Amazing stuff! Very impressive!
Showing posts with label robotics. Show all posts
Showing posts with label robotics. Show all posts
Wednesday, September 02, 2026
The $3 Billion Chinese Startup Solving Elon Musk’s Robotic dexterity Hands Problem
Tuesday, September 01, 2026
Elon Musk Warns Against AI Regulation, Predicts 1 Billion Humanoid Robots in 10 years and 5 times more productive than humans at G20 Ministerial
Recommendable! Elon looks a little bit old!
He makes some startling predictions e.g. that by next year AI software development will beat any human.
He thinks the energy shortfall in AI is very serious and may become very obvious as soon as next year.
TrAct: Bridging Robot Control and Visual Prediction with Visual Tracks
This could be an interesting new paper by Li Fei Fei and her team!
From the abstract:
"Robot actions are inherently embodiment-specific and only weakly aligned with image-space visual changes, limiting their effectiveness as conditioning signals for robot world models.
In contrast, visual tracks provide an embodiment-agnostic representation of how task-relevant points move through a scene, offering dense image-space guidance for accurate and spatially precise future video prediction.
Building on this observation, we propose TrAct, a world-model-based robot decision-making framework that uses visual tracks as an intermediate interface between control and prediction.
TrAct consists of three components:
a Vision-Language-Action-and-Track model (VLAT) that jointly predicts candidate actions and corresponding visual tracks from the current observation and language instruction;
a track-conditioned world model (TWM) that predicts future visual outcomes conditioned on the proposed tracks; and
a vision-language reward model (VLAC) that scores the predicted outcomes.
At inference time, VLAT generates candidate action-track pairs, TWM rolls out their visual consequences, and VLAC selects the track whose predicted outcome best satisfies the instruction; the action paired with the selected track is then executed by the robot.
Experiments on the proposed LIBERO-INTEGRAL benchmark and real-world Franka manipulation show that TrAct improves success rates from 27% to 55% in simulation and from 49% to 76% on real-world tasks compared with the strong VLA baseline π0.5.
Furthermore, TWM consistently improves video prediction quality over the action-conditioned world model (AWM).
These results demonstrate that visual tracks provide an effective shared interface between robot control and visual prediction, enabling more accurate world modeling and stronger robot generalization."
Monday, August 31, 2026
ICE is considering spending up to $2 million on Boston Dynamics’ Spot robot dogs
Good spot on news! I suppose, it is too expensive, too dangerous, and takes too long to train real dogs!
"... for hazardous inspections and situational awareness, using technology the Secret Service deployed at Mar-a-Lago in 2024."
".. The Boston Dynamic SPOT Robots will help "support public safety and law enforcement operations by providing a remotely operated robotic capability for inspection, situational awareness, and hazard assessment in environments that may pose risks to personnel," according to the notice. ..."
ICE plans to purchase robot dogs "The move comes as the agency has also sought to purchase electric shock gloves."
Spot robot dog (Source)
Wednesday, August 26, 2026
Bedrock Robotics' first operator-free retrofitted excavator deployments take off at job sites
Good news!
"Bedrock Robotics Inc. has deployed fully autonomous, retrofitted heavy excavators onto active infrastructure projects in Texas and Nevada. These AI-driven heavy equipment projects directly address the construction industry’s severe labor shortages, safety hazards, and cost overruns. ..."
"... The excavators use Bedrock’s Operator system, which retrofits existing fleets with a sophisticated sensor and compute suite to perform excavation without onboard operators, according to the company. ..."
Bedrock Robotics deploys fully autonomous excavators on jobsites "The company said its retrofit tech, which digs without an operator, is now active on infrastructure projects for firms such as Sundt Construction and Zachry Construction."
Monday, August 24, 2026
World Humanoid Robot Games in Beijing 2026 (only the second games ever since 2025)
Very impressive results so far! There were also some other interesting results reported from these games.
