arXiv subjects

Karl Kruusamäe

Publications and source records attributed to Karl Kruusamäe.

7 recordsLinked to original sources

arXiv2026

PRISM: Projection-Integrated Sampling-Based MPC with Bayesian Cost Tuning for Bimanual Manipulation

Bimanual manipulation in cluttered, contact-rich environments remains challenging because it requires coordinated motion generation, interaction-aware planning, and reliable execution under tight kinematic constraints. We present PRISM, a projection-integrated sampling-based Model Predictive Control (MPC) framework that uses a GPU-accelerated physics simulator as an online world model for complex dual-arm manipulation. The main algorithmic contribution is a QP-guided control sampling strategy that decouples trajectory exploration from kinematic feasibility. At each MPC step, sampled joint-velocity trajectories are projected onto the set of motions satisfying joint position, velocity, acceleration, and jerk bounds, together with an initial-velocity boundary condition, before rollout evaluation. This enables broad yet feasible exploration of coordinated bimanual behaviors. To support efficient online execution, we derive a custom ADMM/Bregman-splitting QP solver that exploits joint-wise separability and reusable matrix factorizations. We further use Bayesian optimization to tune task-cost weights offline, reducing manual parameter selection. We evaluate PRISM on challenging variants of PerAct$^{2}$ tasks, including obstacle-constrained ball transport, tray transport, cube handover, and box lifting. Experiments show improved robustness and task success relative to representative sampling-based baselines, while maintaining real-time or near-real-time execution. We also demonstrate successful sim-to-real transfer on dual UR5e manipulators, highlighting the practical potential of physics-based online planning for contact-rich bimanual manipulation. Project details, including code and supplementary videos, are available at \href{https://sites.google.com/view/prismbimanual}{\texttt{https://sites.google.com/view/prismbimanual}}.

cs.RO↗

arXiv2026

Co-Creating Buildable and Open Social Robot Study Companions with University Students

Open-source social robots offer accessibility, repairability, and student empowerment, yet the build itself often presents a barrier. Existing platforms either ship pre-assembled, foreclosing hands-on learning, or expose students to unfamiliar fasteners, opaque wiring, and inaccessible service points that erode engagement. Whether targeted mechanical redesign can lower this barrier whilst maintaining structural integrity remains untested. Here we show that Design for Assembly (DfA) and Design for Disassembly (DfD) interventions reshape how a build feels before they shorten how long it takes. Working with university students in Guyana and Estonia, we applied the Double Diamond framework to co-create the Robot Study Companion (RSC) v4.1: mapping pain points, then redesigning its chassis around twist-lock fasteners, snap-fit joints, and tool-free service latches. Across two studies with developers and first-time builders, system usability climbed from Poor to Excellent (SUS 59.4 to 89.4), perceived workload trended downward (NASA-TLX 4.29 to 4.00), and mean assembly time trended downward (21.4 to 13.7 minutes, with juniors' learning effect), whilst orientation cues and navigation continuity for first-time builders emerged as the next documentation frontier. Perceived workload, not completion time, appears to govern whether students take up open hardware.

cs.RO↗

arXiv2026

Benchmarking Local Language Models for Social Robots using Edge Devices

Social-educational robots designed for socially interactive pedagogical support, such as the Robot Study Companion (RSC), rely on responsive, privacy-preserving interaction despite severely limited compute. However, there is a gap in systematic benchmarking of language models for edge computing in pedagogical applications. This paper benchmarks 25 open-source language models for local deployment on edge hardware. We evaluate each model across three dimensions: inference efficiency (tokens per second, energy consumption), general knowledge (a six-category MMLU subset), and teaching effectiveness (LLM-rated pedagogical quality), validated against five independent human raters using the Raspberry Pi(RPi)4 as the primary platform, with additional comparisons on the RPi5 and a laptop GPU. Results reveal pronounced trade-offs: throughput and energy efficiency vary by over an order of magnitude across models, MMLU accuracy ranges from near-random to 57.2%, and teaching effectiveness does not correlate monotonically with either metric. Among the evaluated models, Granite4 Tiny Hybrid (7B) achieves a strong overall balance, reaching 2.5 tokens per second, 0.90 tokens per joule, and 54.6% MMLU accuracy; high MMLU accuracy does not appear necessary for strong teaching scores. Human validation on four representative models preserved the automated rank ordering (Pearson r = 0.967, n = 4). Based on these findings, we propose a three-tier local inference architecture for the RSC that balances responsiveness and accuracy on resource-constrained hardware.

cs.RO↗

arXiv2025

Sampling-Based Optimization with Parallelized Physics Simulator for Bimanual Manipulation

