<p> Trajectory planning is a field with widespread utility, and imitation learning pipelines<br> show promise as an accessible training method for trajectory planning. MPNet is the state<br> of the art for imitation learning with respect to success rates. MPNet has two general<br> components to its runtime: a neural network predicts the location of the next anchor point in<br> a trajectory, and then planning infrastructure applies sampling-based techniques to produce<br> near-optimal, collision-less paths. This distinction between the two parts of MPNet prompts<br> investigation into the role of the neural architectures in the Neural Motion Planning pipeline,<br> to discover where improvements can be made. This thesis seeks to explore the importance<br> of neural architecture choice by removing the planning structures, and comparing MPNet’s<br> feedforward anchor point predictor with that of a continuous model trained to output a<br> continuous trajectory from start to goal. A new state of the art model in continuous learning<br> is the Neural Flow model. As a continuous model, it possess a low standard deviation runtime<br> which can be properly leveraged in the absence of planning infrastructure. Neural Flows also<br> output smooth, continuous trajectory curves that serve to reduce noisy path outputs in the<br> absence of lazy vertex contraction. This project analyzes the performance of MPNet, Resnet<br> Flow, and Coupling Flow models when sampling-based planning tools such as dropout, lazy<br> vertex contraction, and replanning are removed. Each neural planner is trained end-to-end in<br> an imitation learning pipeline utilizing a simple feedforward encoder, a CNN-based encoder,<br> and a Pointnet encoder to encode the environment, for purposes of comparison. Results<br> indicate that performance is competitive, with Neural Flows slightly outperforming MPNet’s<br> success rates on our reduced dataset in Simple2D, and being slighty outperformed by MPNet<br> with respect to collision penetration distance in our UR5 Cubby test suite. These results<br> indicate that continuous models can compete with the performance of anchor point predictor<br> models when sampling-based planning techniques are not applied. Neural Flow models also<br> have other benefits that anchor point predictors do not, like continuity guarantees, the ability<br> to select a proportional location in a trajectory to output, and smoothness. </p>

<p dir="ltr">Path-tracking control is an integral part of motion planning in autonomous vehicles, in which the vehicle's lateral and longitudinal positions are controlled by a control system that will provide acceleration and steering angle commands to ensure accurate tracking of longitudinal and lateral movements in reference to a pre-defined trajectory. Extensive research has been conducted to address the growing need for efficient algorithms in this area. In this dissertation, a scenario and machine learning-based data-driven control approach is proposed for a path-tracking controller. Firstly, a Deep Reinforcement Learning model is developed to facilitate the control of longitudinal speed. A Deep Deterministic Policy Gradient algorithm is employed as the primary algorithm in training the reinforcement learning model. The main objective of this model is to maintain a safe distance from a lead vehicle (if present) or track a velocity set by the driver. Secondly, a lateral steering controller is developed using Neural Networks to control the steering angle of the vehicle with the main goal of following a reference trajectory. Then, a path-planning algorithm is developed using a hybrid A* planner. Finally, the longitudinal and lateral control models are coupled together to obtain a complete path-tracking controller that follows a path generated by the hybrid A* algorithm at a wide range of vehicle speeds. The state-of-the-art path-tracking controller is also built using Model Predictive Control and Stanley control to evaluate the performance of the proposed model. The results showed the effectiveness of both proposed models in the same scenario, in terms of velocity error, lateral yaw angle error, and lateral distance error. The results from the simulation show that the developed hybrid A* algorithm has good performance in comparison to the state-of-the-art path planning algorithms.</p>

<p>Soft robotics is concerned with the modeling and designing of devices fabricated from materials with low Young’s moduli—much less than that of metal— that mimic the input/output operation and physical task utility of robotics.  The inherent compliance of soft robots lends these devices an adaptability and a capacity for human-machine interaction beyond that of conventional robotics. Multistable soft robotic grippers are a subset of the technology at the intersection of soft robotics and multistable structures. Multistable structures are continuum systems that exhibit more than one statically stable state, each associated with a strain energy minimum. The existence of these energetic minima allows the structures to adopt different stable configurations that can provide a reference point for open loop control schemes. Multistable soft robotics takes advantage of both the adaptability of soft robotics and the potential for simplified control of multistable structures.</p> <p>Achieving simplified control for soft robotics is a necessary milestone in creating functional and applied soft robots. </p> <p>This work presents a means for simple open-loop control of a multistable soft robotic gripper that is adaptable, controllable, and robust. The behavior is illustrated through a gripper geometry described by specific design parameters resulting in a near infinite design space. An analytical model based on lumped parameter springs is derived, allowing us to search the design space in a tractable fashion. Specifically, we predict the system’s stable states for any given design instance by searching for local minima in the energy landscape formed by a spring lattice representation of our device. The lattice is composed of linear, bistable, and torsional springs—each of which contributes to the energy landscape of the system. We validate our model against Finite Element simulations of our device, showing good agreement with the proposed model. The aptitude of the model sheds light on the fundamental mechanics of our soft robotic gripper topology, laying the foundation for efficient design optimization and simplified control of soft robots.</p>