"A humanoid robot named Tiangong Ultra ran the 100-meter dash in 9.39 seconds Saturday at the World Humanoid Robot Games in Beijing, beating Usain Bolt’s human world record of 9.58 seconds, set in 2009. ...
Another X-Humanoid machine cleared 2.88 meters in the high jump, well above the 2.45-meter human record. More than 2,000 robots from 16 countries are competing in 51 events, including soccer, boxing, and tug of war ..."
Sunday, August 23, 2026
California regulators have approved a major expansion of Waymo’s driverless taxi service across parts of 18 counties including new areas
Good news!
"... The company is now authorized to offer paid robotaxi rides across parts of 18 counties, including new areas around Sacramento, San Diego, Orange County, Napa, and the East Bay of San Francisco."
The US Food and Drug Administration (FDA) has authorized a standalone Dutch robot that can draw blood
Maybe the first approved vampire robot? 😊 And the FDA released the announcement on my birthday! 😊
"... The machine, called Aletta, uses near-infrared light and ultrasound to locate a vein before autonomously applying a tourniquet, disinfecting the skin, inserting the needle, changing collection tubes, and applying a bandage.
The authorization will allow Aletta to be used on adults in clinics and other medical settings where patients do not stay overnight."
"... The device is authorized for use in adults in outpatient settings and must be operated under the oversight of a supervisor trained in phlebotomy.
One phlebotomist can oversee up to three Aletta devices at the same time ..."
"... The Aletta ARPD is designed to autonomously perform diagnostic blood collection. The system combines multimodal imaging, including near-infrared, ultrasound, and Doppler ultrasound to help distinguish between veins and arteries. It combines these capabilities with advanced robotics and artificial intelligence, to identify suitable veins, guide needle insertion, and collect blood samples with high precision and consistency. ..."
US Authorizes First Standalone Robotic Blood Draw Device
FDA Authorizes First-Of-Its-Kind Robotic Blood Draw Device (original news release) "Authorization supports innovative approach to blood collection; may help address growing phlebotomist shortage"
Discover how Aletta® sets a new standard for blood collection "Aletta, the world’s first Autonomous Robotic Phlebotomy Device (ARPD), automates the diagnostic blood collection process."
Vitestro Receives FDA De Novo Authorization for Aletta®, the World’s First Autonomous Robotic Phlebotomy Device (original news release)
Aletta
Saturday, August 22, 2026
Chinese robot smashes Usain Bolt’s 100m record at World Robot Games
Very impressive! But humans do not need stretchers or fire extinguishers after the race. 😊
Robots learn new skills from a single video—in just 29 seconds
This seems to be an interesting new Chinese paper on robotics!
Did these researchers successfully overcome catastrophic learning of previously learned skills?
"... Researchers ... recently developed HOST (Human-to-robot One-Shot Skill AcquisiTion), a new framework that could allow robots to acquire new skills faster and more efficiently. Their proposed learning approach ... allows a robot to acquire a new skill from a single video showing a human demonstration without compromising previously acquired abilities. ..."
"...
Highlights
- Inference-time skill acquisition: one human video, no fine-tuning, and no parameter update.
- Fast acquisition: 29 seconds per novel skill on average, including recording the demonstration.
- Broad real-robot evaluation: acquires executable skills across 50 novel manipulation tasks, each evaluated over 20 trials.
- Strong novel-task performance: achieves 62% average success on the task subset used for baseline comparisons.
- Data and time efficiency: 50 times fewer demonstrations and 507 times faster acquisition than the strongest task-specific fine-tuning baseline evaluated in the paper.
- Skill retention: new skills are supplied through external video context rather than written into the shared policy weights.
- Robust execution: evaluated under lighting changes, unseen objects, scene replacement, and human disturbances during execution.
..."
From the abstract:
"The ability to acquire skills rapidly and effortlessly while retaining those already mastered is essential for robots. However, current methods still rely on a cumbersome training-time loop that is costly and slow, while eroding skills already mastered.
