Robots don’t wait. So why should your model? Large VLAs run in real time… with no training-time changes.

❗️Worth reading if you’re working on real-world deployment of large models in robotics.

I found this write-up on Real-Time Action Chunking (RTC) from Physical Intelligence, a method that lets VLAs like π0 and π0.5 execute actions while still “thinking.”

Instead of waiting for inference to finish, the robot starts acting; and fills in the next steps like inpainting.

The results?
✅ Smoother motion
✅ Faster task completion
✅ Higher precision under latency

Even with 200ms of added delay, RTC kept success rates high, while naive methods collapsed.

📌 Blog + paper:
https://t.co/y3FqUgh7wI

—-

Weekly robotics and AI insights.
Subscribe free: https://t.co/dsa6wcvq6n

[World of Engineering @engineers_feed]

This rarely seen footage shows NASA astronauts struggling to walk on the moon

[Ilir Aliu @IlirAliu_]

A Python-based lightweight robot simulator designed for navigation, control, and reinforcement learning:

[📍 Github Link]

Most robotics simulators are powerful but heavy. Sometimes you just need a lightweight place to test navigation, control, or reinforcement learning without a full physics stack.

IR-Sim is one option worth knowing if you work in robot navigation or AI.

Why it stands out:

✅ Fast to install and simple to set up
✅ Scenarios defined in plain YAML
✅ Real-time visualization with Matplotlib
✅ Built-in collision detection
✅ Good for AI and reinforcement learning workflows

What you can prototype:

- Multi robot collision avoidance
- Lidar based navigation
- Dynamic scenes with moving obstacles
- High level behavior testing before going to a full simulator

The project is open source under MIT, actively maintained, and already used in several research papers on navigation and planning.

📍GitHub: https://t.co/DPySrtCfDe
Docs: https://t.co/vDCVYxfoeB

For students, researchers, or anyone exploring navigation algorithms, IR-Sim offers a simple way to try ideas quickly without large dependencies.

____

Weekly robotics and AI insights.
Subscribe free: https://t.co/dsa6wcvq6n

[Ilir Aliu @IlirAliu_]

🦾 Create 𝗬𝗢𝗨𝗥 𝗢𝗪𝗡 𝗥𝗢𝗕𝗢𝗧𝗜𝗖 𝗔𝗥𝗠.

Source all the parts - 🤖 Step-by-Step Guide:

Build a DIY robotic arm from scratch with a comprehensive, easy-to-follow guide.

🧩 Sourcing Parts:
Obtain all necessary parts for a 3D printed robotic arm, either individually for about $400 or as a complete kit for $270.

💻 Open Source Code:
Utilize our GRBL-based firmware, tailored for 6-axis robotic arms and compatible with various software options for ease of use.

Link: https://t.co/HmhrhZsfOa
Video: https://t.co/aTzvGDKKlQ

—-

Weekly robotics and AI insights. Subscribe free: https://t.co/dsa6wcvq6n

[Ilir Aliu @IlirAliu_]

Most gripper are made to pick or grab just one type of object… often tool changing is the fastest and most efficient option.

Gripper-cushion grips it all❗

Handles sheet metal, liquid tanks, nonwovens, and pipes:

These gripper pads are revolutionizing material handling in manufacturing.

✅ Versatile: Capable of gripping diverse materials from sheet metal to pipes.
✅ Simplified Handling: Reduces the complexity of robotic programming.
✅ Self-Calibrating: Utilizes gravity for error-resistant gripping.
✅ No Special Design Needed: Eliminates the need for precise gripping points on components.

Not something you see every day.

Saw this at: https://t.co/Q2PojIW3Fh

—-

Weekly robotics and AI insights. Subscribe free: https://t.co/dsa6wcvq6n

[Ilir Aliu @IlirAliu_]

1.7× faster inverse kinematics in pure Python, GPU-accelerated! Fully open-source and built for scale
[📍Bookmark for later]

PyRoki, a modular toolkit for robot kinematic optimization, supporting
- inverse kinematics
- trajectory optimization, and
- motion retargeting.

Built for flexibility, but also fast:
✅ Runs on CPU, GPU, and TPU
✅ Outperforms cuRobo in speed, success rate, and accuracy
✅ Fully in Python, easy to integrate and extend

Whether you’re working with industrial arms, simulation environments, or humanoids, PyRoki is designed to fit your stack and scale with your hardware.

Thank you, @ChungMinKim, for sharing!!

