High-dimensional Configuration Space Research Articles

Model predictive manipulation of compliant objects with multi-objective optimizer and adversarial network for occlusion compensation

Background:The manipulation of compliant objects by robotic systems remains a challenging task, largely due to their variable shapes and the complex, high-dimensional nature of their interaction dynamics. Traditional robotic manipulation strategies struggle with the accurate modeling and control necessary to handle such materials, especially in the presence of visual occlusions that frequently occur in dynamic environments. Meanwhile, for most unstructured environments, robots are required to have autonomous interactions with their surroundings. Methods:To solve the shape manipulation of compliant objects in an unstructured environment, we begin by exploring the regression-based algorithm of representing the high-dimensional configuration space of deformable objects in a compressed form that enables efficient and effective manipulation. Simultaneously, we address the issue of visual occlusions by proposing the integration of an adversarial network, enabling guiding the shaping task even with partial observations of the object. Afterwards, we propose a receding-time estimator to coordinate the robot action with the computed shape features while satisfying various performance criteria. Finally, model predictive controller is utilized to compute the robot’s shaping motions subject to safety constraints. Detailed experiments are presented to evaluate the proposed manipulation framework. Significant findings:Our MPC framework utilizes the compressed representation and occlusion-compensated information to predict the object’s behavior, while the multi-objective optimizer ensures that the resulting control actions meet multiple performance criteria. Through rigorous experimental validation, our approach demonstrates superior manipulation capabilities in scenarios with visual obstructions, outperforming existing methods in terms of precision and operational reliability. The findings highlight the potential of our integrated approach to significantly enhance the manipulation of compliant objects in real-world robotic applications.

Machine learning accelerated discovery of corrosion-resistant high-entropy alloys

Corrosion has a wide impact on society, causing catastrophic damage to structurally engineered components. An emerging class of corrosion-resistant materials are high-entropy alloys. However, high-entropy alloys live in high-dimensional composition and configuration space, making materials designs via experimental trial-and-error or brute-force ab initio calculations almost impossible. Here we develop a physics-informed machine-learning framework to identify corrosion-resistant high-entropy alloys. Three metrics are used to evaluate the corrosion resistance, including single-phase formability, surface energy and the compactness of oxide films formed on an alloy surface evaluated by Pilling–Bedworth ratios. We used random forest models to predict the single-phase formability, trained on an experimental dataset. Machine learning inter-atomic potentials were employed to calculate surface energies and Pilling–Bedworth ratios, which are trained on first-principles data fast sampled using embedded atom models. A combination of random forest models and high-fidelity machine learning potentials represents the first of its kind to relate chemical compositions to corrosion resistance of high-entropy alloys, paving the way for automatic design of materials with superior corrosion protection. This framework was demonstrated on AlCrFeCoNi high-entropy alloys and we identified composition regions with high corrosion resistance from a wide range of compositions. Machine learning predicted lattice constants and surface energies are consistent with values by first-principles calculations. The predicted single-phase formability and corrosion-resistant compositions of AlCrFeCoNi agree well with experiments. This framework provides a computationally efficient approach to navigate high-dimensional composition space of high-entropy alloys. It is general in its application and applicable to other complex materials, enabling high-throughput screening of material candidates and potentially accelerating the iteration of integrated computational materials engineering.

Reduced order modeling and model order reduction for continuum manipulators: an overview.

Soft robot's natural dynamics calls for the development of tailored modeling techniques for control. However, the high-dimensional configuration space of the geometrically exact modeling approaches for soft robots, i.e., Cosserat rod and Finite Element Methods (FEM), has been identified as a key obstacle in controller design. To address this challenge, Reduced Order Modeling (ROM), i.e., the approximation of the full-order models, and Model Order Reduction (MOR), i.e., reducing the state space dimension of a high fidelity FEM-based model, are enjoying extensive research. Although both techniques serve a similar purpose and their terms have been used interchangeably in the literature, they are different in their assumptions and implementation. This review paper provides the first in-depth survey of ROM and MOR techniques in the continuum and soft robotics landscape to aid Soft Robotics researchers in selecting computationally efficient models for their specific tasks.

