Autonomous Software Research Articles

A-298 A Novel DNA Array Technology to Detect and Discriminate Complex Panels of Microbial Pathogens

Abstract Background UTI (Urinary Tract Infection) is a prototype for the field of polymicrobial testing. Current testing standards suggest that UTI testing may require the Detection and Discrimination of a panel of at least (@24) bacteria and fungi (@2) along with a panel of antibiotic resistance gene (ARG) markers (>12). Today, such analysis of UTI’s at the DNA level can be obtained with predicates such as large-panel qPCR reactions, e.g. the Thermo OpenArray qPCR assembly, or by NGS based microbiomics. The necessary complexity of such polymicrobial testing pushes the content limits of qPCR and, although NGS has substantial capacity to resolve complex mixtures, NGS remains relatively low throughput and requires a level of data processing that exceeds what is practical in some laboratories. Methods We have developed a new technology for UTI testing, “D3 ArrayTM-UTI”, which obviates most of the present limitations of DNA based polymicrobial testing. The process begins with urine being isolated and purified using standard automated magnetic bead DNA sample prep, already in place in most molecular laboratories, at up to 96 samples in parallel. This is followed by a single multiplex, endpoint PCR reaction on each sample, also at up to 96 samples in parallel, then automated room temperature hybridization of the PCR product in a standard 96 well format, onto a single 21x21 D3 Array at the bottom of each well. Dedicated autonomous software analyzes the hybridization data yielding the complete list of microbial pathogens and ARG’s detected in a urine specimen. The first deployment of the D3 Array-UTI test is now available to support analysis of 96 urine samples in 6hrs. The results of such preliminary testing are presented here. Results aLOD analysis has been performed with purified DNA and found to be in the 0.1-10cp/RXN range among the full set of 24 bacterial species and 12 ARG targets. Concordance analysis for bacterial Species Detection and Discrimination has been performed on 235 clinical isolates obtained from collaborators at Mako Medical laboratories (Thermo Openarray as the comparator) which averaged = 97 +/- 5% for all but two species: A urinae (88%) and M morganii (94%). Concordance analysis for Antibiotic Resistance Gene Detection and Discrimination has been performed on 384 clinical isolates also obtained from Mako Medical laboratories (Thermo Openarray as the comparator) which averaged = 98 +/-5% and was found to be >98% for all but one ARG target: gyrA point mutation (88%). Conclusions The data provided here demonstrates that the D3 Array-UTI platform can perform complex polymicrobial/ARG testing as a high throughput 96 well plate based assay, with Sensitivity and Specificity that matches or exceeds that of qPCR, using standard automation that can be deployed in any molecular laboratory. The resulting performance is unique in that Species and ARG Discrimination are readily scalable to 3-4 dozen Species per test, performed in triplicate in a single well with completely autonomous, cloud enabled data analysis, at a sample-to-answer throughput >96 samples per shift. Beta testing has been initiated among laboratory partners.

Large language models and community legal centres: Could chatbots help reduce Australia’s justice gap?

This article considers the extent to which chatbots based on large language models could reduce unmet demand for services offered by Australian community legal centres. It argues that while chatbots developed to streamline intake processes and assist lawyers with research and document preparation are unlikely to significantly reduce Australia’s justice gap, client-facing legal information chatbots have the potential to do so. However, realising this potential involves solving numerous problems, including uncertainty about how Australia’s Legal Profession Uniform Law applies to autonomous software systems and the risk that hallucinations of large language models will mislead clients.

Pan-Evo: The Evolution of Information and Biology's Part in This.

Many people wonder whether biology, including humans, will benefit or experience harm from new developments in information such as artificial intelligence (AI). Here, it is proposed that biological and non-biological information might be components of a unified process, 'Panevolution' or 'Pan-Evo', based on four basic operations-innovation, transmission, adaptation, and movement. Pan-Evo contains many types of variable objects, from molecules to ecosystems. Biological innovation includes mutations and behavioural changes; non-biological innovation includes naturally occurring physical innovations and innovation in software. Replication is commonplace in and outside biology, including autocatalytic chemicals and autonomous software replication. Adaptation includes biological selection, autocatalytic chemicals, and 'evolutionary programming', which is used in AI. The extension of biological speciation to non-biological information creates a concept called 'Panspeciation'. Panevolution might benefit or harm biology, but the harm might be minimal if AI and humans behave intelligently because humans and the machines in which an AI resides might split into vastly different environments that suit them. That is a possible example of Panspeciation and would be the first speciation event involving humans for thousands of years. This event will not be particularly hostile to humans if humans learn to evaluate information and cooperate better to minimise both human stupidity and artificial simulated stupidity (ASS-a failure of AI).

