Search

Artificial social agents are becoming increasingly present in the real world. For these agents, their success is tied to their ability to remain truly autonomous-to appropriately adapt their decisions and behaviours across changing physical and social contexts-and still remains a key area of research in the field of embodied autonomous agents. Numerous approaches towards models of adaptation have been proposed, including biologically-inspired models based on the regulation of an agent's internal environment. A more recent concept from biological systems, called "(social) allostasis", proposes a mechanism for the anticipatory adaptation of the internal environment through (affect-based) interactions with the (social) environment. Despite the long-term adaptive benefits proposed by social allostasis, current approaches towards models of adaptation for artificial social agents are yet to consider its principles in their approaches. In this manuscript, we address this shortcoming and investigate how mechanisms of social allostasis through affect-based relationships can be used for long-term adaptation of an embodied agent model. Using a biologically-inspired artificial life approach, we systematically investigate numerous hypothesised, hormonal mechanisms that underpin social allostasis and the survival-related benefits associated with affect-based social relationships. We conduct these investigations using a small society of simulated agents across numerous dynamic physical and social contexts. Our results find that adaptation of an embodied agent model through social allostasis mechanisms can provide survival-related advantages across several dynamic conditions in some contexts through physiological and behavioural adaptation. Throughout the manuscript, we also find numerous social interactions and dynamics that mirror biological systems. Our work addresses limitations in current approaches to adaptive models for embodied artificial agents, and presents a novel framework to guide future work in this field. The work also contributes a biologically-inspired artificial life model as a scientific tool to address some methodological challenges in natural systems.

Human-robot interaction (HRI) is the process of humans and robots working together to accomplish a goal with the objective of making the interaction beneficial to humans. Closed loop control and adaptability to individuals are some of the important acceptance criteria for human-robot interaction systems. While designing an HRI interaction scheme, it is important to understand the users of the system and evaluate the capabilities of humans and robots. An acceptable HRI solution is expected to be adaptable by detecting and responding to the changes in the environment and its users. Hence, an adaptive robotic interaction will require a better sensing of the human performance parameters. Human performance is influenced by the state of muscular and mental fatigue during active interactions. Researchers in the field of human-robot interaction have been trying to improve the adaptability of the environment according to the physical state of the human participants. Existing human-robot interactions and robot assisted trainings are designed without sufficiently considering the implications of fatigue to the users. Given this, identifying if better outcome can be achieved during a robot-assisted training by adapting to individual muscular status, i.e. with respect to fatigue, is a novel area of research. This has potential applications in scenarios such as rehabilitation robotics. Since robots have the potential to deliver a large number of repetitions, they can be used for training stroke patients to improve their muscular disabilities through repetitive training exercises. The objective of this research is to explore a solution for a longer and less fatiguing robot-assisted interaction, which can adapt based on the muscular state of participants using fatigue indicators derived from electromyogram (EMG) measurements. In the initial part of this research, fatigue indicators from upper limb muscles of healthy participants were identified by analysing the electromyogram signals from the muscles as well as the kinematic data collected by the robot. The tasks were defined to have point-to-point upper limb movements, which involved dynamic muscle contractions, while interacting with the HapticMaster robot. The study revealed quantitatively, which muscles were involved in the exercise and which muscles were more fatigued. The results also indicated the potential of EMG and kinematic parameters to be used as fatigue indicators. A correlation analysis between EMG features and kinematic parameters revealed that the correlation coefficient was impacted by muscle fatigue. As an extension of this study, the EMG collected at the beginning of the task was also used to predict the type of point-to-point movements using a supervised machine learning algorithm based on Support Vector Machines. The results showed that the movement intention could be detected with a reasonably good accuracy within the initial milliseconds of the task. The final part of the research implemented a fatigue-adaptive algorithm based on the identified EMG features. An experiment was conducted with thirty healthy participants to test the effectiveness of this adaptive algorithm. The participants interacted with the HapticMaster robot following a progressive muscle strength training protocol similar to a standard sports science protocol for muscle strengthening. The robotic assistance was altered according to the muscular state of participants, and, thus, offering varying difficulty levels based on the states of fatigue or relaxation, while performing the tasks. The results showed that the fatigue-based robotic adaptation has resulted in a prolonged training interaction, that involved many repetitions of the task. This study showed that using fatigue indicators, it is possible to alter the level of challenge, and thus, increase the interaction time. In summary, the research undertaken during this PhD has successfully enhanced the adaptability of human-robot interaction. Apart from its potential use for muscle strength training in healthy individuals, the work presented in this thesis is applicable in a wide-range of humanmachine interaction research such as rehabilitation robotics. This has a potential application in robot-assisted upper limb rehabilitation training of stroke patients.

