Tactile perception is vital, especially when distinguishing visually similar objects. We propose an approach to incorporate tactile data into a Vision-Language Model (VLM) for visuo-tactile zero-shot object recognition. Our approach leverages the zero-shot capability of VLMs to infer tactile properties from the names of tactilely similar objects. The proposed method translates tactile data into a textual description solely by annotating object names for each tactile sequence during training, making it adaptable to various contexts with low training costs. The proposed method was evaluated on the FoodReplica and Cube datasets, demonstrating its effectiveness in recognizing objects that are difficult to distinguish by vision alone.

In developing medical interventions using untethered milli- and microrobots, ensuring safety and effectiveness relies on robust methods for detection, real-time tracking, and precise localization within the body. However, the inherent non-transparency of the human body poses a significant obstacle, limiting robot detection primarily to specialized imaging systems such as X-ray fluoroscopy, which often lack crucial anatomical details. Consequently, the robot operator (human or machine) would encounter severe challenges in accurately determining the location of the robot and steering its motion. This study explores the feasibility of circumventing this challenge by creating a simulation environment that contains the precise digital replica (virtual twin) of a model microrobot operational workspace. Synchronizing coordinate systems between the virtual and real worlds and continuously integrating microrobot position data from the image stream into the virtual twin allows the microrobot operator to control navigation in the virtual world. We validate this concept by demonstrating the tracking and steering of a mobile magnetic robot in confined phantoms with high temporal resolution (< 100 ms, with an average of ~20 ms) visual feedback. Additionally, our object detection-based localization approach offers the potential to reduce overall patient exposure to X-ray doses during continuous microrobot tracking without compromising tracking accuracy. Ultimately, we address a critical gap in developing image-guided remote interventions with untethered medical microrobots, particularly for near-future applications in animal models and human patients.

This paper investigates the forecasting performance of COMEX copper futures realized volatility across various high-frequency intervals using both econometric volatility models and deep learning recurrent neural network models. The econometric models considered are GARCH and HAR, while the deep learning models include RNN (Recurrent Neural Network), LSTM (Long Short-Term Memory), and GRU (Gated Recurrent Unit). In forecasting daily realized volatility for COMEX copper futures with a rolling window approach, the econometric models, particularly HAR, outperform recurrent neural networks overall, with HAR achieving the lowest QLIKE loss function value. However, when the data is replaced with hourly high-frequency realized volatility, the deep learning models outperform the GARCH model, and HAR attains a comparable QLIKE loss function value. Despite the black-box nature of machine learning models, the deep learning models demonstrate superior forecasting performance, surpassing the fixed QLIKE value of HAR in the experiment. Moreover, as the forecast horizon extends for daily realized volatility, deep learning models gradually close the performance gap with the GARCH model in certain loss function metrics. Nonetheless, HAR remains the most effective model overall for daily realized volatility forecasting in copper futures.

Robots too small to see by eye have rapidly evolved in recent years thanks to the incorporation of on-board microelectronics. Semiconductor circuits have been used in microrobots capable of executing controlled wireless steering, prescribed legged gait patterns, and user-triggered transitions between digital states. Yet these promising new capabilities have come at the steep price of complicated fabrication. Even though circuit components can be reliably built by semiconductor foundries, currently available actuators for electronically integrated microrobots are built with intricate multi-step cleanroom protocols and use mechanisms like articulated legs or bubble generators that are hard to design and control. Here, we present a propulsion system for electronically integrated microrobots that can be built with a single step of lithographic processing, readily integrates with microelectronics thanks to low current/low voltage operation (1V, 10nA), and yields robots that swim at speeds over one body length per second. Inspired by work on micromotors, these robots generate electric fields in a surrounding fluid, and by extension propulsive electrokinetic flows. The underlying physics is captured by a model in which robot speed is proportional to applied current, making design and control straightforward. As proof, we build basic robots that use on-board circuits and a closed-loop optical control scheme to navigate waypoints and move in coordinated swarms. Broadly, solid-state propulsion clears the way for robust, easy to manufacture, electronically controlled microrobots that operate reliably over months to years.

This work presents a deep reinforcement learning-based approach to train controllers for the autonomous loading of ore piles with a Load-Haul-Dump (LHD) machine. These controllers must perform a complete loading maneuver, filling the LHD's bucket with material while avoiding wheel drift, dumping material, or getting stuck in the pile. The training process is conducted entirely in simulation, using a simple environment that leverages the Fundamental Equation of Earth-Moving Mechanics so as to achieve a low computational cost. Two different types of policies are trained: one with a hybrid action space and another with a continuous action space. The RL-based policies are evaluated both in simulation and in the real world using a scaled LHD and a scaled muck pile, and their performance is compared to that of a heuristics-based controller and human teleoperation. Additional real-world experiments are performed to assess the robustness of the RL-based policies to measurement errors in the characterization of the piles. Overall, the RL-based controllers show good performance in the real world, achieving fill factors between 71-94%, and less wheel drift than the other baselines during the loading maneuvers. A video showing the training environment and the learned behavior in simulation, as well as some of the performed experiments in the real world, can be found in https://youtu.be/jOpA1rkwhDY.