In this paper, we introduce HOST (Human-to-robot One-Shot Skill AcquisiTion), a framework that enables a robot to acquire skills in seconds from a single human video while retaining previously mastered skills.
HOST resolves skill acquisition through a cascade of self-grounded prediction.
It first estimates the robot's progress within the demonstrated task, then translates the upcoming progression into the robot's own future observations, and finally derives actions from these predicted observations.
This cascade is trained on targets coupled to the video demonstration, obtained by mapping the robot trajectory and the video demonstration onto a shared task progress manifold, then redefining each target to align with the future progression of the video.
HOST thereby enables the robot to actively follow the demonstrated procedure and adapt it to the robot's embodiment.
HOST acquires novel skills at inference time from a single human video in an average of 29 seconds and achieves a 62% average success rate. It exceeds the zero-shot baseline by 45% while retaining previously mastered skills.
HOST even exceeds the baseline fine-tuned on 50 robot demonstrations per task while requiring 50 times fewer demonstrations and acquiring each skill 507 times faster. ..."
Project website
Robots Acquire Manipulation Skills in Seconds from a Single Human Video (preprint, open access)
Thursday, August 20, 2026
What does Chinese Unitree Robotics' IPO mean for the humanoid industry?
Another Sputnik shock!
What are some of the competitive advantages of the US: Very competitive and rapid technological innovation and discovery as well as deep and less risk averse financial markets. Is China catching up?
"Shares of Unitree Robotics, China's high-profile humanoid robot maker known for fluid athletic maneuvers, shot up more than 629% to 1,100 yuan apiece upon their debut in Shanghai on Wednesday, after the company raised 6.1 billion yuan ($905 million).
The listing, at a valuation of $9 billion, marks China's first onshore initial public offering by a humanoid robot company. Dozens of rivals, such as Agibot, are reportedly preparing to list in Hong Kong.
The Hangzhou-based company is backed by some of China's most influential technology companies, including Tencent, Alibaba, Meituan and DeepSeek.
State media reported that China is now home to more than 140 humanoid robot companies. Players from a range of industries, particularly electric vehicle makers, are rushing into the sector. Some, such as Xpeng, have announced plans to begin mass production this year, while Chinese EV giant BYD is expected to unveil its first humanoid robot this month.
China's humanoid robots have been drawing growing attention with increasingly sophisticated public demonstrations. ..."
"... Since it was founded in 2016, Unitree has made headlines for its low-cost quadruped and humanoid robots. The Hangzhou, China-based company, formally Yushu Technology Co., is best known for its humanoids, which are used in research labs around the world.
You’ve likely seen videos of Unitree’s G1 robot dancing, doing kung fu, or backflipping, among other things. In addition, the company offers robotic hands, arms, and lidar. ..."
Monday, August 17, 2026
Amazon inks global deal for warehouse robotics with AutoStore
Good news!
"Amazon has entered a strategic global agreement with Norwegian technology group AutoStore to enhance its warehouse automation capabilities. This partnership allows AutoStore to supply its advanced automation systems to Amazon worldwide, although no specific purchasing commitments have been disclosed. The deal aligns with Amazon's ongoing efforts to integrate robotics into its fulfillment processes, aiming to improve delivery speed and reduce costs."
"... AutoStore systems were also being considered for Amazon micro-fulfilment projects as the retailer experiments with increasingly localised fulfilment models. ..."
AutoStore Holdings Ltd – Strategic supply agreement with Amazon (original news release)
Sunday, August 16, 2026
Global shipments of humanoid robots grew to 19,100 units in the first half of 2026 from 5,100 in the first half of 2025
Good news!
"... nearly four times the number shipped during the same period last year. The same report predicts that total shipments will reach 60,000 by the end of the year."
"... China’s AgiBot has surpassed its domestic rival Unitree Robotics to become the world’s largest humanoid robot vendor in the first half of 2026, according to new research, as both industry heavyweights prepare for public listings amid a physical AI boom.
