Supernumerary Robotic Limbs (SRLs) can enhance human capability within close proximity. However, as a wearable device, the generated moment from its operation acts on the human body as an external torque. When the moments increase, more muscle units are activated for balancing, and it can result in reduced muscular null space. Therefore, this paper suggests a concept of a motion planning layer that reduces the generated moment for enhanced Human-Robot Interaction. It modifies given trajectories with desirable angular acceleration and position deviation limits. Its performance to reduce the moment is demonstrated through the simulation, which uses simplified human and robotic system models.
We present a machine learning framework for adaptive source localization in which agents use a structure-preserving digital twin of a coupled hydrodynamic-transport system for real-time trajectory planning and data assimilation. The twin is constructed with conditional neural Whitney forms (CNWF), coupling the numerical guarantees of finite element exterior calculus (FEEC) with transformer-based operator learning. The resulting model preserves discrete conservation, and adapts in real time to streaming sensor data. It employs a conditional attention mechanism to identify: a reduced Whitney-form basis; reduced integral balance equations; and a source field, each compatible with given sensor measurements. The induced reduced-order environmental model retains the stability and consistency of standard finite-element simulation, yielding a physically realizable, regular mapping from sensor data to the source field. We propose a staggered scheme that alternates between evaluating the digital twin and applying Lloyd's algorithm to guide sensor placement, with analysis providing conditions for monotone improvement of a coverage functional. Using the predicted source field as an importance function within an optimal-recovery scheme, we demonstrate recovery of point sources under continuity assumptions, highlighting the role of regularity as a sufficient condition for localization. Experimental comparisons with physics-agnostic transformer architectures show improved accuracy in complex geometries when physical constraints are enforced, indicating that structure preservation provides an effective inductive bias for source identification.
Extreme precipitation events occurring over large spatial domains pose substantial threats to societies because they can trigger compound flooding, landslides, and infrastructure failures across wide areas. A hybrid framework for spatial extreme precipitation modeling and risk zoning is proposed that integrates graph neural networks with r-Pareto processes (GNN-rP). Unlike traditional statistical spatial extremes models, this approach learns nonlinear, nonstationary dependence structures from precipitation-derived spatial graphs and applies a data-driven tail functional to model joint exceedances in a low-dimensional embedding space. Using NASA's IMERG observations (2000-2021) and CMIP6 SSP5-8.5 projections, the framework delineates coherent high-risk zones, quantifies their temporal persistence, and detects emerging hotspots under climate change. Compared with two baseline approaches, the GNN-rP pipeline substantially improves pointwise detection of high-risk grid cells while yielding comparable clustering stability. Results highlight persistent high-risk regions in the tropical belt, especially monsoon and convective zones, and reveal decadal-scale persistence that is punctuated by episodic reconfigurations under high-emission scenarios. By coupling machine learning with extreme value theory, GNN-rP offers a scalable, interpretable tool for adaptive climate risk zoning, with direct applications in infrastructure planning, disaster preparedness, and climate-resilient policy design.
Unmanned aerial vehicles (UAVs) with suspended payloads offer significant advantages for aerial transportation in complex and cluttered environments. However, existing systems face critical limitations, including unreliable perception of the cable-payload dynamics, inefficient planning in large-scale environments, and the inability to guarantee whole-body safety under cable bending and external disturbances. This paper presents Acetrans, an Autonomous, Corridor-based, and Efficient UAV suspended transport system that addresses these challenges through a unified perception, planning, and control framework. A LiDAR-IMU fusion module is proposed to jointly estimate both payload pose and cable shape under taut and bent modes, enabling robust whole-body state estimation and real-time filtering of cable point clouds. To enhance planning scalability, we introduce the Multi-size-Aware Configuration-space Iterative Regional Inflation (MACIRI) algorithm, which generates safe flight corridors while accounting for varying UAV and payload geometries. A spatio-temporal, corridor-constrained trajectory optimization scheme is then developed to ensure dynamically feasible and collision-free trajectories. Finally, a nonlinear model predictive controller (NMPC) augmented with cable-bending constraints provides robust whole-body safety during execution. Simulation and experimental results validate the effectiveness of Acetrans, demonstrating substantial improvements in perception accuracy, planning efficiency, and control safety compared to state-of-the-art methods.
