Explore the vast range of titles available.

Acidic macromolecules are traditionally considered key to calcium carbonate biomineralisation and have long been first choice in the bio-inspired synthesis of crystalline materials. Here, we challenge this view and demonstrate that low-charge macromolecules can vastly outperform their acidic counterparts in the synthesis of nanocomposites. Using gold nanoparticles functionalised with low charge, hydroxyl-rich proteins and homopolymers as growth additives, we show that extremely high concentrations of nanoparticles can be incorporated within calcite single crystals, while maintaining the continuity of the lattice and the original rhombohedral morphologies of the crystals. The nanoparticles are perfectly dispersed within the host crystal and at high concentrations are so closely apposed that they exhibit plasmon coupling and induce an unexpected contraction of the crystal lattice. The versatility of this strategy is then demonstrated by extension to alternative host crystals. This simple and scalable occlusion approach opens the door to a novel class of single crystal nanocomposites. Calcium carbonate biomineralisation has long been linked to acidic macromolecules. Here, the authors challenge this view and show that a huge number of gold nanoparticles coated with hydroxyl-rich proteins can be incorporated into a calcium carbonate crystal while maintaining single crystal character.

Over many millions of years, evolution has provided living organisms with the tools to control the growth and properties of materials from the molecular scale upward. One of the many ways this is achieved is through the introduction of impurities into the solution in which these materials grow. A long-term goal of materials scientists is to harness nature's control mechanisms and apply them in the world of engineering. However, these mechanisms of growth control are highly complex, and understanding them requires insight into physical processes at the molecular scale. While experiments are so-far unable to offer such a high resolution, computer simulations can be used to directly model these physical process with no limit on the resolution. Throughout this thesis, an array of computational methodologies is applied to calcite in an attempt to understand how impurities are able to drive the growth process, and ultimately alter the mechanical properties of the crystal. A series of metadynamics simulations are applied to calcite kink sites, revealing a more complex growth mechanism in which kink-terminating ions do not initially occupy their crystal lattice sites, and only do so upon the adsorption of an additional solute. A combination of metadynamics and Kinetic Monte Carlo simulations are used to examine the adsorption free energies and growth inhibiting properties of amino acids and polyamines, the results of which are compared directly to experiment. This offers a robust insight into the molecular mechanisms that underpin how organic molecules are able to tune the growth of calcite. Simulations are also applied to two case studies of impure calcite. By examining lattice spacings, determining stress distributions and simulating a series of crack propagation events, insight into mechanisms through which biogenic crystals exhibit superior mechanical properties is found. Finally, the nature of non-Markovianity when using reaction coordinates -such as those used in rare event methodologies applied throughout this thesis- are investigated. By introducing non-Markovianity into the system, barrier crossing rates in a coarse-grained system more closely resemble those in the original two-dimensional system. Furthermore, we study the breakdown in rare-events sampling when a poor reaction coordinate is used, and identify which rare-events sampling techniques are more appropriate for detecting poor reaction coordinate choices.

