arXiv subjects

Marc D. Killpack

Publications and source records attributed to Marc D. Killpack.

13 recordsLinked to original sources

arXiv2025

Zero-shot Whole-Body Manipulation with a Large-Scale Soft Robotic Torso via Guided Reinforcement Learning

Whole-body manipulation is a powerful yet underexplored approach that enables robots to interact with large, heavy, or awkward objects using more than just their end-effectors. Soft robots, with their inherent passive compliance, are particularly well-suited for such contact-rich manipulation tasks, but their uncertainties in kinematics and dynamics pose significant challenges for simulation and control. In this work, we address this challenge with a simulation that can run up to 350x real time on a single thread in MuJoCo and provide a detailed analysis of the critical tradeoffs between speed and accuracy for this simulation. Using this framework, we demonstrate a successful zero-shot sim-to-real transfer of a learned whole-body manipulation policy, achieving an 88% success rate on the Baloo hardware platform. We show that guiding RL with a simple motion primitive is critical to this success where standard reward shaping methods struggled to produce a stable and successful policy for whole-body manipulation. Furthermore, our analysis reveals that the learned policy does not simply mimic the motion primitive. It exhibits beneficial reactive behavior, such as re-grasping and perturbation recovery. We analyze and contrast this learned policy against an open-loop baseline to show that the policy can also exhibit aggressive over-corrections under perturbation. To our knowledge, this is the first demonstration of forceful, six-DoF whole-body manipulation using two continuum soft arms on a large-scale platform (10 kg payloads), with zero-shot policy transfer.

cs.RO↗

arXiv2025

Scaling Fabric-Based Piezoresistive Sensor Arrays for Whole-Body Tactile Sensing

Scaling tactile sensing for robust whole-body manipulation is a significant challenge, often limited by wiring complexity, data throughput, and system reliability. This paper presents a complete architecture designed to overcome these barriers. Our approach pairs open-source, fabric-based sensors with custom readout electronics that reduce signal crosstalk to less than 3.3% through hardware-based mitigation. Critically, we introduce a novel, daisy-chained SPI bus topology that avoids the practical limitations of common wireless protocols and the prohibitive wiring complexity of USB hub-based systems. This architecture streams synchronized data from over 8,000 taxels across 1 square meter of sensing area at update rates exceeding 50 FPS, confirming its suitability for real-time control. We validate the system's efficacy in a whole-body grasping task where, without feedback, the robot's open-loop trajectory results in an uncontrolled application of force that slowly crushes a deformable cardboard box. With real-time tactile feedback, the robot transforms this motion into a gentle, stable grasp, successfully manipulating the object without causing structural damage. This work provides a robust and well-characterized platform to enable future research in advanced whole-body control and physical human-robot interaction.

cs.RO↗

arXiv2025

PneuDrive: An Embedded Pressure Control System and Modeling Toolkit for Large-Scale Soft Robots

In this paper, we present a modular pressure control system called PneuDrive that can be used for large-scale, pneumatically-actuated soft robots. The design is particularly suited for situations which require distributed pressure control and high flow rates. Up to four embedded pressure control modules can be daisy-chained together as peripherals on a robust RS-485 bus, enabling closed-loop control of up to 16 valves with pressures ranging from 0-100 psig (0-689 kPa) over distances of more than 10 meters. The system is configured as a C++ ROS node by default. However, independent of ROS, we provide a Python interface with a scripting API for added flexibility. We demonstrate our implementation of PneuDrive through various trajectory tracking experiments for a three-joint, continuum soft robot with 12 different pressure inputs. Finally, we present a modeling toolkit with implementations of three dynamic actuation models, all suitable for real-time simulation and control. We demonstrate the use of this toolkit in customizing each model with real-world data and evaluating the performance of each model. The results serve as a reference guide for choosing between several actuation models in a principled manner. A video summarizing our results can be found here: https://bit.ly/3QkrEqO.

cs.RO↗

arXiv2024

Learning Dynamic Tasks on a Large-scale Soft Robot in a Handful of Trials

Soft robots offer more flexibility, compliance, and adaptability than traditional rigid robots. They are also typically lighter and cheaper to manufacture. However, their use in real-world applications is limited due to modeling challenges and difficulties in integrating effective proprioceptive sensors. Large-scale soft robots ($\approx$ two meters in length) have greater modeling complexity due to increased inertia and related effects of gravity. Common efforts to ease these modeling difficulties such as assuming simple kinematic and dynamics models also limit the general capabilities of soft robots and are not applicable in tasks requiring fast, dynamic motion like throwing and hammering. To overcome these challenges, we propose a data-efficient Bayesian optimization-based approach for learning control policies for dynamic tasks on a large-scale soft robot. Our approach optimizes the task objective function directly from commanded pressures, without requiring approximate kinematics or dynamics as an intermediate step. We demonstrate the effectiveness of our approach through both simulated and real-world experiments.

