arXiv subjects

Marin Kobilarov

Publications and source records attributed to Marin Kobilarov.

At least 19 recordsLinked to original sources

arXiv2026

RL-Guided PAC-NMPC for Probabilistically-Safe Perception-Based Navigation in Unknown Environments

In this paper, we present an approach for combining stochastic nonlinear model predictive control (SNMPC) and reinforcement learning (RL) to enable probabilistically-safe perception-based navigation in unknown environments. Our method first uses RL to train probabilistic actor-critic and sensor prediction models. We then leverage these probabilistic models in a sampling-based SNMPC framework known as Probably Approximately Correct (PAC)-NMPC, which uses hard constraints to enforce finite-time statistical guarantees on the probability of collision and value function improvement. By ensuring that our finite-horizon SNMPC policies decrease the value function in expectation, we can approach the long-horizon performance of the RL approach while satisfying probabilistic safety constraints. Through simulation experiments, we show that our approach can improve the safety of perception-based RL navigation policies and scale to high dimensional systems with large sensor input spaces and complex nonlinear dynamics. We also demonstrate our approach through hardware experiments, showing improved performance for vision-based navigation with an agile fixed-wing aerial vehicle in unknown environments.

cs.RO↗

arXiv2025

Uncertainty-Aware Planning for Heterogeneous Robot Teams using Dynamic Topological Graphs and Mixed-Integer Programming

Multi-robot planning and coordination in uncertain environments is a fundamental computational challenge, since the belief space increases exponentially with the number of robots. In this paper, we address the problem of planning in uncertain environments with a heterogeneous robot team of fast scout vehicles for information gathering and more risk-averse carrier robots from which the scouts vehicles are deployed. To overcome the computational challenges, we represent the environment and operational scenario using a topological graph, where the parameters of the edge weight distributions vary with the state of the robot team on the graph, and we formulate a computationally efficient mixed-integer program which removes the dependence on the number of robots from its decision space. Our formulation results in the capability to generate optimal multi-robot, long-horizon plans in seconds that could otherwise be computationally intractable. Ultimately our approach enables real-time re-planning, since the computation time is significantly faster than the time to execute one step. We evaluate our algorithm in a scenario where the robot team must traverse an environment while minimizing detection by observers in positions that are uncertain to the robot team. We demonstrate that our method is computationally tractable, can improve performance in the presence of imperfect information, and can be adjusted for different risk profiles.

cs.RO↗

arXiv2025

Robust Perception-Based Navigation using PAC-NMPC with a Learned Value Function

Nonlinear model predictive control (NMPC) is typically restricted to short, finite horizons to limit the computational burden of online optimization. As a result, global planning frameworks are frequently necessary to avoid local minima when using NMPC for navigation in complex environments. By contrast, reinforcement learning (RL) can generate policies that minimize the expected cost over an infinite-horizon and can often avoid local minima, even when operating only on current sensor measurements. However, these learned policies are usually unable to provide performance guarantees (e.g., on collision avoidance), especially when outside of the training distribution. In this paper, we augment Probably Approximately Correct NMPC (PAC-NMPC), a sampling-based stochastic NMPC algorithm capable of providing statistical guarantees of performance and safety, with an approximate perception-based value function trained via RL. We demonstrate in simulation that our algorithm can improve the long-term behavior of PAC-NMPC while outperforming other approaches with regards to safety for both planar car dynamics and more complex, high-dimensional fixed-wing aerial vehicle dynamics. We also demonstrate that, even when our value function is trained in simulation, our algorithm can successfully achieve statistically safe navigation on hardware using a 1/10th scale rally car in cluttered real-world environments using only current sensor information.

cs.RO↗

arXiv2025

Survey of Simulators for Aerial Robots: An Overview and In-Depth Systematic Comparisons