In recent years, dual-arm manipulation has become an area of strong interest in robotics, with end-to-end learning emerging as the predominant strategy for solving bimanual tasks. A critical limitation of such learning-based approaches, however, is their difficulty in generalizing to novel scenarios, especially within cluttered environments. This paper presents an alternative paradigm: a sampling-based optimization framework that utilizes a GPU-accelerated physics simulator as its world model. We demonstrate that this approach can solve complex bimanual manipulation tasks in the presence of static obstacles. Our contribution is a customized Model Predictive Path Integral Control (MPPI) algorithm, \textbf{guided by carefully designed task-specific cost functions,} that uses GPU-accelerated MuJoCo for efficiently evaluating robot-object interaction. We apply this method to solve significantly more challenging versions of tasks from the PerAct$^{2}$ benchmark, such as requiring the point-to-point transfer of a ball through an obstacle course. Furthermore, we establish that our method achieves real-time performance on commodity GPUs and facilitates successful sim-to-real transfer by leveraging unique features within MuJoCo. The paper concludes with a statistical analysis of the sample complexity and robustness, quantifying the performance of our approach. The project website is available at: https://sites.google.com/view/bimanualakslabunitartu .

cs.RO↗

arXiv2023

Multimodal Grounding for Embodied AI via Augmented Reality Headsets for Natural Language Driven Task Planning

Recent advances in generative modeling have spurred a resurgence in the field of Embodied Artificial Intelligence (EAI). EAI systems typically deploy large language models to physical systems capable of interacting with their environment. In our exploration of EAI for industrial domains, we successfully demonstrate the feasibility of co-located, human-robot teaming. Specifically, we construct an experiment where an Augmented Reality (AR) headset mediates information exchange between an EAI agent and human operator for a variety of inspection tasks. To our knowledge the use of an AR headset for multimodal grounding and the application of EAI to industrial tasks are novel contributions within Embodied AI research. In addition, we highlight potential pitfalls in EAI's construction by providing quantitative and qualitative analysis on prompt robustness.

cs.RO↗

arXiv2021

GPU Accelerated Batch Multi-Convex Trajectory Optimization for a Rectangular Holonomic Mobile Robot

We present a batch trajectory optimizer that can simultaneously solve hundreds of different instances of the problem in real-time. We consider holonomic robots but relax the assumption of circular base footprint. Our main algorithmic contributions lie in: (i) improving the computational tractability of the underlying non-convex problem and (ii) leveraging batch computation to mitigate initialization bottlenecks and improve solution quality. We achieve both goals by deriving a multi-convex reformulation of the kinematics and collision avoidance constraints. We exploit these structures through an Alternating Minimization approach and show that the resulting batch operation reduces to computing just matrix-vector products that can be trivially accelerated over GPUs. We improve the state-of-the-art in three respects. First, we improve quality of navigation (success-rate, tracking) as compared to baseline approach that relies on computing a single locally optimal trajectory at each control loop. Second, we show that when initialized with trajectory samples from a Gaussian distribution, our batch optimizer outperforms state-of-the-art cross-entropy method in solution quality. Finally, our batch optimizer is several orders of magnitude faster than the conceptually simpler alternative of running different optimization instances in parallel CPU threads. \textbf{Codes:} \url{https://tinyurl.com/a3b99m8}

cs.RO↗

arXiv2020

Fast Adaptation of Manipulator Trajectories to Task Perturbation By Differentiating through the Optimal Solution

Joint space trajectory optimization under end-effector task constraints leads to a challenging non-convex problem. Thus, a real-time adaptation of prior computed trajectories to perturbation in task constraints often becomes intractable. Existing works use the so-called warm-starting of trajectory optimization to improve computational performance. We present a fundamentally different approach that relies on deriving analytical gradients of the optimal solution with respect to the task constraint parameters. This gradient map characterizes the direction in which the prior computed joint trajectories need to be deformed to comply with the new task constraints. Subsequently, we develop an iterative line-search algorithm for computing the scale of deformation. Our algorithm provides near real-time adaptation of joint trajectories for a diverse class of task perturbations such as (i) changes in initial and final joint configurations of end-effector orientation-constrained trajectories and (ii) changes in end-effector goal or way-points under end-effector orientation constraints. We relate each of these examples to real-world applications ranging from learning from demonstration to obstacle avoidance. We also show that our algorithm produces trajectories with quality similar to what one would obtain by solving the trajectory optimization from scratch with warm-start initialization. But most importantly, our algorithm achieves a worst-case speed-up of 160x over the latter approach.

cs.RO↗