Human balance and locomotion control is highly complex and not well understood. To understand how the nervous system controls balance and locomotion works, we test how the body responds to controlled perturbations, the results are analyzed, and control models are developed. However, to recreate this system of control there is a need for a robot with human-like kinematics. Unfortunately, such a robotic testbed does not exist despite the numerous applications such a design would have in mobile robotics, healthcare, and prosthetics. This thesis presents a robotic testbed model of human lower legs. By using MRI and CT scans, I designed joints that require lower force for actuation, are more wear resistant, and are less prone to catastrophic failure than a traditional revolute (or pinned) joints. The result of using this process is the design, construction, and performance analysis of a biologically inspired knee joint for use in bipedal robotics. For the knee joint, the design copies the condylar surfaces of the distal end of the femur and utilizes the same crossed four-bar linkage design the human knee uses. The joint includes a changing center of rotation, a screw-home mechanism, and patella; these are characteristics of the knee that are desirable to copy for bipedal robotics. The design was calculated to have an average sliding to rolling ratio of 0.079, a maximum moment arm of 2.7 inches and a range of motion of 151 degrees. This should reduce joint wear and have kinematics similar to the human knee. I also designed and constructed novel, adjustably-damped hip and ankle joints that use braided pneumatic actuators. These joints provide a wide range of motion and exhibit the same change in stiffness that human joints exhibit as flexion increases, increasing stability, adaptability, and controllability. The theoretical behaviors of the joints make them desirable for use in mobile robotics and should provide a lightweight yet mechanically strong connection that is resistant to unexpected perturbations and catastrophic failure. The joints also bridge the gap between completely soft robotics and completely rigid robotics. These joints will give researchers the ability to test different control schemes and will help to determine how human balance is achieved. They will also lead to robots that are lighter and have lower power requirements while increasing the adaptability of the robot. When applying these design principles to joints used for prosthetics, we reduce the discomfort of the wearer and reduce the effort needed to move. Both of which are serious issues for individuals who need to wear a prosthetic device.

We make two main contributions in this thesis. First, we present our approach to robot manipulation, which emphasizes the benefits of making contact with the world across all the surfaces of a manipulator with whole-arm tactile sensing and compliant actuation at the joints. In contrast, many current approaches to mobile manipulation assume most contact is a failure of the system, restrict contact to only occur at well modeled end effectors, and use stiff, precise control to avoid contact. We develop a controller that enables robots with whole-arm tactile sensing and compliant actuation at the joints to reach to locations in high clutter while regulating contact forces. We assume that low contact forces are benign and our controller does not place any penalty on contact forces below a threshold. Our controller only requires haptic sensing, handles multiple contacts across the surface of the manipulator, and does not need an explicit model of the environment prior to contact. It uses model predictive control with a time horizon of length one, and a linear quasi-static mechanical model that it constructs at each time step. We show that our controller enables both a real and simulated robots to reach goal locations in high clutter with low contact forces. While doing so, the robots bend, compress, slide, and pivot around objects. To enable experiments on real robots, we also developed an inexpensive, flexible, and stretchable tactile sensor and covered large surfaces of two robot arms with these sensors. With an informal experiment, we show that our controller and sensor have the potential to enable robots to manipulate in close proximity to, and in contact with humans while keeping the contact forces low. Second, we present an approach to give robots common sense about everyday forces in the form of probabilistic data-driven object-centric models of haptic interactions. These models can be shared by different robots for improved manipulation performance. We use pulling open doors, an important task for service robots, as an example to demonstrate our approach. Specifically, we capture and model the statistics of forces while pulling open doors and drawers. Using a portable custom force and motion capture system, we create a database of forces as human operators pull open doors and drawers in six homes and one office. We then build data-driven models of the expected forces while opening a mechanism, given knowledge of either its class (e.g, refrigerator) or the mechanism identity (e.g, a particular cabinet in Advait's kitchen). We demonstrate that these models can enable robots to detect anomalous conditions such as a locked door, or collisions between the door and the environment faster and with lower excess force applied to the door compared to methods that do not use a database of forces.