Uncrewed Aerial Vehicle (UAV) research faces challenges with safety, scalability, costs, and ecological impact when conducting hardware testing. High-fidelity simulators offer a vital solution by replicating real-world conditions to enable the development and evaluation of novel perception and control algorithms. However, the large number of available simulators poses a significant challenge for researchers to determine which simulator best suits their specific use-case, based on each simulator's limitations and customization readiness. In this paper we present an overview of 44 UAV simulators, including in-depth, systematic comparisons for 14 of the simulators. Additionally, we present a set of decision factors for selection of simulators, aiming to enhance the efficiency and safety of research endeavors.

cs.RO↗

arXiv2025

Multi-Agent Feedback Motion Planning using Probably Approximately Correct Nonlinear Model Predictive Control

For many tasks, multi-robot teams often provide greater efficiency, robustness, and resiliency. However, multi-robot collaboration in real-world scenarios poses a number of major challenges, especially when dynamic robots must balance competing objectives like formation control and obstacle avoidance in the presence of stochastic dynamics and sensor uncertainty. In this paper, we propose a distributed, multi-agent receding-horizon feedback motion planning approach using Probably Approximately Correct Nonlinear Model Predictive Control (PAC-NMPC) that is able to reason about both model and measurement uncertainty to achieve robust multi-agent formation control while navigating cluttered obstacle fields and avoiding inter-robot collisions. Our approach relies not only on the underlying PAC-NMPC algorithm but also on a terminal cost-function derived from gyroscopic obstacle avoidance. Through numerical simulation, we show that our distributed approach performs on par with a centralized formulation, that it offers improved performance in the case of significant measurement noise, and that it can scale to more complex dynamical systems.

cs.RO↗

arXiv2024

A Feasible Workflow for Retinal Vein Cannulation in Ex Vivo Porcine Eyes with Robotic Assistance

A potential Retinal Vein Occlusion (RVO) treatment involves Retinal Vein Cannulation (RVC), which requires the surgeon to insert a microneedle into the affected retinal vein and administer a clot-dissolving drug. This procedure presents significant challenges due to human physiological limitations, such as hand tremors, prolonged tool-holding periods, and constraints in depth perception using a microscope. This study proposes a robot-assisted workflow for RVC to overcome these limitations. The test robot is operated through a keyboard. An intraoperative Optical Coherence Tomography (iOCT) system is used to verify successful venous puncture before infusion. The workflow is validated using 12 ex vivo porcine eyes. These early results demonstrate a successful rate of 10 out of 12 cannulations (83.33%), affirming the feasibility of the proposed workflow.

cs.RO↗

arXiv2024

Surgical Robot Transformer (SRT): Imitation Learning for Surgical Tasks

We explore whether surgical manipulation tasks can be learned on the da Vinci robot via imitation learning. However, the da Vinci system presents unique challenges which hinder straight-forward implementation of imitation learning. Notably, its forward kinematics is inconsistent due to imprecise joint measurements, and naively training a policy using such approximate kinematics data often leads to task failure. To overcome this limitation, we introduce a relative action formulation which enables successful policy training and deployment using its approximate kinematics data. A promising outcome of this approach is that the large repository of clinical data, which contains approximate kinematics, may be directly utilized for robot learning without further corrections. We demonstrate our findings through successful execution of three fundamental surgical tasks, including tissue manipulation, needle handling, and knot-tying.

cs.RO↗

arXiv2024

Non-Prehensile Aerial Manipulation using Model-Based Deep Reinforcement Learning

With the continual adoption of Uncrewed Aerial Vehicles (UAVs) across a wide-variety of application spaces, robust aerial manipulation remains a key research challenge. Aerial manipulation tasks require interacting with objects in the environment, often without knowing their dynamical properties like mass and friction a priori. Additionally, interacting with these objects can have a significant impact on the control and stability of the vehicle. We investigated an approach for robust control and non-prehensile aerial manipulation in unknown environments. In particular, we use model-based Deep Reinforcement Learning (DRL) to learn a world model of the environment while simultaneously learning a policy for interaction with the environment. We evaluated our approach on a series of push tasks by moving an object between goal locations and demonstrated repeatable behaviors across a range of friction values.