In dynamic and uncertain environments, robotic path planning demands accurate spatiotemporal environment understanding combined with robust decision-making under partial observability. However, current deep reinforcement learning-based path planning methods face two fundamental limitations: (1) insufficient modeling of multi-scale temporal dependencies, resulting in suboptimal adaptability in dynamic scenarios, and (2) inefficient exploration-exploitation balance, leading to degraded path quality. To address these challenges, we propose GundamQ: A Multi-Scale Spatiotemporal Q-Network for Robotic Path Planning. The framework comprises two key modules: (i) the Spatiotemporal Perception module, which hierarchically extracts multi-granularity spatial features and multi-scale temporal dependencies ranging from instantaneous to extended time horizons, thereby improving perception accuracy in dynamic environments; and (ii) the Adaptive Policy Optimization module, which balances exploration and exploitation during training while optimizing for smoothness and collision probability through constrained policy updates. Experiments in dynamic environments demonstrate that GundamQ achieves a 15.3\% improvement in success rate and a 21.7\% increase in overall path quality, significantly outperforming existing state-of-the-art methods.
The goal of Multi-Agent Path Finding (MAPF) is to find a set of paths for a fleet of agents moving in a shared environment such that the agents reach their goals without colliding with each other. In practice, some of the robots executing the plan may get delayed, which can introduce collision risk. Although robust execution methods are used to ensure safety even in the presence of delays, the delays may still have a significant impact on the duration of the execution. At some point, the accumulated delays may become significant enough that instead of continuing with the execution of the original plan, even if it was optimal, there may now exist an alternate plan which will lead to a shorter execution. However, the problem is how to decide when to search for the alternate plan, since it is a costly procedure. In this paper, we propose a holistic architecture for robust execution of MAPF plans, its monitoring and optimization. We exploit a robust execution method called Action Dependency Graph to maintain an estimate of the expected execution duration during the plan's execution. This estimate is used to predict the potential that finding an alternate plan would lead to shorter execution. We empirically evaluate the architecture in experiments in a real-time simulator which we designed to mimic our real-life demonstrator of an autonomous warehouse robotic fleet.
The planned IceCube-Gen2 radio neutrino detector at the South Pole will enhance the detection of cosmic ultra-high-energy neutrinos. It is crucial to utilize the available time until construction to optimize the detector design. A fully differentiable pipeline, from signal generation to detector response, would allow for the application of gradient descent techniques to explore the parameter space of the detector. In our work, we focus on the aspect of signal generation, and propose a modularized deep learning architecture to generate radio signals from in-ice neutrino interactions conditioned on the shower energy and viewing angle. The model is capable of generating differentiable signals with amplitudes spanning multiple orders of magnitude, as well as consistently producing signals corresponding to the same underlying event for different viewing angles. The modularized approach ensures physical consistency of the samples and leads to advantageous computational properties when using the model as part of a bigger optimization pipeline.
Online planning in Markov Decision Processes (MDPs) enables agents to make sequential decisions by simulating future trajectories from the current state, making it well-suited for large-scale or dynamic environments. Sample-based methods such as Sparse Sampling and Monte Carlo Tree Search (MCTS) are widely adopted for their ability to approximate optimal actions using a generative model. However, in practical settings, the generative model is often learned from limited data, introducing approximation errors that can degrade performance or lead to unsafe behaviors. To address these challenges, Robust MDPs (RMDPs) offer a principled framework for planning under model uncertainty, yet existing approaches are typically computationally intensive and not suited for real-time use. In this work, we introduce Robust Sparse Sampling (RSS), the first online planning algorithm for RMDPs with finite-sample theoretical performance guarantees. Unlike Sparse Sampling, which estimates the nominal value function, RSS computes a robust value function by leveraging the efficiency and theoretical properties of Sample Average Approximation (SAA), enabling tractable robust policy computation in online settings. RSS is applicable to infinite or continuous state spaces, and its sample and computational complexities are independent of the state space size. We provide theoretical performance guarantees and empirically show that RSS outperforms standard Sparse Sampling in environments with uncertain dynamics.