LegalAsst: Human-centered and AI-empowered machine to enhance court productivity and legal assistance

We propose autonomous software (namely, LegalAsst ) as a step toward an AI-empowered but human-centered machine focused on enhancing court productivity and legal assistance. LegalAsst aims to provide explainable, traceable, and controllable legal assistance and references for lawyers, judges, government officials, and the general public. To achieve this goal, it collates, processes, distills, and visualizes the whole judgment procedure. It streamlines and semi-automates the judgment procedure through case analysis, legislation analysis, and judicial decision-making. Specifically, to make laws and cases easier to navigate and understand, we incorporate structured representations to perform them. Then based on structured representations, we take a step further by introducing a decision-tree-based judgment, making the entire judging process visible and tractable. Our system not only tracks the procedural aspects of judgments but also incorporates modification capabilities, enabling the consideration of the most up-to-date legislation and societal factors to generate more adaptable judgment outcomes.1

VIRTSI: A novel trust dynamics model enhancing Artificial Intelligence collaboration with human users – Insights from a ChatGPT evaluation study

The rapid integration of intelligent processes and methods into information systems in the Artificial Intelligence (AI) era has led to a substantial shift towards autonomous software decision-making. This evolution necessitates robust human oversight, especially in critical domains like Healthcare, Education, and Energy. Human trust in AI plays a vital role in influencing decision-making processes of users interacting with AI.This paper presents VIRTSI (Variability and Impact of Reciprocal Trust States towards Intelligent systems), a novel rigorous computational model for human-AI Interaction. VIRTSI simulates human trust states, spanning from overtrust to distrust, through user modelling. It comprises: 1. A trust dynamics representational model based on Deterministic Finite State Automata (DFAs), illustrating transitions among cognitive trust states in response to AI-generated replies. 2. A trust evaluation model based on Confusion Matrices, originating from machine learning and Accuracy Metrics, providing a quantitative framework for analysing human trust dynamics. As a result, this is the first time that trust dynamics have been thoroughly traced in a representational model and a method has been developed to assess the impact of possibly harmful states like overtrust and distrust.An empirical study on the recently launched Large Language Model of generative AI, ChatGPT (version 3.5), provides a radical underexplored AI-generated platform for evaluating the human-AI interaction through VIRTSI. The study involved 1200 interactions of real users as well as AI experts together with experts in two very different domains of evaluation, namely software engineering and poetry. This study traces trust dynamics and the emerging human-AI interaction, in concrete examples of real user synergies with generative AI. The research reveals the vital role of maintaining normal trust states for optimal human-AI interaction and that both AI and human users need further steps towards this goal. The real-world implications of this research can guide the creation and evaluation of user interfaces with AI and the incorporation of functionalities in the development of generative AI chatbots in terms of trust by providing a new rigorous DFA representational method of trust dynamics and a corresponding new perspective of confusion matrix evaluation method of the dynamics’ impact in the efficiency of human-AI dialogues.

A Novel Industrial Software Model for Social Manufacturing

Abstract Distinct from the conventional manufacturing paradigm, the functioning of the Social Manufacturing (SocialM) model relies on a vast and dispersed array of social manufacturing resources. SocialM communities/groups emerge autonomously through business interactions, resource self-organization, and production information sharing, all driven by product orders within the SocialM system. This necessitates a more open, distributed, and autonomous software model to facilitate the allocation and operation of social manufacturing resources in support of the SocialM model. Consequently, this paper introduces a novel software development model designed specifically for SocialM. This proposed model employs business interaction software as the system's primary control portal, enabling efficient management of resources and users based on the social manufacturing communities/groups. Additionally, it facilitates the administration and oversight of numerous applications by utilizing an industrial Appstore. Moreover, the paper delves into the technological roadmap for realizing the essential technologies of the software model by integrating multi-agent and Message Queuing Telemetry Transport (MQTT) technologies. A 3D printing task simulation case was employed to demonstrate the model's working principle, which served to confirm the reliability and scalability of the SocialM software, as well as the feasibility of the SocialM theory.