The theory of how humans and machines control and communicate with each other is at the core of the scientific field known as Human-Robot Interaction (HRI). Researchers in this sub-discipline of robotics are therefore particularly interested in developing methods to reduce the inherent friction in this communication and control channel. Just as can be observed in the analogous problem of collaboration between two human partners, solutions in this space require a tight coupling between a human partner and an autonomous partner. A conceptual framework that describes this exact relationship is known as shared control (SC). Shared control defines an abstract link between a set of partners (often a human operator and an autonomous agent) that are both responsible for providing control information to the same robotic device. This paradigm is especially useful as a method of extending the physical capabilities of a human operator, while simultaneously considering important constraints defined by the user and environment.This dissertation is largely motivated by applications of shared control in the fields of assistive and rehabilitation medicine. Therefore, this thesis develops shared control solutions that are designed specifically to improve, or restore, a human operator's ability to control complex mechanical devices. Example motivating systems include powered wheelchairs, exoskeletons, and robotic manipulators. In addition to increasing a human operator's capabilities, a particularly desirable attribute of any interactive system in assistive and rehabilitation medicine is the acceptance, and enjoyment, of the human-in-the-loop. For this reason, the SC algorithms described in this dissertation allocate the majority of the control authority to the human partner, while the autonomous partner is mainly responsible for providing control information to improve the stability and safety of the joint human-machine system.The specific techniques described in this dissertation are motivated by the desire to generalize solutions in shared control to generic pairs of human and machine partners, while simultaneously developing a decision making framework that is responsive to the individual human-in-the-loop. To address this desire, this thesis introduces the notion of data-driven model-based shared control (MbSC). Data-driven MbSC extends the efficacy of standard shared control systems to scenarios in which we do not have any prior knowledge of the system dynamics or the human operator. Instead, data-driven MbSC relies on techniques from (1) machine learning to gain an understanding of the joint human-machine system from observation, and (2) optimal control (OC) to develop a control policy for the autonomous partner. The shared control system then allocates authority to each partner to improve desired outcomes (e.g. task-success, stability, and/or safety). Additionally, this dissertation describes data-driven techniques that further personalize the interaction paradigm to the individual human-in-the-loop. The proposed methodology uses a representation of the autonomy's trust in the human partner's control skill learned from observation data. This data-driven metric is then used to modulate the control authority granted to each partner in real-time. Taken together, the techniques described in this thesis describe a generalizable solution to the shared control problem that can be personalized to the individual human-in-the-loop to improve the capabilities of the joint system.

We present a data-driven shared control algorithm that can be used to improve a human operator's control of complex dynamic machines and achieve tasks that would otherwise be challenging, or impossible, for the user on their own. Our method assumes no a priori knowledge of the system dynamics. Instead, both the dynamics and information about the user's interaction are learned from observation through the use of a Koopman operator. Using the learned model, we define an optimization problem to compute the autonomous partner's control policy. Finally, we dynamically allocate control authority to each partner based on a comparison of the user input and the autonomously generated control. We refer to this idea as model-based shared control (MbSC). We evaluate the efficacy of our approach with two human subjects studies consisting of 32 total participants (16 subjects in each study). The first study imposes a linear constraint on the modeling and autonomous policy generation algorithms. The second study explores the more general, nonlinear variant. Overall, we find that model-based shared control significantly improves task and control metrics when compared to a natural learning, or user only, control paradigm. Our experiments suggest that models learned via the Koopman operator generalize across users, indicating that it is not necessary to collect data from each individual user before providing assistance with MbSC. We also demonstrate the data-efficiency of MbSC and consequently, it's usefulness in online learning paradigms. Finally, we find that the nonlinear variant has a greater impact on a user's ability to successfully achieve a defined task than the linear variant.

Incorporation of guest additives within inorganic single crystals offers a unique strategy for creating nanocomposites with tailored properties. While anionic additives have been widely used to control the properties of crystals, their effective incorporation remains a key challenge. Here, we show that cationic additives are an excellent alterative for the synthesis of nanocomposites, where they are shown to deliver exceptional levels of incorporation of up to 70 wt% of positively charged amino acids, polymer particles, gold nanoparticles, and silver nanoclusters within inorganic single crystals. This high additive loading endows the nanocomposites with new functional properties including plasmon coupling, bright fluorescence, and surface-enhanced Raman scattering (SERS). Cationic additives are also shown to outperform their acidic counterparts, where they are highly active in a wider range of crystal systems, owing to their outstanding colloidal stability in the crystallization media and strong affinity for the crystal surfaces. This work demonstrates that although often overlooked, cationic additives can make valuable crystallization additives to create composite materials with tailored composition-structure-property relationships. This versatile and straightforward approach advances the field of single-crystal composites and provides exciting prospects for the design and fabrication of new hybrid materials with tunable functional properties.