Accurately simulating soft tissue deformation is crucial for surgical training, pre-operative planning, and real-time haptic feedback systems. While physics-based models such as the finite element method (FEM) provide high-fidelity results, they are often computationally expensive and require extensive preprocessing. We propose a graph neural network (GNN) architecture that predicts both tissue surface deformation and applied force from sparse point clouds. The model incorporates internal anatomical information through binary tissue profiles beneath each point and leverages E(n)-equivariant message passing to improve robustness. We collected experimental data that comprises a real silicone and bone-like phantom, and complemented it with synthetic simulations generated using FEM. Our model achieves a comparable performance to a baseline GNN on standard test cases and significantly outperforms it in rotated and cross-resolution scenarios, showing a strong generalization to unseen orientations and point densities. It also achieves a significant speed improvement, offering a solution for real-time applications. When fine-tuned on experimental data, the model maintains sub-millimeter deformation accuracy despite limited sample size and measurement noise. The results demonstrate that our approach offers an efficient, data-driven alternative to traditional simulations, capable of generalizing across anatomical configurations and supporting interactive surgical environments.
Built environment, formed of a plethora of patterns of building, streets, and plots, has a profound impact on how cities are perceived and function. While various methods exist to classify urban patterns, they often lack a strong theoretical foundation, are not scalable beyond a local level, or sacrifice detail for broader application. This paper introduces the Hierarchical Morphotope Classification (HiMoC), a novel, theory-driven, and computationally scalable method of classification of built form. HiMoC operationalises the idea of a morphotope - the smallest locality with a distinctive character - using a bespoke regionalisation method SA3 (Spatial Agglomerative Adaptive Aggregation), to delineate contiguous, morphologically distinct localities. These are further organised into a hierarchical taxonomic tree reflecting their dissimilarity based on morphometric profile derived from buildings and streets retrieved from open data, allowing flexible, interpretable classification of built fabric, that can be applied beyond a scale of a single country. The method is tested on a subset of countries of Central Europe, grouping over 90 million building footprints into over 500,000 morphotopes. The method extends the capabilities of available morphometric analyses, while offering a complementary perspective to existing large scale data products, which are focusing primarily on land use or use conceptual definition of urban fabric types. This theory-grounded, reproducible, unsupervised and scalable method facilitates a nuanced understanding of urban structure, with broad applications in urban planning, environmental analysis, and socio-spatial studies.
Efficient planning and sequence selection are central to intelligence, yet current approaches remain largely incompatible with biological computation. Classical graph algorithms like Dijkstra's or A* require global state and biologically implausible operations such as backtracing, while reinforcement learning methods rely on slow gradient-based policy updates that appear inconsistent with rapid behavioral adaptation observed in natural systems. We propose a biologically plausible algorithm for shortest-path computation that operates through local spike-based message-passing with realistic processing delays. The algorithm exploits spike-timing coincidences to identify nodes on optimal paths: Neurons that receive inhibitory-excitatory message pairs earlier than predicted reduce their response delays, creating a temporal compression that propagates backwards from target to source. Through analytical proof and simulations on random spatial networks, we demonstrate that the algorithm converges and discovers all shortest paths using purely timing-based mechanisms. By showing how short-term timing dynamics alone can compute shortest paths, this work provides new insights into how biological networks might solve complex computational problems through purely local computation and relative spike-time prediction. These findings open new directions for understanding distributed computation in biological and artificial systems, with possible implications for computational neuroscience, AI, reinforcement learning, and neuromorphic systems.