Shanghai-based Agibot captured 44 per cent of the global market after shipping roughly 8,400 humanoid robots from January to June, San Francisco-based research firm Smart Analytics Global (SAG) said in a report published on Sunday. ..."
AgiBot overtakes Unitree as top global humanoid robot vendor in first half amid IPO push "AgiBot’s shipments surged 562 per cent year on year, supported by a portfolio of full-size bipedals, compact units and wheeled robots"
Thursday, August 13, 2026
ZESTy robots do the cartwheel and other feats
Amazing stuff! This could be a breakthrough!
The researchers of this paper are with Robotics and AI Institute, which seems to be in a sort of stealth mode as it appears it has not published anything recently on ZEST. E.g. searching for ZEST on their website results in nothing. None of the authors of this paper are familiar to me.
Notice only about 7 months elapsed between preprint date and journal article date. This is unusually fast.
From the abstract:
"Achieving robust, humanlike whole-body control on humanoid robots for agile, contact-rich behaviors remains a central challenge, demanding heavy per-skill engineering and a brittle process of tuning controllers.
We introduce ZEST (zero-shot embodied skill transfer), a streamlined motion-imitation framework that trains policies via reinforcement learning from diverse sources—high-fidelity motion capture, noisy monocular video, and non–physics-constrained animation—and deploys them to hardware zero-shot.
ZEST generalizes across behaviors and platforms without relying on contact labels, reference or observation windows, state estimators, or extensive reward shaping.
Its training pipeline combines adaptive sampling, which focuses training on difficult motion segments, and an automatic curriculum using a model-based assistive wrench, together enabling dynamic, long-horizon maneuvers.
We further provide a procedure for selecting joint-level gains from approximate analytical armature values for closed-chain actuators, along with a refined model of actuators. Trained entirely in simulation with moderate domain randomization, ZEST demonstrated broad generality.
On Boston Dynamics’ Atlas humanoid, ZEST learned dynamic, multicontact skills (army crawl and breakdancing) from motion capture. It transferred expressive dance and scene-interaction skills, such as box climbing, directly from videos to Atlas and the Unitree G1.
Furthermore, it extended across morphologies to the Spot quadruped, enabling acrobatics, such as a continuous backflip, through animation.
Together, these results demonstrate robust zero-shot deployment across heterogeneous data sources and embodiments, establishing ZEST as a scalable interface between biological movements and their robotic counterparts."
ZEST: Zero-shot embodied skill transfer for athletic robot control (no public access)
ZEST: Zero-shot Embodied Skill Transfer for Athletic Robot Control (preprint, open access)
Robot Recycler Salvages Parts From Broken or end of life Robots
Good news! This seems to be still early, preliminary work!
The famous Austrian economist Joseph Schumpeter coined the term "creative destruction" of free markets! The coming and going of new businesses!
The Achilles heel of this system seems to be the requirement to work off CAD data of the broken robot. What if the CAD data is e.g. not available?
"... There are over 4 million industrial robots in use worldwide, according to the International Federation of Robotics. And researchers predict that number will grow to over 16 million by 2030, as manufacturing rapidly increases. But what’s going to happen when they start breaking down?
A new system designed by researchers at the Karlsruhe Institute of Technology (KIT), in Karlsruhe, Germany, can predict the defect in a broken product and disassemble it while protecting valuable parts from damage. To continue robotic development sustainably, the industry should prepare for the dismantling, recycling, and rebuilding of our robotic systems. ...
KIT’s robotic disassembly system relies on a CAD model of the broken product and of each part, so it can see how the parts should behave and understand if anything is out of the ordinary. It also uses a mathematical model to predict the damage done to a broken part.
When you give the system a broken device and a CAD model, it first guesses how each part of the broken device should move. The axes each part can move along are called degrees of freedom (for example, a screw should rotate, but not move side to side). The disassembler nudges each part to see if it moves as expected. Based on how the part actually moves, it then uses the mathematical model to predict what went wrong with the part: A corroded part might move less than you think it should, a loose screw may move more, and a deformed part might have different degrees of freedom than expected.