<p>Human multi-robot systems constitute a relatively new area of research that focuses on the interaction and collaboration between humans and multiple robots. Well-designed systems can enable a team of humans and robots to effectively work together on complex and sophisticated tasks such as exploration, monitoring, and search and rescue operations. This dissertation introduces an affective workload allocation system capable of adaptively allocating workload in real-time while considering the conditions and work performance of human operators in multi-human multi-robot teams. The proposed system is largely composed of three parts, taking the surveillance scenario involving multi-human operators and multi-robot system as an example. The first part of the system is a framework for an adaptive multi-human multi-robot system that allows real-time measurement and communication between heterogeneous sensors and multi-robot systems. The second part is an algorithm for real-time monitoring of humans' affective states using machine learning techniques and estimation of the affective state from multimodal data that consists of physiological and behavioral signals. The third part is a deep reinforcement learning-based workload allocation algorithm. For the first part of the affective workload allocation system, we developed a robot operating system (ROS)-based affective monitoring framework to enable communication among multiple wearable biosensors, behavioral monitoring devices, and multi-robot systems using the real-time operating system feature of ROS. We validated the sub-interfaces of the affective monitoring framework through connecting to a robot simulation and utilizing the framework to create a dataset. The dataset included various visual and physiological data categorized on the cognitive load level. The targeted cognitive load is stimulated by a closed-circuit television (CCTV) monitoring task on the surveillance scenario with multi-robot systems. Furthermore, we developed a deep learning-based affective prediction algorithm using the physiological and behavioral data captured from wearable biosensors and behavior-monitoring devices, in order to estimate the cognitive states for the second part of the system. For the third part of the affective workload allocation system, we developed a deep reinforcement learning-based workload allocation algorithm to allocate optimal workloads based on a human operator's performance. The algorithm was designed to take an operator's cognitive load, using objective and subjective measurements as inputs, and consider the operator's task performance model we developed using the empirical findings of the extensive user experiments, to allocate optimal workloads to human operators. We validated the proposed system through within-subjects study experiments on a generalized surveillance scenario involving multiple humans and multiple robots in a team. The multi-human multi-robot surveillance environment included an affective monitoring framework and an affective prediction algorithm to read sensor data and predict human cognitive load in real-time, respectively. We investigated optimal methods for affective workload allocations by comparing other allocation strategies used in the user experiments. As a result, we demonstrated the effectiveness and performance of the proposed system. Moreover, we found that the subjective and objective measurement of an operator's cognitive loads and the process of seeking consent for the workload transitions must be included in the workload allocation system to improve the team performance of the multi-human multi-robot teams.</p>

<p dir="ltr">Learning, planning, and representing knowledge at multiple levels of temporal abstractions provide an agent with the ability to predict consequences of different courses of actions, which is essential for improving the performance of sequential decision making. However, discovering effective temporal abstractions, which the agent can use as skills, and adopting the constructed temporal abstractions for efficient policy learning can be challenging. Despite significant advancements in single-agent settings, temporal abstractions in multi-agent systems remains underexplored. This thesis addresses this research gap by introducing novel algorithms for discovering and employing temporal abstractions in both cooperative and competitive multi-agent environments. We first develop an unsupervised spectral-analysis-based discovery algorithm, aiming at finding temporal abstractions that can enhance the joint exploration of agents in complex, unknown environments for goal-achieving tasks. Subsequently, we propose a variational method that is applicable for a broader range of collaborative multi-agent tasks. This method unifies dynamic grouping and automatic multi-agent temporal abstraction discovery, and can be seamlessly integrated into the commonly-used multi-agent reinforcement learning algorithms. Further, for competitive multi-agent zero-sum games, we develop an algorithm based on Counterfactual Regret Minimization, which enables agents to form and utilize strategic abstractions akin to routine moves in chess during strategy learning, supported by solid theoretical and empirical analyses. Collectively, these contributions not only advance the understanding of multi-agent temporal abstractions but also present practical algorithms for intricate multi-agent challenges, including control, planning, and decision-making in complex scenarios.</p>