We introduce and study a novel generalization of the classical Bin Packing Problem (BPP), called the Bin Packing Problem with Setups (BPPS). In this problem, which has many practical applications in production planning and logistics, the items are partitioned into classes and, whenever an item from a given class is packed into a bin, a setup weight and cost are incurred. We present a natural Integer Linear Programming (ILP) formulation for the BPPS and analyze the structural properties of its Linear Programming relaxation. We show that the lower bound provided by the relaxation can be arbitrarily poor in the worst case. We introduce the Minimum Classes Inequalities (MCIs), which strengthen the relaxation and restore a worst-case performance guarantee of 1/2, matching that of the classical BPP. In addition, we derive the Minimum Bins Inequality (MBI) to further reinforce the relaxation, together with an upper bound on the number of bins in any optimal BPPS solution, which leads to a significant reduction in the number of variables and constraints of the ILP formulation. Finally, we establish a comprehensive benchmark of 480 BPPS instances and conduct extensive computational experiments. The results show that the integration of MCIs, the MBI, and the upper bound on the number of bins substantially improves the performance of the ILP formulation in terms of solution time and number of instances solved to optimality.
Artificial intelligence (AI) is transforming financial planning by expanding access, lowering costs, and enabling dynamic, data-driven advice. Yet without clear safeguards, digital platforms risk reproducing longstanding market inefficiencies such as information asymmetry, misaligned incentives, and systemic fragility. This paper develops a framework for responsible AI in financial planning, anchored in five principles: fiduciary duty, adaptive personalization, technical robustness, ethical and fairness constraints, and auditability. We illustrate these risks and opportunities through case studies, and extend the framework into a five-level roadmap of AI financial intermediaries. By linking technological design to economic theory, we show how AI can either amplify vulnerabilities or create more resilient, trustworthy forms of financial intermediation.
NASA's Habitable Worlds Observatory (HWO) aims to achieve starlight suppression to the $10^{-10}$ level for the detection and spectral characterization of Earth-like exoplanets. Broadband ozone absorption features are key biosignatures that appear in the 200-400nm near-ultraviolet (UV) regime. Extending coronagraphy from visible wavelengths to the UV, however, brings with it a number of challenges, including tighter requirements on wavefront sensing and control, optical surface quality, scattered light, and polarization aberrations, among other things. We aim to partially quantify and address these challenges with a combination of modeling, high-resolution metrology to the scales required for UV coronagraphy, and ultimately a demonstration of UV coronagraphy on the Space Coronagraph Optical Bench (SCoOB) vacuum testbed. In these proceedings, we provide a status update on our modeling and contrast budgeting efforts, characterization efforts to understand performance limitations set by key optical components, and our plans to move toward a demonstration of UV coronagraphy.
Vision-Language-Action (VLA) models have recently emerged as a powerful paradigm for robotic manipulation. Despite substantial progress enabled by large-scale pretraining and supervised fine-tuning (SFT), these models face two fundamental challenges: (i) the scarcity and high cost of large-scale human-operated robotic trajectories required for SFT scaling, and (ii) limited generalization to tasks involving distribution shift. Recent breakthroughs in Large Reasoning Models (LRMs) demonstrate that reinforcement learning (RL) can dramatically enhance step-by-step reasoning capabilities, raising a natural question: Can RL similarly improve the long-horizon step-by-step action planning of VLA? In this work, we introduce SimpleVLA-RL, an efficient RL framework tailored for VLA models. Building upon veRL, we introduce VLA-specific trajectory sampling, scalable parallelization, multi-environment rendering, and optimized loss computation. When applied to OpenVLA-OFT, SimpleVLA-RL achieves SoTA performance on LIBERO and even outperforms $\pi_0$ on RoboTwin 1.0\&2.0 with the exploration-enhancing strategies we introduce. SimpleVLA-RL not only reduces dependence on large-scale data and enables robust generalization, but also remarkably surpasses SFT in real-world tasks. Moreover, we identify a novel phenomenon ``pushcut'' during RL training, wherein the policy discovers previously unseen patterns beyond those seen in the previous training process. Github: https://github.com/PRIME-RL/SimpleVLA-RL
The drift-plus-penalty method is a Lyapunov optimisation technique commonly applied to network routing problems. It reduces the original stochastic planning task to a sequence of greedy optimizations, enabling the design of distributed routing algorithms which stabilize data queues while simultaneously optimizing a specified penalty function. While drift-plus-penalty methods have desirable asymptotic properties, they tend to incur higher network delay than alternative control methods, especially under light network load. In this work, we propose a learned variant of the drift-plus-penalty method that can preserve its theoretical guarantees, while being flexible enough to learn routing strategies directly from a model of the problem. Our approach introduces a novel mechanism for learning routing decisions and employs an optimal transport-based method for link scheduling. Applied to the joint task of transmit-power allocation and data routing, the method achieves consistent improvements over common baselines under a broad set of scenarios.