cs.RO↗

arXiv2023

Probably Approximately Correct Nonlinear Model Predictive Control (PAC-NMPC)

Approaches for stochastic nonlinear model predictive control (SNMPC) typically make restrictive assumptions about the system dynamics and rely on approximations to characterize the evolution of the underlying uncertainty distributions. For this reason, they are often unable to capture more complex distributions (e.g., non-Gaussian or multi-modal) and cannot provide accurate guarantees of performance. In this paper, we present a sampling-based SNMPC approach that leverages recently derived sample complexity bounds to certify the performance of a feedback policy without making assumptions about the system dynamics or underlying uncertainty distributions. By parallelizing our approach, we are able to demonstrate real-time receding-horizon SNMPC with statistical safety guarantees in simulation and on hardware using a 1/10th scale rally car and a 24-inch wingspan fixed-wing unmanned aerial vehicle (UAV).

cs.RO↗

arXiv2023

A Small Form Factor Aerial Research Vehicle for Pick-and-Place Tasks with Onboard Real-Time Object Detection and Visual Odometry

This paper introduces a novel, small form-factor, aerial vehicle research platform for agile object detection, classification, tracking, and interaction tasks. General-purpose hardware components were designed to augment a given aerial vehicle and enable it to perform safe and reliable grasping. These components include a custom collision tolerant cage and low-cost Gripper Extension Package, which we call GREP, for object grasping. Small vehicles enable applications in highly constrained environments, but are often limited by computational resources. This work evaluates the challenges of pick-and-place tasks, with entirely onboard computation of object pose and visual odometry based state estimation on a small platform, and demonstrates experiments with enough accuracy to reliably grasp objects. In a total of 70 trials across challenging cases such as cluttered environments, obstructed targets, and multiple instances of the same target, we demonstrated successfully grasping the target in 93% of trials. Both the hardware component designs and software framework are released as open-source, since our intention is to enable easy reproduction and application on a wide range of small vehicles.

cs.RO↗

arXiv2023

Towards Deep Learning Guided Autonomous Eye Surgery Using Microscope and iOCT Images

Recent advancements in retinal surgery have paved the way for a modern operating room equipped with a surgical robot, a microscope, and intraoperative optical coherence tomography (iOCT)- a depth sensor widely used in retinal surgery. Integrating these tools raises the fundamental question of how to effectively combine them to enable surgical autonomy. In this work, we tackle this question by developing a unified framework that facilitates real-time autonomous surgical workflows leveraging these devices. The system features: (1) a novel imaging system that integrates the microscope and iOCT in real-time by dynamically tracking the surgical instrument via a small iOCT scanning region, providing real-time depth feedback; (2) implementation of convolutional neural networks (CNN) that automatically detect and segment task-relevant information for surgical autonomy; (3) intuitive selection of goal waypoints within both the microscope and iOCT views through simple mouse-click interactions; and (4) integration of model predictive control (MPC) for trajectory generation, ensuring patient safety by implementing safety-related kinematic constraints. The system's utility is demonstrated by automating subretinal injection (SI), a challenging procedure with high accuracy and depth perception requirements. We validate our system by conducting 30 successful SI trials on pig eyes, achieving mean needle insertion accuracy of 26 micrometers to various subretinal goals and mean duration of 55 seconds. Preliminary comparisons to a human operator performing SI in robot-assisted mode highlight the enhanced safety of our system. Project website is here: https://sites.google.com/view/eyesurgerymicroscopeoct/home

cs.RO↗

arXiv2023

Micromanipulation in Surgery: Autonomous Needle Insertion Inside the Eye for Targeted Drug Delivery