This thesis puts forward a novel way of control for robotic morphologies. Taking inspiration from Behaviour Based robotics and self-organisation principles, we present an interfacing mechanism, capable of adapting both to the user and the robot, while enabling a paradigm of intuitive control for the user. A transparent mechanism is presented, allowing for a seamless integration of control signals and robot behaviours. Instead of the user adapting to the interface and control paradigm, the proposed architecture allows the user to shape the control motifs in their way of preference, moving away from the cases where the user has to read and understand operation manuals or has to learn to operate a specific device. The seminal idea behind the work presented is the coupling of intuitive human behaviours with the dynamics of a machine in order to control and direct the machine dynamics. Starting from a tabula rasa basis, the architectures presented are able to identify control patterns (behaviours) for any given robotic morphology and successfully merge them with control signals from the user, regardless of the input device used. We provide a deep insight in the advantages of behaviour coupling, investigating the proposed system in detail, providing evidence for and quantifying emergent properties of the models proposed. The structural components of the interface are presented and assessed both individually and as a whole, as are inherent properties of the architectures. The proposed system is examined and tested both in vitro and in vivo, and is shown to work even in cases of complicated environments, as well as, complicated robotic morphologies. As a whole, this paradigm of control is found to highlight the potential for a change in the paradigm of robotic control, and a new level in the taxonomy of human in the loop systems.

The IoT concept has revolutionised the vision of the future Internet with the advent of standards such as 6LoWPAN making it feasible to extend the Internet into previously isolated environments, e.g., WSNs. The abstraction of resources as services, has opened these environments to a new plethora of potential applications. Moreover, the web service paradigm can be used to provide interoperability by offering a standard interface to interact with these services to enable WoT paradigm. However, these networks pose many challenges, in terms of limited resources, that make the adaptability of existing IP-based solutions infeasible. As traditional service discovery and selection solutions demand heavy communication and use bulky formats, which are unsuitable for these resource-constrained devices incorporating sleep cycles to save energy. Even a registry based approach exhibits burdensome traffic in maintaining the availability status of the devices. The feasible solution for service discovery and selection is instrumental to enable the wide application coverage of these networks in the future. This research project proposes, TRENDY, a new compact and adaptive registry-based SDP with context awareness for the IoT, with more emphasis given to constrained networks, e.g., 6LoWPAN It uses CoAP-based light-weight and RESTful web services to provide standard interoperable interfaces, which can be easily translated from HTTP. TRENDY's service selection mechanism collects and intelligently uses the context information to select appropriate services for user applications based on the available context information of users and services. In addition, TRENDY introduces an adaptive timer algorithm to minimise control overhead for status maintenance, which also reduces energy consumption. Its context-aware grouping technique divides the network at the application layer, by creating location-based groups. This grouping of nodes localises the control overhead and provides the base for service composition, localised aggregation and processing of data. Different grouping roles enable the resource-awareness by offering profiles with varied responsibilities, where high capability devices can implement powerful profiles to share the load of other low capability devices. Thus, it allows the productive usage of network resources. Furthermore, this research project proposes APPUB, an adaptive caching technique, that has the following benefits: it allows service hosts to share their load with the resource directory and also decreases the service invocation delay. The performance of TRENDY and its mechanisms is evaluated using an extensive number of experiments performed using emulated Tmote sky nodes in the COOJA environment. The analysis of the results validates the benefit of performance gain for all techniques. The service selection and APPUB mechanisms improve the service invocation delay considerably that, consequently, reduces the traffic in the network. The timer technique consistently achieved the lowest control overhead, which eventually decreased the energy consumption of the nodes to prolong the network lifetime. Moreover, the low traffic in dense networks decreases the service invocations delay, and makes the solution more scalable. The grouping mechanism localises the traffic, which increases the energy efficiency while improving the scalability. In summary, the experiments demonstrate the benefit of using TRENDY and its techniques in terms of increased energy efficiency and network lifetime, reduced control overhead, better scalability and optimised service invocation time.