We present a novel object detection pipeline for localization and recognition in three dimensional environments. Our approach makes use of an RGB-D sensor and combines state-of-the-art techniques from the robotics and computer vision communities to create a robust, real-time detection system. We focus specifically on solving the object detection problem for tabletop scenes, a common environment for assistive manipulators. Our detection pipeline locates objects in a point cloud representation of the scene. These clusters are subsequently used to compute a bounding box around each object in the RGB space. Each defined patch is then fed into a Convolutional Neural Network (CNN) for object recognition. We also demonstrate that our region proposal method can be used to develop novel datasets that are both large and diverse enough to train deep learning models, and easy enough to collect that end-users can develop their own datasets. Lastly, we validate the resulting system through an extensive analysis of the accuracy and run-time of the full pipeline.

We present a shared control paradigm that improves a user's ability to operate complex, dynamic systems in potentially dangerous environments without a priori knowledge of the user's objective. In this paradigm, the role of the autonomous partner is to improve the general safety of the system without constraining the user's ability to achieve unspecified behaviors. Our approach relies on a data-driven, model-based representation of the joint human-machine system to evaluate, in parallel, a significant number of potential inputs that the user may wish to provide. These samples are used to (1) predict the safety of the system over a receding horizon, and (2) minimize the influence of the autonomous partner. The resulting shared control algorithm maximizes the authority allocated to the human partner to improve their sense of agency, while improving safety. We evaluate the efficacy of our shared control algorithm with a human subjects study (n=20) conducted in two simulated environments: a balance bot and a race car. During the experiment, users are free to operate each system however they would like (i.e., there is no specified task) and are only asked to try to avoid unsafe regions of the state space. Using modern computational resources (i.e., GPUs) our approach is able to consider more than 10,000 potential trajectories at each time step in a control loop running at 100Hz for the balance bot and 60Hz for the race car. The results of the study show that our shared control paradigm improves system safety without knowledge of the user's goal, while maintaining high-levels of user satisfaction and low-levels of frustration. Our code is available online at https://github.com/asbroad/mpmi_shared_control.

We describe a shared control methodology that can, without knowledge of the task, be used to improve a human's control of a dynamic system, be used as a training mechanism, and be used in conjunction with Imitation Learning to generate autonomous policies that recreate novel behaviors. Our algorithm introduces autonomy that assists the human partner by enforcing safety and stability constraints. The autonomous agent has no a priori knowledge of the desired task and therefore only adds control information when there is concern for the safety of the system. We evaluate the efficacy of our approach with a human subjects study consisting of 20 participants. We find that our shared control algorithm significantly improves the rate at which users are able to successfully execute novel behaviors. Experimental results suggest that the benefits of our safety-aware shared control algorithm also extend to the human partner's understanding of the system and their control skill. Finally, we demonstrate how a combination of our safety-aware shared control algorithm and Imitation Learning can be used to autonomously recreate the demonstrated behaviors.

We present a novel approach to shared control of human-machine systems. Our method assumes no a priori knowledge of the system dynamics. Instead, we learn both the dynamics and information about the user's interaction from observation through the use of the Koopman operator. Using the learned model, we define an optimization problem to compute the optimal policy for a given task, and compare the user input to the optimal input. We demonstrate the efficacy of our approach with a user study. We also analyze the individual nature of the learned models by comparing the effectiveness of our approach when the demonstration data comes from a user's own interactions, from the interactions of a group of users and from a domain expert. Positive results include statistically significant improvements on task metrics when comparing a user-only control paradigm with our shared control paradigm. Surprising results include findings that suggest that individualizing the model based on a user's own data does not effect the ability to learn a useful dynamic system. We explore this tension as it relates to developing human-in-the-loop systems further in the discussion.

We present a structured neural network architecture that is inspired by linear time-varying dynamical systems. The network is designed to mimic the properties of linear dynamical systems which makes analysis and control simple. The architecture facilitates the integration of learned system models with gradient-based model predictive control algorithms, and removes the requirement of computing potentially costly derivatives online. We demonstrate the efficacy of this modeling technique in computing autonomous control policies through evaluation in a variety of standard continuous control domains.