Visual navigation using only a single camera and a topological map has recently become an appealing alternative to methods that require additional sensors and 3D maps. This is typically achieved through an "image-relative" approach to estimating control from a given pair of current observation and subgoal image. However, image-level representations of the world have limitations because images are strictly tied to the agent's pose and embodiment. In contrast, objects, being a property of the map, offer an embodiment- and trajectory-invariant world representation. In this work, we present a new paradigm of learning "object-relative" control that exhibits several desirable characteristics: a) new routes can be traversed without strictly requiring to imitate prior experience, b) the control prediction problem can be decoupled from solving the image matching problem, and c) high invariance can be achieved in cross-embodiment deployment for variations across both training-testing and mapping-execution settings. We propose a topometric map representation in the form of a "relative" 3D scene graph, which is used to obtain more informative object-level global path planning costs. We train a local controller, dubbed "ObjectReact", conditioned directly on a high-level "WayObject Costmap" representation that eliminates the need for an explicit RGB input. We demonstrate the advantages of learning object-relative control over its image-relative counterpart across sensor height variations and multiple navigation tasks that challenge the underlying spatial understanding capability, e.g., navigating a map trajectory in the reverse direction. We further show that our sim-only policy is able to generalize well to real-world indoor environments. Code and supplementary material are accessible via project page: https://object-react.github.io/
Astronomical observations at millimeter and submillimeter wavelengths heavily depend on the amount of Precipitable Water Vapor (PWV) in the atmosphere, directly affecting the sky transparency and degrading the quality of the signals received by radio telescopes. Predictions of PWV at different forecasting horizons is crucial to support telescope operations, engineering planning, and observational scheduling and efficiency of radio observatories installed in the Chajnantor area in northern Chile. We developed and validated a Long Short-Term Memory (LSTM) deep learning-based model to predict PWV at forecasting horizons of 12, 24, 36, and 48 hours using historical data from two 183 GHz radiometers and a weather station in the Chajnantor area. We find the LSTM method is able to predict PWV in the 12 and 24 hours forecasting horizons with Mean Absolute Percentage Error (MAPE) of 22% compared to 36% of the traditional Global Forecast System (GFS) method used by Atacama Pathfinder EXperiment (APEX) and the Root Mean Square Error (RMSE) in mm are reduced by 50%. We present a first application of deep learning techniques for preliminary predictions of PWV in the Chajnantor area. The prediction performance shows significant improvements to traditional methods in 12 and 24 hours time windows. We also propose upgrades to improve our method in short (< 1 hour) and long (> 36 hours) forecasting timescales for future work.
Nor'easters frequently impact the North American East Coast, bringing hazardous precipitation, winds, and coastal flooding. Accurate simulation of their pressure and wind fields is essential for forecasting, risk assessment, and infrastructure planning, yet remains challenging due to their complex, asymmetric structure. This study introduces a novel hybrid analytical-data-driven model designed to efficiently simulate Nor'easter pressure and boundary layer wind fields. The pressure field is modeled using an adapted Holland-type formulation, with azimuthally varying parameters estimated through Kriging surrogate models informed by sensitivity analysis of reanalysis data. The wind field is then derived analytically from the momentum equations by decomposing the wind flow into gradient and frictional components. Model performance is assessed against ERA-Interim reanalysis data and surface wind observations from a historical event. Results show that the proposed pressure model accurately reproduces elliptical isobars and key asymmetrical patterns, while the wind model captures the fundamental structure and intensity of the boundary layer flow, including the presence of supergradient winds. Owing to its physical basis, computational efficiency, and ability to represent critical storm asymmetries, the model offers a valuable alternative to computationally expensive numerical simulations for hazard assessment and scenario analysis of extreme Nor'easters.