Fibrillar adhesion, observed in animals like beetles, spiders, and geckos, relies on nanoscopic or microscopic fibrils to enhance surface adhesion via 'contact splitting.' This concept has inspired engineering applications across robotics, transportation, and medicine. Recent studies suggest that functional grading of fibril properties can improve adhesion, but this is a complex design challenge that has only been explored in simplified geometries. While machine learning (ML) has gained traction in adhesive design, no previous attempts have targeted fibril-array scale optimization. In this study, we propose an ML-based tool that optimizes the distribution of fibril compliance to maximize adhesive strength. Our tool, featuring two deep neural networks (DNNs), recovers previous design results for simple geometries and introduces novel solutions for complex configurations. The Predictor DNN estimates adhesive strength based on random compliance distributions, while the Designer DNN optimizes compliance for maximum strength using gradient-based optimization. Our method significantly reduces test error and accelerates the optimization process, offering a high-performance solution for designing fibrillar adhesives and micro-architected materials aimed at fracture resistance by achieving equal load sharing (ELS).

Airships, best recognized for their unique quality of payload/energy ratio, present a fascinating challenge for the field of engineering. Their construction and operation require a delicate balance of materials and rules, making them a compelling object of study. They embody a distinct intersection of physics, design, and innovation, offering a wide array of possibilities for future transportation and exploration. Thanks to their long-flight endurance, they are suited for long-term missions. To operate in complex environments such as indoor cluttered spaces, their membrane and mechatronics need to be protected from impacts. This paper presents a new indoor airship design inspired by origami and the Kresling pattern. The airship structure combines a carbon fiber exoskeleton and UV resin micro-lattices for shock absorption. Our design strengthens the robot while granting the ability to access narrow spaces by folding the structure - up to a volume expansion ratio of 19.8. To optimize the numerous parameters of the airship, we present a pipeline for design, manufacture, and assembly. It takes into account manufacturing constraints, dimensions of the target deployment area, and aerostatics, allowing for easy and quick testing of new configurations. We also present unique features made possible by combining origami with airship design, which reduces the chances of mission-compromising failures. We demonstrate the potential of the design with a complete simulation including an effective control strategy leveraging lightweight mechatronics to optimize flight autonomy in exploration missions of unstructured environments.

Estimating the Most Important Person (MIP) in any social event setup is a challenging problem mainly due to contextual complexity and scarcity of labeled data. Moreover, the causality aspects of MIP estimation are quite subjective and diverse. To this end, we aim to address the problem by annotating a large-scale `in-the-wild' dataset for identifying human perceptions about the `Most Important Person (MIP)' in an image. The paper provides a thorough description of our proposed Multimodal Large Language Model (MLLM) based data annotation strategy, and a thorough data quality analysis. Further, we perform a comprehensive benchmarking of the proposed dataset utilizing state-of-the-art MIP localization methods, indicating a significant drop in performance compared to existing datasets. The performance drop shows that the existing MIP localization algorithms must be more robust with respect to `in-the-wild' situations. We believe the proposed dataset will play a vital role in building the next-generation social situation understanding methods. The code and data is available at https://github.com/surbhimadan92/MIP-GAF.

Materials design often relies on human-generated hypotheses, a process inherently limited by cognitive constraints such as knowledge gaps and limited ability to integrate and extract knowledge implications, particularly when multidisciplinary expertise is required. This work demonstrates that large language models (LLMs), coupled with prompt engineering, can effectively generate non-trivial materials hypotheses by integrating scientific principles from diverse sources without explicit design guidance by human experts. These include design ideas for high-entropy alloys with superior cryogenic properties and halide solid electrolytes with enhanced ionic conductivity and formability. These design ideas have been experimentally validated in high-impact publications in 2023 not available in the LLM training data, demonstrating the LLM's ability to generate highly valuable and realizable innovative ideas not established in the literature. Our approach primarily leverages materials system charts encoding processing-structure-property relationships, enabling more effective data integration by condensing key information from numerous papers, and evaluation and categorization of numerous hypotheses for human cognition, both through the LLM. This LLM-driven approach opens the door to new avenues of artificial intelligence-driven materials discovery by accelerating design, democratizing innovation, and expanding capabilities beyond the designer's direct knowledge.

A key challenge in artificial intelligence is the creation of systems capable of autonomously advancing scientific understanding by exploring novel domains, identifying complex patterns, and uncovering previously unseen connections in vast scientific data. In this work, we present SciAgents, an approach that leverages three core concepts: (1) the use of large-scale ontological knowledge graphs to organize and interconnect diverse scientific concepts, (2) a suite of large language models (LLMs) and data retrieval tools, and (3) multi-agent systems with in-situ learning capabilities. Applied to biologically inspired materials, SciAgents reveals hidden interdisciplinary relationships that were previously considered unrelated, achieving a scale, precision, and exploratory power that surpasses traditional human-driven research methods. The framework autonomously generates and refines research hypotheses, elucidating underlying mechanisms, design principles, and unexpected material properties. By integrating these capabilities in a modular fashion, the intelligent system yields material discoveries, critique and improve existing hypotheses, retrieve up-to-date data about existing research, and highlights their strengths and limitations. Our case studies demonstrate scalable capabilities to combine generative AI, ontological representations, and multi-agent modeling, harnessing a `swarm of intelligence' similar to biological systems. This provides new avenues for materials discovery and accelerates the development of advanced materials by unlocking Nature's design principles.