Cloud Configuration Optimization for Recurring Batch-Processing Applications

Recognizing the diversity of Big Data analytic jobs, cloud providers offer a wide range of VM instance types or even clusters to cater for different use cases. The choice of cloud configurations can have a significant impact on the response time and running cost of batch-processing applications, which may need to be re-run regularly with cloud-scale resources. However, identifying the best cloud configuration with a low search cost is quite challenging due to i) the large and high-dimensional configuration space, ii) the time-varying cloud service cost (e.g., AWS Spot instances), and iii) job response time variation even given the same configuration. To tackle these challenges, we design and implement <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Accordia</i> , a system that enables Adaptive Cloud Configuration Optimization for Recurring Data-Intensive Applications. By leveraging recent algorithmic advances in Gaussian Process UCB techniques, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Accordia</i> can unearth the cost-optimal configuration with a deadline constraint (i.e., maximum tolerated running time) under the time-varying cloud service cost. More importantly, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Accordia</i> manages to achieve a theoretical performance guarantee, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">sub-linearly increasing dynamic regret</i> of the job completion cost. Using extensive trace-driven simulations and empirical measurements of our Kubernetes-based implementation, we demonstrate that <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Accordia</i> can identify a near-cost-optimal configuration (i.e., within 10% of the optimum) after fewer than 20 runs from over 7000 candidate choices, which translates to a 2X-speedup and up to 17.9% cost-savings, when comparing to the state-of-the-art approach, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">CherryPick</i> .

High performance assembly of complex structural parts in special environments – research on space manipulator assisted module docking method

PurposeCompared with the robotic manipulation in structured environment, high performance assembly of complex parts in extreme special environment is facing great challenges because of the uncertainty in the environment, and the decline of the control accuracy of the robot and the sensor accuracy. The assembly and construction of the space station is a typical case. An important step in the construction of the space station is the module positioning and docking with the auxiliary of the space manipulator. The operation of the manipulator is faced with many problems, such as low sensing information accuracy, large end position deviation and the requirement of weak impact in the docking process. The purpose of this paper is to design a docking method at the strategy level to effectively solve the problems that may be faced in the docking process.Design/methodology/approachInspired by the research of robotic high-precision compliant assembly, this paper introduces the concept of Attractive Region in Environment (ARIE) into the space manipulator–assisted module docking. The contact configuration space of the docking mechanism and the existence of ARIE are systematically analyzed. The docking strategy based on ARIE framework is proposed, in which the impedance control is used to ensure the weak impact during the docking process.FindingsFor the androgynous peripheral spacecraft docking mechanism, a large range of attractive region exists in the high-dimensional contact configuration space. The docking strategy based on ARIE framework can be designed according to the geometric characteristics of the constraint region and the structural characteristics of the docking mechanism. The virtual models and the simulation environment are established, and the effectiveness of the proposed method is preliminarily verified.Originality/valueBased on the research results of robotic precision compliant manipulation, in this paper, the theory of ARIE is first systematically applied to the analysis of spacecraft docking problem and the design of docking scheme. The effectiveness of the proposed docking method is preliminarily verified for the requirements of large position tolerance and weak impact. The research results will provide theoretical support and technical reference for the assembly and construction of space station and other space manipulator operations.

Simulating short-range order in compositionally complex materials.

In multicomponent materials, short-range order (SRO) is the development of correlated arrangements of atoms at the nanometer scale. Its impact in compositionally complex materials has stimulated an intense debate within the materials science community. Understanding SRO is critical to control the properties of technologically relevant materials, from metallic alloys to functional ceramics. In contrast to long-range order, quantitative characterization of the nature and spatial extent of SRO evades most of the experimentally available techniques. Simulations at the atomistic scale have full access to SRO but face the challenge of accurately sampling high-dimensional configuration spaces to identify the thermodynamic and kinetic conditions at which SRO is formed and what impact it has on material properties. Here we highlight recent progress in computational approaches, such as machine learning-based interatomic potentials, for quantifying and understanding SRO in compositionally complex materials. We briefly recap the key theoretical concepts and methods.