Objective. Proton beams enable localized dose delivery. Accurate range estimation is essential, but planning still relies on X-ray CT, which introduces uncertainty in stopping power and range. Proton CT measures water equivalent thickness directly but suffers resolution loss from multiple Coulomb scattering. We develop a data driven method that reconstructs water equivalent path length (WEPL) maps from energy resolved proton radiographs, bypassing intermediate reconstructions. Approach. We present a machine learning pipeline for WEPL from high dimensional radiographs. Data were generated with the TOPAS Monte Carlo toolkit, modeling a clinical nozzle and a patient CT. Proton energies spanned 70-230 MeV across 72 projection angles. Principal component analysis reduced input dimensionality while preserving signal. A conditional GAN with gradient penalty was trained for WEPL prediction using a composite loss (adversarial, MSE, SSIM, perceptual) to balance sharpness, accuracy, and stability. Main results. The model reached a mean relative WEPL deviation of 2.5 percent, an SSIM of 0.97, and a proton radiography gamma index passing rate of 97.1 percent (2 percent delta WEPL, 3 mm distance-to-agreement) on a simulated head phantom. Results indicate high spatial fidelity and strong structural agreement. Significance. WEPL can be mapped directly from proton radiographs with deep learning while avoiding intermediate steps. The method mitigates limits of analytic techniques and may improve treatment planning. Future work will tune the number of PCA components, include detector response, explore low dose settings, and extend multi angle data toward full proton CT reconstruction; it is compatible with clinical workflows.
Bagging tasks, commonly found in industrial scenarios, are challenging considering deformable bags' complicated and unpredictable nature. This paper presents an automated bagging system from the proposed adaptive Structure-of-Interest (SOI) manipulation strategy for dual robot arms. The system dynamically adjusts its actions based on real-time visual feedback, removing the need for pre-existing knowledge of bag properties. Our framework incorporates Gaussian Mixture Models (GMM) for estimating SOI states, optimization techniques for SOI generation, motion planning via Constrained Bidirectional Rapidly-exploring Random Tree (CBiRRT), and dual-arm coordination using Model Predictive Control (MPC). Extensive experiments validate the capability of our system to perform precise and robust bagging across various objects, showcasing its adaptability. This work offers a new solution for robotic deformable object manipulation (DOM), particularly in automated bagging tasks. Video of this work is available at https://youtu.be/6JWjCOeTGiQ.
Gliomas are the most prevalent type of primary brain tumors, and their accurate segmentation from MRI is critical for diagnosis, treatment planning, and longitudinal monitoring. However, the scarcity of high-quality annotated imaging data in Sub-Saharan Africa (SSA) poses a significant challenge for deploying advanced segmentation models in clinical workflows. This study introduces a robust and computationally efficient deep learning framework tailored for resource-constrained settings. We leveraged a 3D Attention UNet architecture augmented with residual blocks and enhanced through transfer learning from pre-trained weights on the BraTS 2021 dataset. Our model was evaluated on 95 MRI cases from the BraTS-Africa dataset, a benchmark for glioma segmentation in SSA MRI data. Despite the limited data quality and quantity, our approach achieved Dice scores of 0.76 for the Enhancing Tumor (ET), 0.80 for Necrotic and Non-Enhancing Tumor Core (NETC), and 0.85 for Surrounding Non-Functional Hemisphere (SNFH). These results demonstrate the generalizability of the proposed model and its potential to support clinical decision making in low-resource settings. The compact architecture, approximately 90 MB, and sub-minute per-volume inference time on consumer-grade hardware further underscore its practicality for deployment in SSA health systems. This work contributes toward closing the gap in equitable AI for global health by empowering underserved regions with high-performing and accessible medical imaging solutions.