At the beginning of disassembly, the system formulates a plan. It guesses what might be wrong with the device it’s taking apart, and then can change its guess based on observing each piece it takes apart. ..."
From the abstract:
"To support the circular economy, robotic systems must not only assemble new products but also disassemble end-of-life (EOL) ones for reuse, recycling, or safe disposal.
Existing approaches to disassembly sequence planning often assume deterministic and fully observable product models, yet real EOL products frequently deviate from their initial designs due to wear, corrosion, or undocumented repairs.
We argue that disassembly should therefore be formulated as a Partially Observable Markov Decision Process (POMDP), which naturally captures uncertainty about the product's internal state. We present a mathematical formulation of disassembly as a POMDP, in which hidden variables represent uncertain structural or physical properties.
Building on this formulation, we propose a task and motion planning framework that automatically derives specific POMDP models from CAD data, robot capabilities, and inspection results.
To obtain tractable policies, we approximate this formulation with a reinforcement-learning approach that operates on stochastic action outcomes informed by inspection priors, while a Bayesian filter continuously maintains beliefs over latent EOL conditions during execution.
Using three products on two robotic systems, we demonstrate that this probabilistic planning framework outperforms deterministic baselines in terms of average disassembly time and variance, generalizes across different robot setups, and successfully adapts to deviations from the CAD model, such as missing or stuck parts."
From CAD to POMDP: Probabilistic Planning for Robotic Disassembly of End-of-Life Products (preprint, open access)
Tuesday, August 11, 2026
Why Does Action Chunking Improve Behavioral Cloning Performance in Robotic Control?
This could be an interesting, new paper by Sergey Levine and his team!
From the abstract:
"Action chunking---predicting and executing multiple actions instead of a single action---has proven to be a critical component for learning effective robotic control policies.
However, our precise understanding of why action chunking improves performance has remained limited.
In this work we seek to close this gap. Through rigorous experimental evaluations in both simulated and real-world settings, we show that existing hypotheses for the success of action chunking---temporal consistency, horizon reduction, and representation learning---fail to explain the success of action chunking.
Instead, we find that action chunking benefits from greater non-Markovian expressivity and reduced compounding error compared to Markovian policies, but, in many settings of interest, these effects can be fully captured by delayed policies, which at each step predict a single action based on the observation k steps in the past.
We then show that there exists an additional benefit of action chunking that we refer to as implicit ensembling. In particular, by learning a diversity of temporal relationships (that is, at|ot,at|ot−1,…), action-chunked policies exhibit behavior matching that of a model ensemble, increasing their robustness and generalization ability over policies that only learn a single temporal relationship.
Building on these insights, we show that in simulated and real-world robotic control settings, we can match the performance of action chunking without action chunking---by deploying an action chunking policy as an ensemble of policies with randomized delays.
Furthermore, we propose a policy class that amplifies the benefits of action chunking by explicitly instantiating an ensemble, and which we show significantly improves over the performance of action chunking in many domains."
Friday, August 07, 2026
Americans Can’t Buy/afford Chinese EVs. Waymo Is Importing Thousands
Recommendable! However, the most likely explanation is that Waymo was able to buy these vehicles at a significant discount.
I have already seen several of these new Waymo cars driving around in the Phoenix, Arizona metro area. I wondered what they were.
Wednesday, August 05, 2026
Humanoide Roboter: Was sich jetzt für Unternehmen ändert
Neues aus der Bananenrepublik Deutschland! Kann Deutschland aufholen was Roboter angeht?
"... Das neue Faktenblatt „Humanoide Roboter – Entwicklungsstand, Potenziale und Zukunftsaussichten Humanoider Roboter für die industrielle Produktion“ vom ifaa – Institut für angewandte Arbeitswissenschaft e. V. ordnet den Stand der Technik ein, zeigt Einsatzfelder und benennt offene Fragen für die betriebliche Einführung. ..."