Algorithms and Systems for Manipulating Multiple Objects

Robot manipulation of multiple objects is an important topic for applications including warehouse automation, service robots performing cleaning, and large-scale object sorting. Although problems can range in complexity from a few objects to large disordered piles, autonomy remains a significant technical challenge due to the high-dimensional joint configuration space of the robot and all objects, the complex dynamics of object interaction, and the ambiguity and occlusion caused by clutter. This article surveys a broad range of classical and state-of-the-art research in multiobject manipulation and categorizes them along the dimensions of tasks, perception, predictive models, and decision-making algorithms. It also covers emerging trends and open problems faced in the ongoing effort to realize robust multiobject manipulation systems in practice.

Reweighted Manifold Learning of Collective Variables from Enhanced Sampling Simulations.

Enhanced sampling methods are indispensable in computational chemistry and physics, where atomistic simulations cannot exhaustively sample the high-dimensional configuration space of dynamical systems due to the sampling problem. A class of such enhanced sampling methods works by identifying a few slow degrees of freedom, termed collective variables (CVs), and enhancing the sampling along these CVs. Selecting CVs to analyze and drive the sampling is not trivial and often relies on chemical intuition. Despite routinely circumventing this issue using manifold learning to estimate CVs directly from standard simulations, such methods cannot provide mappings to a low-dimensional manifold from enhanced sampling simulations, as the geometry and density of the learned manifold are biased. Here, we address this crucial issue and provide a general reweighting framework based on anisotropic diffusion maps for manifold learning that takes into account that the learning data set is sampled from a biased probability distribution. We consider manifold learning methods based on constructing a Markov chain describing transition probabilities between high-dimensional samples. We show that our framework reverts the biasing effect, yielding CVs that correctly describe the equilibrium density. This advancement enables the construction of low-dimensional CVs using manifold learning directly from the data generated by enhanced sampling simulations. We call our framework reweighted manifold learning. We show that it can be used in many manifold learning techniques on data from both standard and enhanced sampling simulations.

On Free Velocity Cones of Narrow Passages of High-DoF Kinematic Chains

Narrow passages in free space pose great challenges to many sampling-based motion planners. In problems with high-dimensional configuration spaces (C-spaces), narrow passages in free space (C-free) can be hard to find and navigate and, sometimes, even counterintuitive. In this article, we present an algorithm to construct cones of available instantaneous velocities (“free velocity cones” or, briefly, “free cones”) on the boundary of C-free to facilitate finding and navigating narrow passages. This is accomplished by first developing free cones and associated measures of local C-space narrowness for a single rigid link. These results are then extended to kinematic chains (open and closed) of arbitrary degrees of freedom. It turns out that the degeneracy of the free cones dictates the existence of C-space narrow passages, and the locations and orientations of the links in the workspace relate intimately to the corresponding connected component of C-free. This observation leads us to a modified probabilistic roadmap (M-PRM) algorithm and a modified rapidly exploring random vine (M-RRV) algorithm of combining the enumeration of topological components and random sampling of the free cones. Experimental results from applying our algorithms to several challenging examples show that our new algorithms are more efficient than some variants of the PRM algorithms, and several variants of the rapidly exploring random tree (RRT) algorithms, such as RRT-CONNECT and RRV.

Microstructure design using a human computation game

The design of microstructures that are optimized for a given engineering applications requires exploration of a rough and high-dimensional configuration space. Gradient-based algorithms are efficient, but suffer from a propensity to get stuck in local minima. Global-optimization algorithms are better at finding global minima, but are generally slow to converge. We developed and tested a Human Computation Game (HCG) for microstructure design where players interactively manipulate the microstructure to optimize an effective macroscopic material property. We investigate the impact of various game mechanics on solution quality and efficiency, and compare the HCG player solutions to those of a traditional global optimization algorithm—Simulated Annealing (SA). We show that organizing players into Synchronous teams performed better on more complex problems on average than players working Asynchronously or Solo. We also show that in the best cases, players can find microstructures that outperform those obtained by SA by up to 25% using the same number of computations, or achieve the same performance using up to 307 times fewer computational steps. By studying the optimization strategies employed by HCG players, we anticipate that improved optimization algorithms for microstructure design (and other configurational optimization problems) can be developed.