cs.RO↗

arXiv2024

Baloo: A Large-Scale Hybrid Soft Robotic Torso for Whole-Arm Manipulation

Soft robotic actuators can simplify the design of controllers when operating in contact-rich environments. Importantly, their passive compliance fundamentally alters contact mechanics by smoothing impacts and distributing forces over large areas. By integrating soft actuators, we can perform high-impact, dynamic, and contact-rich tasks that are challenging or impossible for traditional rigid robots. In order to explore the benefits of passive structural compliance and learn to utilize it effectively, we present a prototype robotic torso named Baloo. Baloo's hybrid soft-rigid design incorporates both adaptability from soft components and strength from rigid components with two meter-long, pneumatic robot arms mounted on a rigid torso. The hybrid design is capable of lifting end effector payloads of up to 19 kg, far exceeding many hybrid robot designs. Such payloads are competitive with similar-sized rigid robots, but with a much higher strength-to-weight ratio. Through 30 physical whole-body grasping experiments, we also demonstrate how a simple control strategy can generalize for effective lifting across six challenging objects with diverse shapes, sizes, and weights. A 100% success rate across all objects--achieved with the simple control strategy--underscores the potential of our hybrid soft-rigid robot design for contact-rich, whole-body tasks.

cs.RO↗

arXiv2024

A Decomposition of Interaction Force for Multi-Agent Co-Manipulation

Multi-agent human-robot co-manipulation is a poorly understood process with many inputs that potentially affect agent behavior. This paper explores one such input known as interaction force. Interaction force is potentially a primary component in communication that occurs during co-manipulation. There are, however, many different perspectives and definitions of interaction force in the literature. Therefore, a decomposition of interaction force is proposed that provides a consistent way of ascertaining the state of an agent relative to the group for multi-agent co-manipulation. This proposed method extends a current definition from one to four degrees of freedom, does not rely on a predefined object path, and is independent of the number of agents acting on the system and their locations and input wrenches (forces and torques). In addition, all of the necessary measures can be obtained by a self-contained robotic system, allowing for a more flexible and adaptive approach for future co-manipulation robot controllers.

cs.RO↗

arXiv2020

A Versatile Multi-Robot Monte Carlo Tree Search Planner for On-Line Coverage Path Planning

Mobile robots hold great promise in reducing the need for humans to perform jobs such as vacuuming, seeding,harvesting, painting, search and rescue, and inspection. In practice, these tasks must often be done without an exact map of the area and could be completed more quickly through the use of multiple robots working together. The task of simultaneously covering and mapping an area with multiple robots is known as multi-robot on-line coverage and is a growing area of research. Many multi-robot on-line coverage path planning algorithms have been developed as extensions of well established off-line coverage algorithms. In this work we introduce a novel approach to multi-robot on-line coverage path planning based on a method borrowed from game theory and machine learning- Monte Carlo Tree Search. We implement a Monte Carlo Tree Search planner and compare completion times against a Boustrophedon-based on-line multi-robot planner. The MCTS planner is shown to perform on par with the conventional Boustrophedon algorithm in simulations varying the number of robots and the density of obstacles in the map. The versatility of the MCTS planner is demonstrated by incorporating secondary objectives such as turn minimization while performing the same coverage task. The versatility of the MCTS planner suggests it is well suited to many multi-objective tasks that arise in mobile robotics.

cs.MA↗

arXiv2020

Parameterized and GPU-Parallelized Real-Time Model Predictive Control for High Degree of Freedom Robots

This work presents and evaluates a novel input parameterization method which improves the tractability of model predictive control (MPC) for high degree of freedom (DoF) robots. Experimental results demonstrate that by parameterizing the input trajectory more than three quarters of the optimization variables used in traditional MPC can be eliminated with practically no effect on system performance. This parameterization also leads to trajectories which are more conservative, producing less overshoot in underdamped systems with modeling error. In this paper we present two MPC solution methods that make use of this parameterization. The first uses a convex solver, and the second makes use of parallel computing on a graphics processing unit (GPU). We show that both approaches drastically reduce solve times for large DoF, long horizon MPC problems allowing solutions at real-time rates. Through simulation and hardware experiments, we show that the parameterized convex solver MPC has faster solve times than traditional MPC for high DoF cases while still achieving similar performance. For the GPU-based MPC solution method, we use an evolutionary algorithm and that we call Evolutionary MPC (EMPC). EMPC is shown to have even faster solve times for high DoF systems. Solve times for EMPC are shown to decrease even further through the use of a more powerful GPU. This suggests that parallelized MPC methods will become even more advantageous with the improvement and prevalence of GPU technology.

eess.SY↗

arXiv2020

Human-robot co-manipulation of extended objects: Data-driven models and control from analysis of human-human dyads