The capacity of humankind to mimic fish-like locomotion for engineering applications depends mainly on the availability of suitable actuators. Researchers have recently focused on developing robotic fish using smart materials, particularly Ionic Polymer-Metal Composites (IPMCs), as a compliant, noise-free, and scalable alternative to conventional motor-based propulsion systems. In this thesis, we investigate fish-like self propulsion using flexible bimorphs made of Macro-Fiber Composite (MFC) piezoelectric laminates. Similar to IPMCs, MFCs also exhibit high efficiency in size, energy consumption, and noise reduction. In addition, MFCs offer large dynamic forces in bending actuation, strong electromechanical coupling as well as both low-frequency and high-frequency performance capabilities. The experimental component of the presented work focuses on the characterization of an MFC bimorph propulsor for thrust generation in a quiescent fluid as well as the development of a preliminary robotic fish prototype incorporating a microcontroller and a printed-circuit-board (PCB) amplifier to generate high actuation voltage for battery-powered free locomotion. From the theoretical standpoint, a reliable modeling framework that couples the actuator dynamics, hydroelasticity, and fish locomotion theory is essential to both design and control of robotic fish. Therefore, a distributed-parameter electroelastic model with fluid effects and actuator dynamics is coupled with the elongated body theory. Both in-air and underwater experiments are performed to verify the incorporation of hydrodynamic effects in the linear actuation regime. For electroelastically nonlinear actuation levels, experimentally obtained underwater vibration response is coupled with the elongated body theory to predict the thrust output. Experiments are conducted to validate the electrohydroelastic modeling approach employed in this work and to characterize the performance of an MFC bimorph propulsor. Finally, a wireless battery-powered preliminary robotic fish prototype is developed and tested in free locomotion at different frequency and voltage levels.

Joint attention, a crucial component in interaction and an important milestone in human development, has drawn a lot of attention from the robotics community recently. Robotics researchers have studied and implemented joint attention for robots for the purposes of achieving natural human-robot interaction and facilitating social learning. Most previous work on the realization of joint attention in the robotics community has focused only on responding to joint attention and/or initiating joint attention. Responding to joint attention is the ability to follow another's direction of gaze and gestures in order to share common experience. Initiating joint attention is the ability to manipulate another's attention to a focus of interest in order to share experience. A third important component of joint attention is ensuring, where by the initiator ensures that the responders has changed their attention. However, to the best of our knowledge, there is no work explicitly addressing the ability for a robot to ensure that joint attention is reached by interacting agents. We refer to this ability as ensuring joint attention and recognize its importance in human-robot interaction. We propose a computational model of joint attention consisting of three parts: responding to joint attention, initiating joint attention, and ensuring joint attention. This modular decomposition is supported by psychological findings and matches the developmental timeline of humans. Infants start with the skill of following a caregiver's gaze, and then they exhibit imperative and declarative pointing gestures to get a caregiver's attention. Importantly, as they aged and social skills matured, initiating actions often come with an ensuring behavior that is to look back and forth between the caregiver and the referred object to see if the caregiver is paying attention to the referential object. We conducted two experiments to investigate joint attention in human-robot interaction. The first experiment explored effects of responding to joint attention. We hypothesize that humans will find that robots responding to joint attention are more transparent, more competent, and more socially interactive. Transparency helps people understand a robot's intention, facilitating a better human-robot interaction, and positive perception of a robot improves the human-robot relationship. Our hypotheses were supported by quantitative data, results from questionnaire, and behavioral observations. The second experiment studied the importance of ensuring joint attention. The results confirmed our hypotheses that robots that ensure joint attention yield better performance in interactive human-robot tasks and that ensuring joint attention behaviors are perceived as natural behaviors by humans. The findings suggest that social robots should use ensuring joint attention behaviors.