We consider a micromanipulation problem in eye surgery, specifically retinal vein cannulation (RVC). RVC involves inserting a microneedle into a retinal vein for the purpose of targeted drug delivery. The procedure requires accurately guiding a needle to a target vein and inserting it while avoiding damage to the surrounding tissues. RVC can be considered similar to the reach or push task studied in robotics manipulation, but with additional constraints related to precision and safety while interacting with soft tissues. Prior works have mainly focused developing robotic hardware and sensors to enhance the surgeons' accuracy, leaving the automation of RVC largely unexplored. In this paper, we present the first autonomous strategy for RVC while relying on a minimal setup: a robotic arm, a needle, and monocular images. Our system exclusively relies on monocular vision to achieve precise navigation, gentle placement on the target vein, and safe insertion without causing tissue damage. Throughout the procedure, we employ machine learning for perception and to identify key surgical events such as needle-vein contact and vein punctures. Detecting these events guides our task and motion planning framework, which generates safe trajectories using model predictive control to complete the procedure. We validate our system through 24 successful autonomous trials on 4 cadaveric pig eyes. We show that our system can navigate to target veins within 22 micrometers of XY accuracy and under 35 seconds, and consistently puncture the target vein without causing tissue damage. Preliminary comparison to a human demonstrates the superior accuracy and reliability of our system.

cs.RO↗

arXiv2023

Deep Learning Guided Autonomous Surgery: Guiding Small Needles into Sub-Millimeter Scale Blood Vessels

We propose a general strategy for autonomous guidance and insertion of a needle into a retinal blood vessel. The main challenges underpinning this task are the accurate placement of the needle-tip on the target vein and a careful needle insertion maneuver to avoid double-puncturing the vein, while dealing with challenging kinematic constraints and depth-estimation uncertainty. Following how surgeons perform this task purely based on visual feedback, we develop a system which relies solely on \emph{monocular} visual cues by combining data-driven kinematic and contact estimation, visual-servoing, and model-based optimal control. By relying on both known kinematic models, as well as deep-learning based perception modules, the system can localize the surgical needle tip and detect needle-tissue interactions and venipuncture events. The outputs from these perception modules are then combined with a motion planning framework that uses visual-servoing and optimal control to cannulate the target vein, while respecting kinematic constraints that consider the safety of the procedure. We demonstrate that we can reliably and consistently perform needle insertion in the domain of retinal surgery, specifically in performing retinal vein cannulation. Using cadaveric pig eyes, we demonstrate that our system can navigate to target veins within 22$μm$ XY accuracy and perform the entire procedure in less than 35 seconds on average, and all 24 trials performed on 4 pig eyes were successful. Preliminary comparison study against a human operator show that our system is consistently more accurate and safer, especially during safety-critical needle-tissue interactions. To the best of the authors' knowledge, this work accomplishes a first demonstration of autonomous retinal vein cannulation at a clinically-relevant setting using animal tissues.

cs.RO↗

arXiv2023

Learn Proportional Derivative Controllable Latent Space from Pixels

Recent advances in latent space dynamics model from pixels show promising progress in vision-based model predictive control (MPC). However, executing MPC in real time can be challenging due to its intensive computational cost in each timestep. We propose to introduce additional learning objectives to enforce that the learned latent space is proportional derivative controllable. In execution time, the simple PD-controller can be applied directly to the latent space encoded from pixels, to produce simple and effective control to systems with visual observations. We show that our method outperforms baseline methods to produce robust goal reaching and trajectory tracking in various environments.

cs.LG↗

arXiv2023

Autonomous Needle Navigation in Retinal Microsurgery: Evaluation in ex vivo Porcine Eyes

Important challenges in retinal microsurgery include prolonged operating time, inadequate force feedback, and poor depth perception due to a constrained top-down view of the surgery. The introduction of robot-assisted technology could potentially deal with such challenges and improve the surgeon's performance. Motivated by such challenges, this work develops a strategy for autonomous needle navigation in retinal microsurgery aiming to achieve precise manipulation, reduced end-to-end surgery time, and enhanced safety. This is accomplished through real-time geometry estimation and chance-constrained Model Predictive Control (MPC) resulting in high positional accuracy while keeping scleral forces within a safe level. The robotic system is validated using both open-sky and intact (with lens and partial vitreous removal) ex vivo porcine eyes. The experimental results demonstrate that the generation of safe control trajectories is robust to small motions associated with head drift. The mean navigation time and scleral force for MPC navigation experiments are 7.208 s and 11.97 mN, which can be considered efficient and well within acceptable safe limits. The resulting mean errors along lateral directions of the retina are below 0.06 mm, which is below the typical hand tremor amplitude in retinal microsurgery.