This work focuses on the need for new knowledge to allow hierarchical multi-project management to be conducted in the construction industry, which is characterised by high uncertainty, fragmentation, complex decisions, dynamic changes and long-distance communication. A dynamic integrated project management approach is required at strategic, tactical and operational levels in order to achieve adaptability. The work sees the multi-project planning and control problem in the context of supply chain management at main contractor companies. A portfolio manager must select and prioritise the projects, bid and negotiate with a wide range of clients, while project managers are dealing with subcontractors, suppliers, etc whose relationships and collaborations are critical to the optimisation of schedules in which time, cost and safety (etc) criteria must be achieved. Literature review and case studies were used to investigate existing approaches to hierarchical multi-project management, to identify the relationships and interactions between the parties concerned, and to investigate the possibilities for integration. A system framework was developed using a multi-agent-system architecture and utilising procedures adapted from literature to deal with short, medium and long-term planning. The framework is based on in-depth case study and integrates time-cost trade-off for project optimisation with multi-attribute utility theory to facilitate project scheduling, subcontractor selection and bid negotiation at the single project level. In addition, at the enterprise level, key performance indicator rule models are devised to align enterprise supply chain configuration (strategic decision) with bid selection and bid preparation/negotiation (tactical decision) and project supply chain selection (operational decision). Across the hierarchical framework the required quantitative and qualitative methods are integrated for project scheduling, risk assessment and subcontractor evaluation. Thus, experience sharing and knowledge management facilitate project planning across the scattered construction sites. The mathematical aspects were verified using real data from in-depth case study and a test case. The correctness, usefulness and applicability of the framework for users was assessed by creating a prototype Multi Agent System-Decision Support System (MAS-DSS) which was evaluated empirically with four case studies in national, international, large and small companies. The positive feedback from these cases indicates strong acceptance of the framework by experienced practitioners. It provides an original contribution to the literature on planning and supply chain management by integrating a practical solution for the dynamic and uncertain complex multi-project environment of the construction industry.

This thesis develops an approach to improving the representation, form and timeliness of data in a complex process control interface. In traditional interfaces, a mapping is made at design time between the process parameters and an appropriate rendering at the interface. This mapping is usually the best all-purpose mapping under a set of general constraints. It is not, however, the only mapping—others may have been discarded which might have represented a better mapping under a different set of constraints. In the general case of the system functioning under normal conditions, the general mapping may be appropriate in most instances. However, if the process moves into a disturbed state, one of the other discarded mappings may be more appropriate for the new conditions. The goal of this thesis is to investigate if these other mappings can be implemented in a flexible mapping system, so that an adaptive system can make a decision on which mapping to use at run-time, based on the current state of process, the environment, the actions of the operator team, and access to a human factors database.

The COVID-19 pandemic has challenged student success and the ability of colleges to deliver the education students need for the workplace. The purpose of this qualitative dissertation case study was to investigate how the organizational resiliency of a Midwest community college impacted student success during the COVID-19 pandemic by examining the contributions of static and dynamic resiliency characteristics and dimensions. Exploration and analysis of the resilient characteristics and dimensions of student success focus on the key factors of instructional modality, instructor-student engagement, and socioeconomic influences. Conclusions are based on thematic analysis of semi-structured interviews with the college’s senior leaders triangulated with information from public documents and a student survey. The researcher concludes that the college’s static resiliency components of solid planning and infrastructural preparation for probable events, a longstanding collaborative commitment to achieving key elements of student success, and effective internal communication processes produced the robust capacity for flexibility and innovation that distinguishes dynamic resiliency. Innovations that accelerated decision processes, faculty and staff encouraged to experiment based on regular feedback, and energetic action on non-academic stresses had positive effects on student success during the COVID-19 pandemic. One result was that for the fiscal year 2021-2022, the college awarded the highest number of degrees in its history to all students and to African American males, with 99 percent of students rating the quality of their education as good to excellent.

I designed an explanatory sequential mixed-method study to explore the relationship between leadership practices, collective teacher efficacy (CTE), and educational outcomes of low SES students in an open school system. Four data sources were analyzed: K-5 student Developmental Reading Assessment (DRA) results of 1,170 students, Collective Teacher Belief Scale (CTBS) survey to measure CTE perceptions of 48 teachers, focus group interviews composed of a stratified sample of 11 K-5 teachers, and two one-on-one principal interviews. The study took place in two elementary schools in north Texas during the 2020-2021 school year. A Spearman's rank-order correlation analysis indicated that the relationship between CTBS scores and student reading scores was mixed. While one school showed a positive association between CTE and the reading data of low SES students, the other school showed a weak correlation between the variables. The quantitative data indicated that CTBS scores did not independently explain reading achievements at both campuses. The data also showed that while teachers had a large effect size on the reading performance of low SES students, as measured through a Cohen's d for paired sample t-test, achievement gaps continued to widen. Two themes emerged through a grounded theory approach when principals described their sensemaking and framing process: many variables and teamwork. In a complex open system, school principals must consider the many needs of teachers and students before enacting reform efforts. To achieve ambitious goals, school principals promote teamwork and help develop supportive structures to assist teachers and students. Furthermore, the theme of support emerged when teachers described their CTE perceptions concerning their principal's leadership actions. The overall data suggest that leadership practices directly impact CTE levels and indirectly impact student performance. When teachers felt supported, CTE levels increased, but when teachers felt overwhelmed or failed to see how school and district mandates support teaching and learning efforts, CTE levels dropped.