Industrial robot arms are an essential part of automated manufacturing, and perform tasks such as component assembly, welding, light machining, spray painting, etc. They are highly repeatable, can be calibrated to be sufficiently accurate and they eliminate human error. The serial robot architecture is by far the most ubiquitous in modern day manufacturing, as the technology is highly refined in its current state; the machine architecture provides great dexterity and it has a large useful workspace. This architecture however does have some problems, one of which is a large machine moving mass. The primary reason for this lies in the location of its motors and gearboxes. Due to the robot's significant inertia it utilizes a large amount of energy. This thesis focused on the mechanical design, mathematical modelling and simulation of 2 robotic arm designs which had a hybrid nature. They were classified as hybrid due to the fact that their architectures departed from both the classic definitions of serial kinematics manipulators/machines (SKMs) and parallel kinematics manipulators/machines (PKMs). The primary design goal was to merge some of the advantages of both architectures, i.e. a large workspace to footprint ratio and high end-effector dexterity which was found in serial robots, combined with the low inertia of a parallel robot for improved dynamics. Serial and parallel robots were complementary, and these design goals could not co-exist in a single purist robot architecture. The designs had a full complement of 6 DOFs (degrees of freedom), 3 DOFs for spatial position of the wrist and 3 DOFs for orientation of that wrist. They also had a lower machine moving mass, a fact that was thought to improve speed and energy usage. A major contribution of this research PhD project was a comparative energy usage study, which was performed against the serial robot as a measure. This was done for both hybrid designs as well as another model which represented 2 existing patented designs. The purpose of that was to determine if lowering the machine moving mass would improve energy efficiency, and to determine which design was best. / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2012.

The development of a humanoid robot that mimics human motion requires extensive programming as well as understanding the motion limitations of the robot. Programming the countless possibilities for a robot’s response to observed human motion can be time consuming. To simplify this process, this thesis presents a new approach for mimicking captured human motion data through the development of a composition routine. This routine is built upon a behavior-based framework and is coupled with optimization by calculus to determine the appropriate weightings of predetermined motion behaviors. The completion of this thesis helps to fill a void in human/robot interactions involving mimicry and behavior-based design. Technological advancements in the way computers and robots identify human motion and determine for themselves how to approximate that motion have helped make possible the mimicry of observed human subjects. In fact, many researchers have developed humanoid systems that are capable of mimicking human motion data; however, these systems do not use behavior-based design. This thesis will explain the framework and theory behind our optimal behavior composition algorithm and the selection of sinusoidal motion primitives that make up a behavior library. This algorithm breaks captured motion data into various time intervals, then optimally weights the defined behaviors to best approximate the captured data. Since this routine does not reference previous or following motion sequences, discontinuities may exist between time intervals. To address this issue, the addition of a PI controller to regulate and smooth out the transitions between time intervals will be shown. The effectiveness of using the optimal behavior composition algorithm to create an approximated motion that mimics capture motion data will be demonstrated through an example configuration of hardware and a humanoid robot platform. An example of arm motion mimicry will be presented and includes various image sequences from the mimicry as well as trajectories containing the joint positions for both the human and the robot.

What does it mean to do the disco? Or perform a cheerleading routine? Or move in a style appropriate for a given mode of human interaction? Answering these questions requires an interpretation of what differentiates two distinct movement styles and a method for parsing this difference into quantitative parameters. Furthermore, such an understanding of principles of style has applications in control, robotics, and dance theory. This thesis present a definition for “style of motion” that is rooted in dance theory, a framework for stylistic motion generation that separates basic movement ordering from its precise trajectory, and an inverse optimal control method for extracting these stylistic parameters from real data. On the part of generation, the processes of sequencing and scaling are modulated by the stylistic parameters enumerated: an automation that lists basic primary movements, sets which determine the final structure of the state machine that encodes allowable sequences, and weights in an optimal control problem that generates motions of the desired quality. This generation framework is demonstrated on a humanoid robotic platform for two distinct case studies – disco dancing and cheerleading. In order to extract the parameters that comprise the stylistic definition put forth, two inverse optimal control problems are posed and solved -- one to classify individual movements and one to segment longer movement sequences into smaller motion primitives. The motion of a real human leg (recorded via motion capture) is classified in an example. Thus, the contents of the thesis comprise a tool to produce and understand stylistic motion.