cs.RO↗

arXiv2022

Closed-Form Minkowski Sum Approximations for Efficient Optimization-Based Collision Avoidance

Motion planning methods for autonomous systems based on nonlinear programming offer great flexibility in incorporating various dynamics, objectives, and constraints. One limitation of such tools is the difficulty of efficiently representing obstacle avoidance conditions for non-trivial shapes. For example, it is possible to define collision avoidance constraints suitable for nonlinear programming solvers in the canonical setting of a circular robot navigating around M convex polytopes over N time steps. However, it requires introducing (2+L)MN additional constraints and LMN additional variables, with L being the number of halfplanes per polytope, leading to larger nonlinear programs with slower and less reliable solving time. In this paper, we overcome this issue by building closed-form representations of the collision avoidance conditions by outer-approximating the Minkowski sum conditions for collision. Our solution requires only MN constraints (and no additional variables), leading to a smaller nonlinear program. On motion planning problems for an autonomous car and quadcopter in cluttered environments, we achieve speedups of 4.8x and 8.7x respectively with significantly less variance in solve times and negligible impact on performance arising from the use of outer approximations.

eess.SY↗

arXiv2020

Towards Autonomous Eye Surgery by Combining Deep Imitation Learning with Optimal Control

During retinal microsurgery, precise manipulation of the delicate retinal tissue is required for positive surgical outcome. However, accurate manipulation and navigation of surgical tools remain difficult due to a constrained workspace and the top-down view during the surgery, which limits the surgeon's ability to estimate depth. To alleviate such difficulty, we propose to automate the tool-navigation task by learning to predict relative goal position on the retinal surface from the current tool-tip position. Given an estimated target on the retina, we generate an optimal trajectory leading to the predicted goal while imposing safety-related physical constraints aimed to minimize tissue damage. As an extended task, we generate goal predictions to various points across the retina to localize eye geometry and further generate safe trajectories within the estimated confines. Through experiments in both simulation and with several eye phantoms, we demonstrate that our framework can permit navigation to various points on the retina within 0.089mm and 0.118mm in xy error which is less than the human's surgeon mean tremor at the tool-tip of 0.180mm. All safety constraints were fulfilled and the algorithm was robust to previously unseen eyes as well as unseen objects in the scene. Live video demonstration is available here: https://youtu.be/n5j5jCCelXk

cs.RO↗

arXiv2020

Autonomously Navigating a Surgical Tool Inside the Eye by Learning from Demonstration

A fundamental challenge in retinal surgery is safely navigating a surgical tool to a desired goal position on the retinal surface while avoiding damage to surrounding tissues, a procedure that typically requires tens-of-microns accuracy. In practice, the surgeon relies on depth-estimation skills to localize the tool-tip with respect to the retina in order to perform the tool-navigation task, which can be prone to human error. To alleviate such uncertainty, prior work has introduced ways to assist the surgeon by estimating the tool-tip distance to the retina and providing haptic or auditory feedback. However, automating the tool-navigation task itself remains unsolved and largely unexplored. Such a capability, if reliably automated, could serve as a building block to streamline complex procedures and reduce the chance for tissue damage. Towards this end, we propose to automate the tool-navigation task by learning to mimic expert demonstrations of the task. Specifically, a deep network is trained to imitate expert trajectories toward various locations on the retina based on recorded visual servoing to a given goal specified by the user. The proposed autonomous navigation system is evaluated in simulation and in physical experiments using a silicone eye phantom. We show that the network can reliably navigate a needle surgical tool to various desired locations within 137 microns accuracy in physical experiments and 94 microns in simulation on average, and generalizes well to unseen situations such as in the presence of auxiliary surgical tools, variable eye backgrounds, and brightness conditions.

cs.RO↗