Human teams are able to easily perform collaborative manipulation tasks. However, for a robot and human to simultaneously manipulate an extended object is a difficult task using existing methods from the literature. Our approach in this paper is to use data from human-human dyad experiments to determine motion intent which we use for a physical human-robot co-manipulation task. We first present and analyze data from human-human dyads performing co-manipulation tasks. We show that our human-human dyad data has interesting trends including that interaction forces are non-negligible compared to the force required to accelerate an object and that the beginning of a lateral movement is characterized by distinct torque triggers from the leader of the dyad. We also examine different metrics to quantify performance of different dyads. We also develop a deep neural network based on motion data from human-human trials to predict human intent based on past motion. We then show how force and motion data can be used as a basis for robot control in a human-robot dyad. Finally, we compare the performance of two controllers for human-robot co-manipulation to human-human dyad performance.

cs.RO↗

arXiv2017

Estimating Human Intent for Physical Human-Robot Co-Manipulation

Human teams can be exceptionally efficient at adapting and collaborating during manipulation tasks using shared mental models. However, the same shared mental models that can be used by humans to perform robust low-level force and motion control during collaborative manipulation tasks are non-existent for robots. For robots to perform collaborative tasks with people naturally and efficiently, understanding and predicting human intent is necessary. However, humans are difficult to predict and model. We have completed an exploratory study recording motion and force for 20 human dyads moving an object in tandem in order to better understand how they move and how their movement can be predicted. In this paper, we show how past motion data can be used to predict human intent. In order to predict human intent, which we equate with the human team's velocity for a short time horizon, we used a neural network. Using the previous 150 time steps at a rate of 200 Hz, human intent can be predicted for the next 50 time steps with a mean squared error of 0.02 (m/s)^2. We also show that human intent can be estimated in a human-robot dyad. This work is an important first step in enabling future work of integrating human intent estimation on a robot controller to execute a short-term collaborative trajectory.

cs.RO↗

arXiv2017

Analysis of Rigid Extended Object Co-Manipulation by Human Dyads: Lateral Movement Characterization

During co-manipulation involving humans and robots, it is necessary to base robot controllers on human behaviors to achieve comfortable and coordinated movement between the human-robot dyad. In this paper, we describe an experiment between human-human dyads and we record the force and motion data as the leader-follower dyads moved in translation and rotation. The force/motion data was then analyzed for patterns found during lateral translation only. For extended objects, lateral translation and in-place rotation are ambiguous, but this paper determines a way to characterize lateral translation triggers for future use in human-robot interaction. The study has 4 main results. First, interaction forces are apparent and necessary for co-manipulation. Second, minimum-jerk trajectories are found in the lateral direction only for lateral movement. Third, the beginning of a lateral movement is characterized by distinct force triggers by the leader. Last, there are different metrics that can be attributed to determine which dyads moved most effectively in the lateral direction.

cs.RO↗

arXiv2014

Automated Tracking and Estimation for Control of Non-rigid Cloth

This report is a summary of research conducted on cloth tracking for automated textile manufacturing during a two semester long research course at Georgia Tech. This work was completed in 2009. Advances in current sensing technology such as the Microsoft Kinect would now allow me to relax certain assumptions and generally improve the tracking performance. This is because a major part of my approach described in this paper was to track features in a 2D image and use these to estimate the cloth deformation. Innovations such as the Kinect would improve estimation due to the automatic depth information obtained when tracking 2D pixel locations. Additionally, higher resolution camera images would probably give better quality feature tracking. However, although I would use different technology now to implement this tracker, the algorithm described and implemented in this paper is still a viable approach which is why I am publishing this as a tech report for reference. In addition, although the related work is a bit exhaustive, it will be useful to a reader who is new to methods for tracking and estimation as well as modeling of cloth.

cs.CV↗

arXiv2013

Manipulation in Clutter with Whole-Arm Tactile Sensing

We begin this paper by presenting our approach to robot manipulation, which emphasizes the benefits of making contact with the world across the entire manipulator. We assume that low contact forces are benign, and focus on the development of robots that can control their contact forces during goal-directed motion. Inspired by biology, we assume that the robot has low-stiffness actuation at its joints, and tactile sensing across the entire surface of its manipulator. We then describe a novel controller that exploits these assumptions. The controller only requires haptic sensing and does not need an explicit model of the environment prior to contact. It also handles multiple contacts across the surface of the manipulator. The controller uses model predictive control (MPC) with a time horizon of length one, and a linear quasi-static mechanical model that it constructs at each time step. We show that this controller enables both real and simulated robots to reach goal locations in high clutter with low contact forces. Our experiments include tests using a real robot with a novel tactile sensor array on its forearm reaching into simulated foliage and a cinder block. In our experiments, robots made contact across their entire arms while pushing aside movable objects, deforming compliant objects, and perceiving the world.

cs.RO↗