Website: https://t.co/YRlgDbTJ0Y
Code: https://t.co/8D1pcE0oxq
Paper: https://t.co/SuFyT8XwHc

This one’s for researchers, builders, and anyone tired of debugging their IK stack.

—-

Weekly robotics and AI insights. Subscribe free: https://t.co/dsa6wcvq6n

[Ilir Aliu @IlirAliu_]

Most robots fail at the last millimeter.
That’s where precision, not power, decides everything.

This system aims to close that gap.
It executes tolerance-critical motions with steady, millimeter-level control, without manual tuning or retraining.

✅ Adapts instantly to new parts and surfaces
✅ Works across robot types and grippers
✅ Runs in real production, not just simulation
✅ Reduces manual positioning and touch-ups

I’ve seen production lines where operators still guide fragile parts by hand because robots couldn’t hold the tolerance.

That’s where real autonomy breaks down.

With Cortex (@SereactAI‘s Vision-Language-Action (VLA) model), those same operations start to look different.

The robot holds its own precision, even when the environment changes.

Robots got faster.
But when control slips, speed only hides the problem.

Precision is what actually scales.

—-
Weekly robotics and AI insights.
Subscribe free: https://t.co/dsa6wcvq6n

[Humanoids daily @humanoidsdaily]

Zurich's Mimic Robotics has raised a $16M seed round led by Elaia & Speedinvest. Their plan: Skip the full-body robot and build an AI foundation model for dexterous hands, trained on data captured directly from skilled human workers on factory floors.

[Chestnut Robotics @ChestnutRobotic]

Our open-source robotic hand just won a thumb wrestling match 🤖💪

Human-like motion. Real-time control. A surprisingly competitive thumb 👍

Want to try it yourself? 👇
🔗 https://t.co/QeohUpAnUB
💻 https://t.co/H0UhUwB154

#robotics #AI #opensource #teleoperation #dexteroushand

[RoboHub🤖 @XRoboHub]

VLAI Robotics Launches High-Dexterity Dual-Arm Robot Starting at ~$5,500 USD
VLAI Robotics, utilizing the open-source design of Japan’s OpenArm team, has introduced a high-dexterity, low-cost bimanual robotic system. The product is priced from 39,900 RMB (approximately 5,500 USD), making high-level research tools accessible to institutions and developers.

The robot features a human-scale, arm-hand integrated design with up to 16 DOF (8 DOF per arm, including the gripper). It can handle a dual-arm peak payload of 12 kg while maintaining high precision. A key breakthrough is its ability to replicate human upper limb movement trajectories, which enhances data quality for imitation learning and remote operation.

VLAI Robotics handled the domestic engineering, manufacturing, and VLA (Vision-Language-Action) algorithm integration, with quality control organized according to the strict standards of the OpenArm R&D framework. The robot is highly adaptable for research, education, and Physical AI training, supporting easy expansion for features like VR teleoperation.

[CyberRobo @CyberRobooo]

Every time I see the demonstration of GITAI's lunar operation humanoid robot, I feel like a human settlement has already been built there.

[Massimo @Rainmaker1973]

Air displacement

[Humanoid Scott @GoingBallistic5]

It's All in the Hips

Can we learn anything from the evolution of hip kinematics in humanoid bots?

Is there a growing consensus of the optimal design?

Yes. And there is data to prove it.

First, a primer on Hip Kinematics

🧵

[Ilir Aliu @IlirAliu_]

First foundation model that zero-shot runs on any robot arm.

[📍 Fully open source]

What if robot arms could learn new tasks with zero data collection or fine-tuning?

That is what the team behind RDT2 just demonstrated.

RDT2 is the first foundation model that zero-shot deploys across robot arms, unseen scenes, objects, and instructions. Plug and play robotics…

No extra training.

✅ 10,000 hours of human manipulation data from 100 real homes with redesigned UMI hardware
✅ 7B one-step diffusion policy running at 23 Hz for real-time closed-loop control
✅ Zero-shot: pick, place, press, wipe, block arrows at 30 m/s, play ping-pong, extinguish burning incense

The result is a clear milestone:

models that generalize across embodiments and environments, moving toward embodied superintelligence.

Amazing work by the team, thanks for sharing, @songming_liu!

📍Project: https://t.co/KloV8tjPck

Code: https://t.co/Vf3HLDMRbP

[Siyuan Huang @siyuanhuang95]

Watching the G1 robot doing Webster flip by the side🙀. Such moments make humanoid robot research so much fun 😃😃