Recent advances in multimodal large language models (MLLMs) have opened new opportunities for embodied intelligence, enabling multimodal understanding, reasoning, and interaction, as well as continuous spatial decision-making. Nevertheless, current MLLM-based embodied systems face two critical limitations. First, Geometric Adaptability Gap: models trained solely on 2D inputs or with hard-coded 3D geometry injection suffer from either insufficient spatial information or restricted 2D generalization, leading to poor adaptability across tasks with diverse spatial demands. Second, Embodiment Constraint Gap: prior work often neglects the physical constraints and capacities of real robots, resulting in task plans that are theoretically valid but practically infeasible. To address these gaps, we introduce OmniEVA -- an embodied versatile planner that enables advanced embodied reasoning and task planning through two pivotal innovations: (1) a Task-Adaptive 3D Grounding mechanism, which introduces a gated router to perform explicit selective regulation of 3D fusion based on contextual requirements, enabling context-aware 3D grounding for diverse embodied tasks. (2) an Embodiment-Aware Reasoning framework that jointly incorporates task goals and embodiment constraints into the reasoning loop, resulting in planning decisions that are both goal-directed and executable. Extensive experimental results demonstrate that OmniEVA not only achieves state-of-the-art general embodied reasoning performance, but also exhibits a strong ability across a wide range of downstream scenarios. Evaluations of a suite of proposed embodied benchmarks, including both primitive and composite tasks, confirm its robust and versatile planning capabilities. Project page: https://omnieva.github.io
Swept volume computation, the determination of regions occupied by moving objects, is essential in graphics, robotics, and manufacturing. Existing approaches either explicitly track surfaces, suffering from robustness issues under complex interactions, or employ implicit representations that trade off geometric fidelity and face optimization difficulties. We propose a novel inversion of motion perspective: rather than tracking object motion, we fix the object and trace spatial points backward in time, reducing complex trajectories to efficiently linearizable point motions. Based on this, we introduce a multi field tetrahedral framework that maintains multiple distance fileds per element, preserving fine geometric details at trajectory intersections where single field methods fail. Our method robustly computes swept volumes for diverse motions, including translations and screw motions, and enables practical applications in path planning and collision detection.
We present and evaluate Vejde; a framework which combines data abstraction, graph neural networks and reinforcement learning to produce inductive policy functions for decision problems with richly structured states, such as object classes and relations. MDP states are represented as data bases of facts about entities, and Vejde converts each state to a bipartite graph, which is mapped to latent states through neural message passing. The factored representation of both states and actions allows Vejde agents to handle problems of varying size and structure. We tested Vejde agents on eight problem domains defined in RDDL, with ten problem instances each, where policies were trained using both supervised and reinforcement learning. To test policy generalization, we separate problem instances in two sets, one for training and the other solely for testing. Test results on unseen instances for the Vejde agents were compared to MLP agents trained on each problem instance, as well as the online planning algorithm Prost. Our results show that Vejde policies in average generalize to the test instances without a significant loss in score. Additionally, the inductive agents received scores on unseen test instances that on average were close to the instance-specific MLP agents.
Accurate motion prediction of surrounding agents is crucial for the safe planning of autonomous vehicles. Recent advancements have extended prediction techniques from individual agents to joint predictions of multiple interacting agents, with various strategies to address complex interactions within future motions of agents. However, these methods overlook the evolving nature of these interactions. To address this limitation, we propose a novel progressive multi-scale decoding strategy, termed ProgD, with the help of dynamic heterogeneous graph-based scenario modeling. In particular, to explicitly and comprehensively capture the evolving social interactions in future scenarios, given their inherent uncertainty, we design a progressive modeling of scenarios with dynamic heterogeneous graphs. With the unfolding of such dynamic heterogeneous graphs, a factorized architecture is designed to process the spatio-temporal dependencies within future scenarios and progressively eliminate uncertainty in future motions of multiple agents. Furthermore, a multi-scale decoding procedure is incorporated to improve on the future scenario modeling and consistent prediction of agents' future motion. The proposed ProgD achieves state-of-the-art performance on the INTERACTION multi-agent prediction benchmark, ranking $1^{st}$, and the Argoverse 2 multi-world forecasting benchmark.