As an increasingly important part of our social, professional and financial lives happens online, the frequency with which we have to deal with privacy problems is ever on the rise. In this dissertation I answer the question: How can we help users to balance the benefits and risks of information disclosure in a user-friendly manner, so that they can make good privacy decisions? After briefly motivating this question in Chapter 1, I first discuss problems with existing answers to this question in Chapters 2 and 3. In Chapter 2, I explain how providing transparency and control does sufficiently not help users in making better privacy decisions. Specifically, I demonstrate that people’s privacy decisions fall prey to all sorts of decision biases, and that most privacy decisions are too complex for people to fathom. In effect, many people refrain from exploiting the provided transparency and control altogether.In Chapter 3, I explain how “privacy nudging” is also not sufficient in its presently studied form. Specifically, I demonstrate that although nudges relieve some of the burden of privacy decision making, they tend to overlook the inherent diversity of users’ privacy preferences and the context-dependency of their decisions.The main argument of this dissertation is that because of these shortcomings of transparency-and-control and privacy nudges, privacy scholars need to move beyond the “one-size-fits-all” approach to privacy embodied in both nudges and transparency and control. I argue that because of the high variability and context-dependency of people’s privacy decisions, nudges need to be tailored to the user and her context.In several studies, I contextualize users’ privacy decisions by showing how disclosure depended on the person’s privacy profile, the type of information, and the recipient of the information (Chapter 4). Then, I present the idea of a “privacy adaptation procedure” and demonstrate its merit in Chapter 5. Finally, I test a complete implementation of the privacy adaptation procedure in Chapter 6. The results of this final study cause reserved optimism regarding the feasibility of user-tailored privacy decision support.

The Quality-based Adaptive Resource Management Architecture (QARMA) consists of a framework for describing resource management solutions and a collection of CORBA-based middleware services for the management of distributed, real-time systems. The framework can be used to: (1) characterize existing resource management architectures and tools and (2) assist in integrating existing tools into coherent resource management solutions. The middleware components are an instantiation of a resource management solution based directly on the framework. The main contributions of this thesis include an analysis of the information life-cycle in a resource management system, a collection of algorithmic models that serve as a basis for resource allocation algorithms in the QARMA middleware, and a description of the greedy algorithm used by the Resource Management Service. Experimental results demonstrate that QARMA can control existing application systems and can be integrated with existing management middleware.

Networks and their applications have been evolving substantially in the last two decades. Network-based control systems, also called Networked Control Systems (NCS), are one of the hottest topics in network-based applications. By networked control system we mean controlling the system through the network, i.e., using networks for remote control and automation. One major concern of using networks in remote control and automation is the Quality of Service (QoS) deterioration effects. These QoS deterioration effects such as: network-induced time delays, bandwidth constraints, multiple-packet transmission, and dropping network packets are serious tribulations. Without considering the effect of the QoS deterioration the performance of the NCS closed-loop control system could be degraded and even destabilized. In this thesis, an adaptive fuzzy logic modulation scheme is proposed to compensate for the network-induced time delays in the network-based PI control systems. This scheme is based on modulating the control signal provided by the PI controller and applied to the remote controller via the network. A network-based controlled DC motor is used in this thesis to illustrate the effectiveness of the proposed scheme.

The application of this thesis compares three different manufacturing theories, varying on resource allocation, to decide which performs best. Three Resource Allocation methods are analyzed including a static resource allocation method, a semidynamic resource allocation method, and a dynamic resource allocation method. The last methodology used in this thesis develops a resource allocation method intended to display complex behavior. A Base Case manufacturing simulation is developed to represent a static resource allocation system. A Theory of Constraints manufacturing model using the Theory of Constraints process represents a semi-dynamic resource allocation system. A Complex Adaptive System manufacturing model using autonomous agents represents a dynamic resource allocation system. The systems were analyzed at multiple stress levels for system to system performance, as well as for complex behavior. Dynamic resource allocation outperformed the semi-dynamic allocation and static allocation systems unilaterally. Some complex behavior was displayed for elements of some models, but as a whole, a trend of increasing complexity did not emerge as the models were analyzed from the Base Case, to Theory of Constraints, to the Complex Adaptive System.