Human stance in quiet mode, relies on feedback from eyes, skin, muscles and the inner ear and the control produced is a combination of strategies which enable a person to stay standing. This thesis presents the simulation and control of a hip actuated robotic model of human standing posture. / The first part of the thesis is devoted to recalling basic elements of the human balance system and to describe the balance strategies it uses to maintain an upright stance. Of the strategies presented, we consider the hip strategy which motivated the formulation of a hip actuated robot. An investigation into the control of nonlinear underactuated robots by linear controllers is done to verify the range and efficiency of the controlled system. / The second part of the thesis includes the investigation of two simplified models of the robot. Results using linear state feedback control are presented. The two models used are compared to clarify the use of one over the other. / We found that for linear controls, the size of the region of convergence decreased underactuated systems of increasing complexity. For our four degrees of freedom robot, the region of convergence is of 2.3 degrees for the actuated joints and of 1 degree for the unactuated joints. Our system is Lyapunov stable when the fully simplified model is assumed.

While the locomotion performance of legged robots over flat terrain or known obstacles has improved over the past few decades, they have yet to equal the performance of their animal counterparts over variable terrain. This work analyzes a multi-legged reduced order model of cockroach locomotion on variable slopes which will be used as an inspiration for a future sprawled posture legged robot. The cockroach is modeled as a point mass, and each leg of the cockroach is modeled as a massless, tangentially rigid, linearly elastic spring attached at the center of mass. All of the springs are actuated to allow changes in energy to the system. This is accomplished by varying the force free length of each leg in a feed-forward manner without reliance on feedback to change the actuation scheme. Fixed points of the model are found using a numerical solver that varies the velocity and phase shift parameters while leaving all other parameters at fixed values selected to match true cockroach motion. Each fixed point is checked for stability and robustness representing how effective the model is at staying on the predetermined gait, and transport cost as a measure of how efficient this gait is. Stable and robust fixed points were successfully found for the range of heading angles encompassing those of representative cockroach motion at each slope. Cockroaches may select the gait used based on stability or efficiency. Thus, additional fixed points were found in combination with a search routine that varies the leg actuation parameters in order to optimize either stability or metabolic efficiency, gaining insights into why cockroaches use the gaits that they do. Optimized fixed points were found based on four different leg functional combination families depending on whether each leg pushes or pulls. Optimized fixed point gaits exist for every incline slope studied between level ground and vertical slopes, at a range of initial heading angles that encompass those typically used by cockroaches. The selected gaits using both a stability based and an efficiency based optimization on the modeled cockroach are very similar. Both are also similar to gaits used by real cockroaches. The forces generated by the model are qualitatively similar to the experimental forces. / Graduation date: 2012

We present a novel control strategy for running which is robust to disturbances, and makes excellent use of passive dynamics for energy economy. The motivation for our control strategy is based on observations of animals, which are able to economically walk and run over varying terrain and ground dynamics. It is well-known that steady-state animal running can be approximated by spring-mass models, but these passive dynamic models describe only steady-state running and are sensitive to disturbances that animals can accommodate. While animals rely on their passive dynamics for energy economy, they also incorporate active control for disturbance rejection. The same approach can be used for spring-mass walking and running, but an active controller is needed that interferes minimally with the passive dynamics of the system. We demonstrate, in simulation, how force control combined with a leg spring stiffness tuned for the desired hopping frequency provides robustness to disturbances on a model for robot hopping, while maintaining the energy economy of a completely passive system during steady-state operation. Our strategy is promising for robotics applications, because there is a clear distinction between the passive dynamic behavior of the model and the active controller, it does not require sensing of the environment, and it is based on a sound theoretical background that is compatible with existing high-level controllers for ideal spring-mass models. / Graduation date: 2012