Accurately reasoning about future parking spot availability and integrated planning is critical for enabling safe and efficient autonomous valet parking in dynamic, uncertain environments. Unlike existing methods that rely solely on instantaneous observations or static assumptions, we present an approach that predicts future parking spot occupancy by explicitly distinguishing between initially vacant and occupied spots, and by leveraging the predicted motion of dynamic agents. We introduce a probabilistic spot occupancy estimator that incorporates partial and noisy observations within a limited Field-of-View (FoV) model and accounts for the evolving uncertainty of unobserved regions. Coupled with this, we design a strategy planner that adaptively balances goal-directed parking maneuvers with exploratory navigation based on information gain, and intelligently incorporates wait-and-go behaviors at promising spots. Through randomized simulations emulating large parking lots, we demonstrate that our framework significantly improves parking efficiency, safety margins, and trajectory smoothness compared to existing approaches.
In this work, we propose a novel flow field-based motion planning method that drives a robot from any initial state to a desired reference trajectory such that it converges to the trajectory's end point. Despite demonstrated efficacy in using Koopman operator theory for modeling dynamical systems, Koopman does not inherently enforce convergence to desired trajectories nor to specified goals -- a requirement when learning from demonstrations (LfD). We present KoopMotion which represents motion flow fields as dynamical systems, parameterized by Koopman Operators to mimic desired trajectories, and leverages the divergence properties of the learnt flow fields to obtain smooth motion fields that converge to a desired reference trajectory when a robot is placed away from the desired trajectory, and tracks the trajectory until the end point. To demonstrate the effectiveness of our approach, we show evaluations of KoopMotion on the LASA human handwriting dataset and a 3D manipulator end-effector trajectory dataset, including spectral analysis. We also perform experiments on a physical robot, verifying KoopMotion on a miniature autonomous surface vehicle operating in a non-static fluid flow environment. Our approach is highly sample efficient in both space and time, requiring only 3\% of the LASA dataset to generate dense motion plans. Additionally, KoopMotion provides a significant improvement over baselines when comparing metrics that measure spatial and temporal dynamics modeling efficacy.
Variant calling is the first step in analyzing a human genome and aims to detect variants in an individual's genome compared to a reference genome. Due to the computationally-intensive nature of variant calling, genomic data are increasingly processed in cloud environments as large amounts of compute and storage resources can be acquired with the pay-as-you-go pricing model. In this paper, we address the problem of efficiently executing a variant calling pipeline for a workload of human genomes on graphics processing unit (GPU)-enabled machines. We propose a novel machine learning (ML)-based approach for optimizing the workload execution to minimize the total execution time. Our approach encompasses two key techniques: The first technique employs ML to predict the execution times of different stages in a variant calling pipeline based on the characteristics of a genome sequence. Using the predicted times, the second technique generates optimal execution plans for the machines by drawing inspiration from the flexible job shop scheduling problem. The plans are executed via careful synchronization across different machines. We evaluated our approach on a workload of publicly available genome sequences using a testbed with different types of GPU hardware. We observed that our approach was effective in predicting the execution times of variant calling pipeline stages using ML on features such as sequence size, read quality, percentage of duplicate reads, and average read length. In addition, our approach achieved 2X speedup (on an average) over a greedy approach that also used ML for predicting the execution times on the tested workload of sequences. Finally, our approach achieved 1.6X speedup (on an average) over a dynamic approach that executed the workload based on availability of resources without using any ML-based time predictions.
Cyber warfare has become a central element of modern conflict, especially within multi-domain operations. As both a distinct and critical domain, cyber warfare requires integrating defensive and offensive technologies into coherent strategies. While prior research has emphasized isolated tactics or fragmented technologies, a holistic understanding is essential for effective resource deployment and risk mitigation. Game theory offers a unifying framework for this purpose. It not only models attacker-defender interactions but also provides quantitative tools for equilibrium analysis, risk assessment, and strategic reasoning. Integrated with modern AI techniques, game-theoretic models enable the design and optimization of strategies across multiple levels of cyber warfare, from policy and strategy to operations, tactics, and technical implementations. These models capture the paradoxical logic of conflict, where more resources do not always translate into greater advantage, and where nonlinear dynamics govern outcomes. To illustrate the approach, this chapter examines RedCyber, a synthetic cyber conflict, demonstrating how game-theoretic methods capture the interdependencies of cyber operations. The chapter concludes with directions for future research on resilience, cros-echelon planning, and the evolving role of AI in cyber warfare.