Imitation learning has shown promise in learning from large-scale real-world datasets. However, pretrained policies usually perform poorly without sufficient in-domain data. Besides, human-collected demonstrations entail substantial labour and tend to encompass mixed-quality data and redundant information. As a workaround, human-in-the-loop systems gather domain-specific data for policy post-training, and exploit closed-loop policy feedback to offer informative guidance, but usually require full-time human surveillance during policy rollout. In this work, we devise ARMADA, a multi-robot deployment and adaptation system with human-in-the-loop shared control, featuring an autonomous online failure detection method named FLOAT. Thanks to FLOAT, ARMADA enables paralleled policy rollout and requests human intervention only when necessary, significantly reducing reliance on human supervision. Hence, ARMADA enables efficient acquisition of in-domain data, and leads to more scalable deployment and faster adaptation to new scenarios. We evaluate the performance of ARMADA on four real-world tasks. FLOAT achieves nearly 95% accuracy on average, surpassing prior state-of-the-art failure detection approaches by over 20%. Besides, ARMADA manifests more than 4$\times$ increase in success rate and greater than 2$\times$ reduction in human intervention rate over multiple rounds of policy rollout and post-training, compared to previous human-in-the-loop learning methods.

We study distributed training of Graph Neural Networks (GNNs) on billion-scale graphs that are partitioned across machines. Efficient training in this setting relies on min-edge-cut partitioning algorithms, which minimize cross-machine communication due to GNN neighborhood sampling. Yet, min-edge-cut partitioning over large graphs remains a challenge: State-of-the-art (SoTA) offline methods (e.g., METIS) are effective, but they require orders of magnitude more memory and runtime than GNN training itself, while computationally efficient algorithms (e.g., streaming greedy approaches) suffer from increased edge cuts. Thus, in this work we introduce Armada, a new end-to-end system for distributed GNN training whose key contribution is GREM, a novel min-edge-cut partitioning algorithm that can efficiently scale to large graphs. GREM builds on streaming greedy approaches with one key addition: prior vertex assignments are continuously refined during streaming, rather than frozen after an initial greedy selection. Our theoretical analysis and experimental results show that this refinement is critical to minimizing edge cuts and enables GREM to reach partition quality comparable to METIS but with 8-65x less memory and 8-46x faster. Given a partitioned graph, Armada leverages a new disaggregated architecture for distributed GNN training to further improve efficiency; we find that on common cloud machines, even with zero communication, GNN neighborhood sampling and feature loading bottleneck training. Disaggregation allows Armada to independently allocate resources for these operations and ensure that expensive GPUs remain saturated with computation. We evaluate Armada against SoTA systems for distributed GNN training and find that the disaggregated architecture leads to runtime improvements up to 4.5x and cost reductions up to 3.1x.

--Dynamic and contact-rich object manipulation, such as striking, snatching, or hammering, remains challenging for robotic systems due to hardware limitations. Most existing robots are constrained by high-inertia design, limited compliance, and reliance on expensive torque sensors. T o address this, we introduce ARMADA (Affordable Robot for Manipulation and Dynamic Actions), a 6 degrees-of-freedom bimanual robot designed for dynamic manipulation research. ARMADA combines low-inertia, back-drivable actuators with a lightweight design, using readily available components and 3D-printed links for ease of assembly in research labs. The entire system, including both arms, is built for just $6,100. Each arm achieves speeds up to 6.16m/s, almost twice that of most collaborative robots, with a comparable payload of 2.5kg. We demonstrate ARMADA can perform dynamic manipulation like snatching, hammering, and bimanual throwing in real-world environments. We also showcase its effectiveness in reinforcement learning (RL) by training a non-prehensile manipulation policy in simulation and transferring it zero-shot to the real world, as well as human motion shadowing for dynamic bimanual object throwing. ARMADA is fully open-sourced with detailed assembly instructions, CAD models, URDFs, simulation, and learning codes. We highly recommend viewing the supplementary video at https://sites.google.com/view/im2-humanoid-arm. I NTRODUCTION Humans use a rich set of action repertoire to manipulate objects: we not only pick-and-place objects but also toss laundry, slide a box, snatch a pen, hammer a nail, otherwise arXiv:2502.16908v1 In contrast, most manipulators today are limited to picking, where a robot simply grasps an object to resist the frictional force. While kinematic pick-and-place is sufficient for static, controlled tasks, dynamic manipulation is necessary for building an effective general-purpose robot.

What would you say if I told you that an armada of highly advanced alien craft was en-route to Earth from some mysterious distant world, with an estimated arrival time in the next century -- perhaps in as little as 20 years. Would you grow frustrated with the lack of answers to questions like: Could we hold the aliens off, or appeal to them to keep our best interests at heart? The truth is, we know of no such armada, notwithstanding the current interest in UAPs (UFOs). Humanity faces a similar scenario, but the aliens will be created by us right here on Earth. Many prominent thinkers -- including top researchers in the field of A.I., or artificial intelligence -- believe that we are in the early stages of building our own successors, silicon-based life forms that -- from our limited human perspective -- could achieve god-like intelligence and power.

The maritime and scientific communities have set themselves the ambitious target of 2030 to map Earth's entire ocean floor. You can argue about the numbers but it's in the region of 80% of the global seafloor that's either completely unknown or has had no modern measurement applied to it. The international GEBCO 2030 project was set up to close the data gap and has announced a number of initiatives to get it done. What's clear, however, is that much of this work will have to leverage new technologies or at the very least max the existing ones. Which makes the news from Ocean Infinity - that it's creating a fleet of ocean-going robots - all the more interesting. US-based OI is a relatively new exploration and survey company.

A fleet of 11 uncrewed vessels will traverse the world's oceans over the next ten years in a bid to map the sea floor. The bottom of the world's oceans remains a mystery, with around 80 per cent either poorly imaged or not visualised at all. Ocean Infinity launched in 2016 and has pledged its support to an international collaboration to try and map every inch of the ocean floor within the next decade. It has also attempted to use its technology to try and locate the missing Malaysian Airlines MH370 flight that tragically went missing with 239 people on board nearly six years ago. It has announced it has bought a fleet of 11 uncrewed vessels will traverse the world's oceans over the next ten years in a bid to map the sea floor Uncrewed Surface Vessels (USV) are the latest technology which open up the possibility for long-term marine missions. They have no humans on board and are controlled by computers via a satellite link and a central computer base.