Robot Precision Assembly Combining With Passive and Active Compliant Motions

Precision assembly is one of the primary goals of robots in manufacturing. Assembly strategies combining active and passive compliant control are presented herein. In this article, we construct a high-dimensional configuration space of robot assembly and then divide the configuration space to its subspaces. We further map the active compliant motion and the passive compliant motion of the manipulator into different subspaces. In one subspace, we construct the constraint function and design the passive compliant motion of the manipulator in the constraint region, where the uncertainties of the system should be eliminated by the environment constraints. In another subspace, we design a force controller based on the low-resolution force sensory information to control the position of the robot. The proposed method avoids the design of precision mechanism systems and the usage of high-quality sensors. Several experiments pertaining to peg-in-hole insertions are conducted to demonstrate the efficiency of the proposed method.

ClassyTune: A Performance Auto-Tuner for Systems in the Cloud

Performance tuning can improve the system performance and thus enable the reduction of cloud computing resources needed to support an application. Due to the ever increasing number of parameters and complexity of systems, there is a necessity to automate performance tuning for the complicated systems in the cloud. The state-of-the-art tuning methods are adopting either the experience-driven tuning approach or the data-driven one. Data-driven tuning is attracting increasing attentions, as it has wider applicability. But existing data-driven methods cannot fully address the challenges of sample scarcity and high dimensionality simultaneously. We present ClassyTune, a data-driven automatic configuration tuning tool for cloud systems. ClassyTune exploits the machine learning model of classification for auto-tuning. This exploitation enables the induction of more training samples without increasing the input dimension. Experiments on seven popular systems in the cloud show that ClassyTune can effectively tune system performance to seven times higher for high-dimensional configuration space, outperforming expert tuning and the state-of-the-art auto-tuning solutions. We also describe a use case in which performance tuning enables the reduction of 33% computing resources needed to run an online stateless service.

Exact Topology of the Dynamic Probability Surface of an Activated Process by Persistent Homology.

To gain insight into the reaction mechanism of activated processes, we introduce an exact approach for quantifying the topology of high-dimensional probability surfaces of the underlying dynamic processes. Instead of Morse indexes, we study the homology groups of a sequence of superlevel sets of the probability surface over high-dimensional configuration spaces using persistent homology. For alanine-dipeptide isomerization, a prototype of activated processes, we identify locations of probability peaks and connecting ridges, along with measures of their global prominence. Instead of a saddle point, the transition state ensemble (TSE) of conformations is at the most prominent probability peak after reactants/products, when proper reaction coordinates are included. Intuition-based models, even those exhibiting a double-well, fail to capture the dynamics of the activated process. Peak occurrence, prominence, and locations can be distorted upon subspace projection. While principal component analysis accounts for conformational variance, it inflates the complexity of the surface topology and destroys the dynamic properties of the topological features. In contrast, TSE emerges naturally as the most prominent peak beyond the reactant/product basins, when projected to a subspace of minimum dimension containing the reaction coordinates. Our approach is general and can be applied to investigate the topology of high-dimensional probability surfaces of other activated processes.

Dynamic Manipulation of Deformable Objects With Implicit Integration

Due to their complex dynamics and high-dimensional configuration spaces, non-rigid objects such as cables, garments, bedding and various food items remain notoriously challenging for robots to manipulate effectively. In this letter, we therefore develop, validate and analyze model-based optimal control techniques for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dynamic</i> manipulation of deformable objects. We study, in particular, the application of both the batch Newton method and the stagewise Differential Dynamic Programming (DDP) approach to this challenging problem domain. On a technical level, we derive analytic formulations for all necessary derivatives, noting that numerically stable simulation of deformable objects demands implicit integration schemes, which do not have closed form solutions. While both DDP and Newton's method converge quadratically, our experiments and analysis show that the relative overall performance of these two approaches depends heavily on the dimensions of the control problems being solved. We demonstrate the efficacy of our trajectory optimization formulations through a variety of simulation and real-world experiments.