Advancements in Reinforcement Learning Algorithms for Autonomous Systems

Reinforcement learning, often known as RL, has developed as a strong paradigm to teach autonomous software agents to make choices in contexts that are both complicated and dynamic. This abstract investigates recent developments and uses of RL in a variety of fields, showing both its transformational potential and the constraints that it faces at present. Recent developments in reinforcement learning (RL) algorithms, in particular deep reinforcement learning (DRL), have made it possible to make major advancements in autonomous decision- making tasks. DRL algorithms can learn complicated representations of state-action spaces by using deep neural networks. This allows for more efficient exploration and exploitation methods to be implemented. Additionally, advancements in algorithmic enhancements, such as prioritized experience replay and distributional reinforcement learning, have improved the stability and sample efficiency of reinforcement learning algorithms, which has made it possible for these algorithms to be used in real-world applications. Robotics, autonomous cars, game playing, finance, and healthcare are just a few of the many fields that may benefit from the use of RL. In the field of robotics, reinforcement learning (RL) makes it possible for autonomous agents to learn how to navigate, manipulate, and move about in environments that are both complicated and unstructured. To improve both safety and efficiency on the road, autonomous cars make use of reinforcement learning (RL) to make decisions in dynamic traffic situations. In finance, RL algorithms are used for portfolio optimization, algorithmic trading, and risk management. These applications serve to improve investment techniques and decision-making procedures. Furthermore, in the field of healthcare, RL supports individualized treatment planning, clinical decision support, and medical image analysis, which enables physicians to provide patients with care that is specifically suited to their needs. Despite the promising improvements and applications, RL is still confronted with several difficulties that restrict its capacity to be widely adopted and scaled. Among these problems are the inefficiency of the sample, the trade-offs between exploration and exploitation, concerns about safety and dependability, and the need for explainability and interpretability in decision-making processes. To effectively address these difficulties, it is necessary to engage in collaborative efforts across several disciplines, conduct research on algorithmic developments, and establish extensive assessment frameworks (Anon, 2022).

Cross-domain flood risk assessment for smart cities using dynamic knowledge graphs

This paper investigates the usage of knowledge graphs to bridge the gap between current data silos in deriving a holistic perspective on the impact of flooding. It builds on the idea of connected digital twins based on the World Avatar dynamic knowledge graph to deploy an ecosystem of autonomous software agents to continuously ingest new real-world information and operate on it. Multiple publicly available yet isolated data sources, including geospatial building information and property sales data as well as real-time river levels, weather observations, and flood warnings, are connected to instantiate a semantically rich ecosystem of knowledge, data, and computational capabilities to provide cross-domain insights in projected flooding events and their potential impact on population and built infrastructure. The extensibility of the proposed approach is highlighted by further integrating power, water, and telecoms infrastructure as part of the very same system, in order to analyse flood-induced asset failures and their propagation across networks. The World Avatar promotes evidence-based decision making during several disaster management phases, supporting both tactical and strategic risk assessments, which supports the United Nations Sustainable Development Goal 11 to improve the assessment of vulnerability, exposure, and risk of communities imposed by flooding events.

Baseline energy modeling for improved measurement and verification through the use of ensemble artificial intelligence models

Accurate estimation of energy savings is crucial for the effective implementation of energy conservation measures (ECMs). Simultaneously, the integration of Artificial Intelligence (AI) has revolutionized software engineering by imbuing software with intelligent capabilities and autonomy. In this paper, we propose an ensemble model for precisely estimating baseline energy consumption within the realm of AI-empowered autonomous software. The ensemble model combines predictions from three tree-based Machine Learning models, namely Random Forest, XGBoost, and LightGBM. Notably, our model emphasizes the provision of explainability, granting transparency and insights into the key factors influencing baseline energy consumption. To validate its effectiveness, we conduct experimental evaluations on a diverse cluster of real-world buildings in Latvia. The results demonstrate the superiority of our proposed ensemble model over individual models and even a deep learning network tailored for energy consumption estimation. These findings underscore the efficacy of AI-empowered models in the energy sector, offering a robust and interpretable solution for estimating energy savings.