Made available in DSpace on 2015-03-05T13:53:43Z (GMT). No. of bitstreams: 0 Previous issue date: 28 / Nenhuma / Neste trabalho foi desenvolvido um sistema de controle robusto para robôs móveis autônomos que é capaz de operar e de se adaptar a diferentes ambientes e condições. Para isso foi proposta uma arquitetura de controle híbrida (COHBRA), integrando as duas principais técnicas de controle robótico (controle deliberativo e controle reativo). Esta arquitetura de controle utiliza uma abordagem de três camadas para integrar uma camada vital (controle reativo), uma camada funcional (seqüenciador) e uma camada deliberativa (controle deliberativo). A comunicação entre as diversas camadas é realizada através de uma área de memória compartilhada, inspirada na abordagem Blackboard. A arquitetura de controle possui um esquema de múltiplas representações internas do ambiente: representação poligonal, representação matricial e representação topológica / semântica. O sistema de controle desenvolvido tem a capacidade de navegar em um ambiente dinâmico, desviando tanto de obstáculos estáticos como de obstáculos móveis / In this work we developed a robust control system for autonomous mobile robots capable of operating and adapting in various environments and conditions. In order to accomplish this objective an hybrid control architecture (COHBRA) was proposed, integrating the two main techniques of robotic control: deliberative control and reactive control. This control architecture uses a three layers approach to integrate a vital layer (reactive control), a functional layer (sequencer) and a deliberative layer (deliberative control). The communication between the three layers uses a shared memory approach, inspired in the Blackboard approach. The control architecture has a structure of multiple internal representations of the environment: polygonal representation, matricial representation and topological/semantic representation. The control system has the ability to navigate in a dynamic environment, avoiding static obstacles and unexpected mobile obstacles. The deliberative layer uses the A* algorithm to calcu

<div>Multi-robot systems have demonstrated strong capability in handling environmental operations. In this study, We examine how a team of robots can be utilized in covering and removing spill patches in a dynamic environment by executing three consecutive stages: deployment, coverage, and rendezvous. </div><div> </div><div>For the deployment problem, we aim for robot allocation based on the discreteness of the patches that need to be covered. With the deep neural network (DNN) based spill detector and remote sensing facilities such as drones with vision sensors and satellites, we are able to obtain the spill distribution in the workspace. Then, we formulate the allocation problem in a general optimization form and provide solutions using an integer linear programming (ILP) solver under several realistic constraints. After the allocation process is completed and the robot team is divided according to the number of spills, we deploy robots to their computed optimal goal positions. In the robot deployment part, control laws based on artificial potential field (APF) method are proposed and practiced on robots with a common unicycle model. </div><div> </div><div>For the coverage control problem, we show two strategies that are tailored for a wirelessly networked robot team. We propose strategies for coverage with and without path planning, depending on the availability of global information. Specifically, in terms of coverage with path planning, we partition the workspace from the aerial image into pieces and let each robot take care of one of the pieces. However, path-planning-based coverage relies on GPS signals or other external positioning systems, which are not applicable for indoor or GPS-denied circumstances. Therefore, we propose an asymptotic boundary shrink control that enables a collective coverage operation with the robot team. Such a strategy does not require a planned path, and because of its distributedness, it shows many advantages, including system scalability, dynamic spill adaptability, and collision avoidance. In case of a large-scale patch that poses challenges to robot connectivity maintenance during the operation, we propose a pivot-robot coverage strategy by mean of an a priori geometric tessellation (GT). In the pivot-robot-based coverage strategy, a team of robots is sent to perform complete coverage to every packing area of GT in sequence. Ultimately, the entire spill in the workspace can be covered and removed.</div><div> </div><div>For the rendezvous problem, we investigate the use of graph theory and propose control strategies based on network topology to motivate robots to meet at a designated or the optimal location. The rendezvous control strategies show a strong robustness to some common failures, such as mobility failure and communication failure. To expedite the rendezvous process and enable herding control in a distributed way, we propose a multi-robot multi-point rendezvous control strategy. </div><div> </div><div>To verify the validity of the proposed strategies, we carry out simulations in the Robotarium MATLAB platform, which is an open source swarm robotics experiment testbed, and conduct real experiments involving multiple mobile robots.</div>