Permut - Spatially Resolved Hydration Entropies from Atomistic Simulations

The thermodynamics of the first few surface water layers profoundly affects the overall free energy change in protein folding and unfolding, bilayer self-assembly, and ligand binding, and many other biomolecular processes. Whereas the enthalpy contribution can, in principle, be directly calculated from atomistic molecular dynamics simulations, the solvation shell entropy calculation suffers particularly from the poor sampling of the high dimensional configuration space. To furthermore understand, e.g., the effects of water molecules inside a binding pocket or the contribution of individual side chains to the hydration entropy, spatial resolution is required. To address these problems, we developed Per|Mut, a new method to calculate spatially resolved hydration entropies. In Per|Mut, we first employ a permutation reduction, which exploits the permutation symmetry of the identical water molecules to alleviate the sampling problem by the Gibbs factor N! without changing the physics. Next, we use a third-order mutual information expansion to obtain the hydration shell entropy. The spatial resolution and the decomposition into one- two- and three-body correlations of the translational and rotational entropies, as well as of the translation-rotation correlation, allows for an easy physical interpretation and characterization of the resolved entropy changes. In test-simulations, Per|Mut yielded accurate entropies for an argon gas test system and solvated alkanes. Applied to the solvation statistical mechanics of hydrated octanol and the protein Crambin, Per|Mut revealed the local effects of individual chemical groups or side chains on the solvation entropy.

Capturing the dynamic target by the robot manipulator in high-dimensional configuration space map

In robotics, the tough problem about the dynamic target capturing consists of tracking the target by the robot manipulator and grasping the target by the robot finger. For the sake of space, this article deals with only the first problem, tracking the dynamic target by the robot manipulator. The traditional approaches of capturing the dynamic target may work well when they are employed in low-dimensional space by reinforcement learning or physical modeling. However, they fail to work well in high-dimensional space. The traditional approaches have four limitations with respect to Cartesian space, configuration space, reinforcement learning, and physical modeling. To overcome these limitations, this article implements improved dynamic A* algorithm in high-dimensional configuration space map to capture the target. First, a space injection model injects the collision detection and target position from the Cartesian space into the configuration space to construct a high-dimensional map. Then, the target capturing method including the improved dynamic A* algorithm is applied on the map to track and capture the target. Finally, the experiment performed in time-varying environment and the dynamic target achieves a reliable result. This article has proposed an approach that makes the robot manipulator motion planning more accurate in high-dimensional dynamic configuration space. This approach enables the multi-joint manipulator to avoid the obstacle while tracking the target in high-dimensional configuration space. It takes the advantages of heuristic algorithms in the process of target capturing method designing. It adds precision and speed to target tracking. The success of the approach may apply to any industrial robot tracking target, surgical operation, and space probes. And, it may lay a solid foundation for dynamics control with a scope for future investigations.

Searching Multiple Approximate Solutions in Configuration Space to Guide Sampling-Based Motion Planning

High-dimensional configuration space is usually searched using sampling-based motion planning methods. The well-known issue of sampling-based planners is the narrow passage problem caused by small regions of the configuration space that are difficult to cover by random samples. Practically, the presence of narrow passages decreases the probability of finding a solution, and to cope with it, the number of random samples has to be significantly increased, which also increases the planning time. By dilating the free space, e.g., by scaling-down or thinning the robot (or obstacles), narrow passages become wider, which allows us to compute an approximate solution. Then, the configuration space can be sampled densely around the approximate solution to find the solution of the original problem. However, this process may fail if the final solution is too far from the approximate one. In this paper, we propose a method to find multiple approximate solutions in the configuration space to increase the chance of finding the final solution. The approximate solutions are computed by repeated search of the configuration space while avoiding, if possible, the already discovered solutions. This enables us to search for distinct solutions leading through different parts of the configuration space. The number of approximate solutions is automatically determined based on their similarity. All approximate solutions are then used to guide the sampling of the configuration space. The performance of the proposed approach is verified in scenarios with multiple narrow passages and the benefits of the method are demonstrated by comparing the results with the state-of-the-art planners.