"Kernaussagen
Entwicklungsstand von Humanoiden
Humanoide Roboter befinden sich im Übergang von der Forschung zur industriellen Anwendung. Auf den voraussichtlich stark wachsenden Markt drängen viele Hersteller aus den USA und China. Europa und Deutschland sind beteiligt, u. a. durch Neura Robotics.
Einsatz zunächst in Logistik und Produktion
Pilotprojekte in Automobilbau und Logistik zeigen die industrielle Einsatzfähigkeit. Hier werden Humanoide für repetitive Tätigkeiten und in Bereichen mit hoher Fluktuation genutzt. Fortschritte bei KI, Trainingsmethoden und Hardware erweitern ihre Flexibilität und Einsatzmöglichkeiten.
Große Flexibilität und hoher Nutzen erwartet
Steigende Einsatzflexibilität und sinkende Preise lassen ein hohes wirtschaftliches Nutzenpotenzial für Humanoide Roboter erwarten – insbesondere in kleinen und mittleren Unternehmen (KMU), die klassische, weniger flexible Robotik häufig weder finanzieren noch angemessen nutzen können.
Zukunftsaussichten für Humanoide
Humanoide werden bislang vor allem von internationalen Großunternehmen erprobt und für spezifische Aufgaben eingesetzt. Eine flexible Nutzung – insbesondere in KMU – ist bisher kaum erprobt, jedoch notwendig, um das damit verbundene breite wirtschaftliche Potenzial zu erschließen. ..."
Tuesday, August 04, 2026
Roboter-Start-up Humanoid wird zum Einhorn – auch Bosch und Schaeffler steigen ein
Auch die Bananenrepublik D hat Einhorns! Na sowas! 😊
Monday, August 03, 2026
Making robots faster by helping them think ahead
This seems to be an interesting new paper (however, this paper was first published in November 2025)!
"A new method developed by ... researchers makes robots better at thinking ahead while they are acting, leading to smoother motions and quicker reactions.
This technique enables the artificial intelligence model that plans a robot’s motion to forecast its future position. The model uses this prediction to seamlessly transition current movements into the next actions.
Many existing methods cause a robot to stop and think about what it needs to do next, leading to slow and jerky motions. By basing its calculations on the future state of the robot, rather than its current position, the ... method helps robots operate much faster. ..."
From the abstract:
"Vision-Language-Action models (VLAs) are becoming increasingly capable across diverse robotic tasks. However, these models are typically deployed under synchronous inference, where the robot waits for model inference to complete before acting, and cannot perceive or respond to environmental changes during action execution. This not only introduces noticeable action stalls, but also significantly increases reaction latency, fundamentally limiting the applicability of VLAs to dynamic, real-time tasks.
Asynchronous inference offers a promising solution to achieve continuous and low-latency control by enabling robots to execute actions and perform inference simultaneously.
However, because the robot and environment continue to evolve during inference, a temporal misalignment arises between the prediction and execution intervals. This leads to significant action instability, while existing asynchronous methods either degrade accuracy or introduce runtime overhead to mitigate it.
We propose VLASH, a simple yet effective method for asynchronous VLA inference that delivers smooth, accurate, and fast reaction control without architectural changes or additional runtime overhead.
VLASH leverages the future execution-time state by rolling the robot state forward with the previous action chunk, thereby bridging the gap between prediction and execution.
Experiments show that VLASH reduces reaction latency by up to 11.8x compared to synchronous inference and consistently outperforms all asynchronous baselines in accuracy.
With action quantization, it further achieves 1.5-2.0x task completion speedup with minimal accuracy loss.
Moreover, it empowers state-of-the-art VLAs such as to handle fast-reaction, high-precision tasks including playing ping-pong and playing whack-a-mole, where traditional synchronous inference fails. ..."
VLASH: Real-Time VLAs via Future-State-Aware Asynchronous Inference (preprint, open access)
Subscribe to:
Posts (Atom)