<p>Robots have become increasingly common in various sectors, such as manufacturing, healthcare, and service industries. With the growing demand for automation and the expectation for interactive and assistive capabilities, robots must learn to adapt to unpredictable environments like humans can. This necessitates the development of learning methods that can effectively enable robots to collaborate with humans, learn from them, and provide guidance. Human experts commonly teach their collaborators to perform tasks via a few demonstrations, often followed by episodes of coaching that refine the trainee’s performance during practice. Adopting a similar approach that facilitates interactions to teaching robots is highly intuitive and enables task experts to teach the robots directly. Learning from Demonstration (LfD) is a popular method for robots to learn tasks by observing human demonstrations. However, for contact-rich tasks such as cleaning, cutting, or writing, LfD alone is insufficient to achieve a good performance. Further, LfD methods are developed to achieve observed goals while ignoring actions to maximize efficiency. By contrast, we recognize that leveraging human social learning strategies of practice and coaching in conjunction enables learning tasks with improved performance and efficacy. To address the deficiencies of learning from demonstration, we propose a Coaching by Demonstration (CbD) framework that integrates LfD-based practice with sparse coaching interactions from a human expert.</p> <p><br></p> <p>The LfD-based practice in CbD was implemented as an end-to-end off-policy reinforcement learning (RL) agent with the action space and rewards inferred from the demonstration. By modeling the reward as a similarity network trained on expert demonstrations, we eliminate the need for designing task-specific engineered rewards. Representation learning was leveraged to create a novel state feature that captures interaction markers necessary for performing contact-rich skills. This LfD-based practice was combined with coaching, where the human expert can improve or correct the objectives through a series of interactions. The dynamics of interaction in coaching are formalized using a partially observable Markov decision process. The robot aims to learn the true objectives by observing the corrective feedback from the human expert. We provide an approximate solution by reducing this to a policy parameter update using KL divergence between the RL policy and a Gaussian approximation based on coaching. The proposed framework was evaluated on a dataset of 10 contact-rich tasks from the assembly (peg-insertion), service (cleaning, writing, peeling), and medical domains (cricothyroidotomy, sonography). Compared to baselines of behavioral cloning and reinforcement learning algorithms, CbD demonstrates improved performance and efficiency.</p> <p><br></p> <p>During the learning process, the demonstrations and coaching feedback imbue the robot with expert knowledge of the task. To leverage this expertise, we develop a reverse coaching model where the robot can leverage knowledge from demonstrations and coaching corrections to provide guided feedback to human trainees to improve their performance. Providing feedback adapted to individual trainees' "style" is vital to coaching. To this end, we have proposed representing style as objectives in the task null space. Unsupervised clustering of the null-space trajectories using Gaussian mixture models allows the robot to learn different styles of executing the same skill. Given the coaching corrections and style clusters database, a style-conditioned RL agent was developed to provide feedback to human trainees by coaching their execution using virtual fixtures. The reverse coaching model was evaluated on two tasks, a simulated incision and obstacle avoidance through a haptic teleoperation interface. The model improves human trainees’ accuracy and completion time compared to a baseline without corrective feedback. Thus, by taking advantage of different human-social learning strategies, human-robot collaboration can be realized in human-centric environments. </p> <p><br></p>

<p dir="ltr">This paper aims to gain a deep understanding of up-to-date research and development on modular self-reconfigurable robots (MSRs) through a thorough survey of market demands and published works on <i>design methodologies</i>, <i>system integration</i>, <i>advanced controls</i>, and <i>new applications</i>. Some limitations of existing mobile MSR are discussed from the reconfigurability perspective of mechanical structures, and a novel MSR system is proposed to address the identified limitations of existing MSRs. The comprehensive set of <i>Functional Requirements</i> (FRs) of MSRs is discussed, from which the mechanical designs of MSR were created, and the system was prototyped and built for testing. Three main innovations of the designed modules for MSR are to (1) share torque power, (2) customize the size for a given task, and (3) have a low number of actuated motors while still maintain a motion with high <i>Degrees of Freedom</i> (DoF) to overcome the constraints by the power capacities of individual motors; this helps to increase reconfigurability, reduce cost, and reduce the size of conventional MSRs.</p>

<p>This dissertation examines the impact of active shooters upon patrons attending large outdoor events. There has been a spike in shooters targeting densely populated spaces in recent years, to include open-air venues. The 2019 Gilroy Garlic Festival was selected for modeling replication using AnyLogic software to test various experiments designed to reduce casualties in the event of an active shooter situation. Through achievement of validation to produce identical outcomes of the real-world Gilroy Garlic Festival shooting, the researcher established a reliable foundational model for experimental purposes. This active shooter research project identifies the need for rapid response efforts to neutralize the shooter(s) as quickly as possible to minimize casualties. Key findings include the importance of armed officers patrolling event grounds to reduce response time, the need for adequate exits during emergency evacuations, incorporation of modern technology to identify the shooter’s location, and applicability of a 1:548 police to patron ratio.</p>