Robot path planning optimization method based on heuristic multi-directional rapidly-exploring tree

Aiming at the problems of low exploration efficiency and high optimal solution cost for existing robot path planning methods, a robot path planning optimization method based on heuristic multi-directional rapidly-exploring tree is implemented. In high-dimensional configuration space, drawing on the Rapid-exploration Random Tree (RRT) related algorithm idea, directional sampling control module works under the guidance of the robot goal course as a heuristic exploration. A flexible multi-directional rapidly-exploring tree construction method is used due to a degree of directional instability. In accordance with the principle of the centripetal growth of the tree, new multi-directional trees will be built on demand to arrive at specific coverage of the space. Then based on the previous path exploration vertices, through the merging method of the trees, a closed loop path is formed and optimized to finally generate a relative optimal path. Simulation experiment results show that this method could effectively improve the exploring efficiency with low computational cost.

A Novel Path Planner for Steerable Bevel-Tip Needles to Reach Multiple Targets With Obstacles

Steerable bevel-tip needles offer higher maneuverability independent of insertion depth and consequently are preferred for many needle-steering applications compared with symmetric-tip needles. Using these needles, the clinician can reach previously inaccessible targets using traditional stiff needles, thus helping improve the efficiency of needle insertion procedures significantly. However, due to their nonholonomic kinematics inside biological tissue, path planning of these needles is complicated and requires a great deal of care. Rapidly exploring random-tree (RRT)-based approaches are proper candidates for intraoperative planning of needle motion due to their fast computation and simple implementations. They also work well in high-dimensional configuration spaces and under nonholonomic kinematic constraints, both of which are the characteristics of steerable bevel-tip needle motion inside soft tissue. We developed a new heuristic-based RRT planner to reach multiple targets inside soft tissue without having to completely retract, reorient, and reinsert the needle toward each separate target, resulting in significantly less tissue damage compared with the conventional sequential insertion of the needle toward each target. Moreover, the proposed planner can have real clinical applications, where the limited size of the workspace as well as the needle's limited natural curvature imposes significant limitations on the needle path-planning problem inside soft tissue. Simulations demonstrate the efficiency of the proposed planner. The maximum targeting error of all targets is 1 mm and the needle's inserted length is decreased up to 35% compared with the sequential insertion of the needle for each target.

A Two-Stage Reinforcement Learning Approach for Multi-UAV Collision Avoidance Under Imperfect Sensing

Unlike autonomous ground vehicles (AGVs), unmanned aerial vehicles (UAVs) have a higher dimensional configuration space, which makes the motion planning of multi-UAVs a challenging task. In addition, uncertainties and noises are more significant in UAV scenarios, which increases the difficulty of autonomous navigation for multi-UAV. In this letter, we proposed a two-stage reinforcement learning (RL) based multi-UAV collision avoidance approach without explicitly modeling the uncertainty and noise in the environment. Our goal is to train a policy to plan a collision-free trajectory by leveraging local noisy observations. However, the reinforcement learned collision avoidance policies usually suffer from high variance and low reproducibility, because unlike supervised learning, RL does not have a fixed training set with ground-truth labels. To address these issues, we introduced a two-stage training method for RL based collision avoidance. For the first stage, we optimize the policy using a supervised training method with a loss function that encourages the agent to follow the well-known reciprocal collision avoidance strategy. For the second stage, we use policy gradient to refine the policy. We validate our policy in a variety of simulated scenarios, and the extensive numerical simulations demonstrate that our policy can generate time-efficient and collision-free paths under imperfect sensing, and can well handle noisy local observations with unknown noise levels.