Glare analysis of an integral daylighting and lighting control strategy for offices

Complex Fenestration Systems (CFSs) can significantly impact both the visual and non-visual daylight effects on the occupants as well as the energy performance of buildings. To ensure that those impacts improve the overall situation, proper control algorithms are required if the CFS can be operated. The state of the CFS directly and immediately effects the indoor lighting situation. The thermal condition of the building is usually strongly but not directly influenced by the condition of the façade system due to thermal capacities.In this study an experimental integral control strategy, called Integrated Lighting Module (ILM) is investigated in terms of it’s glare prevention capabilities and compared to a cut-off- and a simple rule based control logic. The ILM is an autonomous software package capable to run on building management hardware to control façade systems and artificial lighting in real buildings in an energetically optimised way. The adaptive sampling algorithm raytraverse is employed to calculate Daylight Glare Probability (DGP) values, which are used as a benchmark for the control strategy comparison. The other aspect of this comparison is the opening of the façade, which is evaluated as statistics how often each possible façade state is selected by each control strategy.

A systematic review of hardware technologies for small-scale self-driving cars

Autonomous vehicle (AV) technology has the potential to revolutionize the transportation and logistics industry, making it more efficient and safer. However, testing such technologies is often limited by time, space, and cost constraints. Therefore, in recent years, several initiatives have emerged to test autonomous software and hardware on scaled vehicles. In order to provide guidance for future research, this systematic literature review was conducted to provide an overview of the literature surrounding small-scale self-driving cars, summarizing the current autonomous platforms deployed and focusing on the hardware developments in this field. Through the use of databases such as Web of Science, Scopus, Springer Link, Wiley, ACM Digital Library, and the TRID, 38 eligible studies that present small-scale testing of self-driving cars were identified and reviewed. The results indicated that publications on the topic are relatively new, with only the last four years showing an increase in the number of publications. Additionally, most papers only presented preliminary results, highlighting the potential for further research and development in the field. Research papers predominantly focused on software rather than hardware.

Demonstrating Autonomy for Complex Space Missions: A Europa Lander Mission Autonomy Prototype

There is a desire for robotic spacecraft to perform exploration in unknown, dynamic environments. The Europa Lander Mission Concept is one such mission that needs to deal with an extremely limited lifetime and energy supply, manage intermittent communications with long blackouts, face numerous environmental dangers, and ultimately take place too far from Earth to rely on human control. No missions to date have operated with the required level of autonomy and under the same level of communication constraints, uncertainty, and mission concept complexity as this mission. As a result, the viability of the autonomy must be demonstrated before it will be trusted with mission-critical planning. In this paper, we present an autonomous software prototype that can demonstrate and test the ability of different planners and executives to carry out complex, science-centric missions with limited interventions from humans. The prototype uses a hierarchical utility model that is used to maximize both the amount of expected science return as well as the number of mission objectives imposed by the ground. We demonstrate how this system handles some of the autonomous tasks expected of complex space missions such as decision making, in-situ data acquisition and analysis, data prioritization, resource management, and failure response handling in both simulation and on actual hardware. Through several scenario-based experiments we show how different planners and executives can meet the challenges of the Europa Lander Mission Concept. We also demonstrate that this system can be used in concert with a hardware prototype for autonomy field tests.

Simulating of human physiological supersystems: interactions of cardiovascular, thermoregulatory and respiratory systems

Special quantitative model of the human thermoregulatory system (MTS) functioning with cardiovascular and lung systems is created. These systems form a human physiological supersystem (HPSS). MTS describes thermoregulatory responses to alterations of both external environmental physical characteristics and internal biological characteristics. Algorithms provide designing of scenarios including simulation of either short-time or long-time (hours or days) observations. Input data include different combinations of environmental variables (air or water temperature, air humidity, wind or water flow speed, light intensity, infrared radiation) for a naked or wear human, as well as for given dynamics of biological characteristics (rate of heat production including its components associated with metabolism and ATP molecules leasing during mental and physical activities). Human body is presented by a core, blood, and a skin compartments. Skin and lung evaporation are under hypothalamic control based on afferent impulse patterns from internal, and skin heat and cold receptors. Dynamic output data include blood, hypothalamic, and skin temperatures, hemodynamic parameters like heart rate, cardiac output, regional blood flows, vascular resistances, blood pressures, and regional blood volumes. Serotonin and melatonin concentrations modulating biological heat production rate are associated with a day/night light intensity. Currently, the PCbased simulator is autonomous software to be used both for educational purposes and for providing of special computer research. In a near future, this simulator has to be widened by models of kidneys, and a mechanism of liverpancreas interaction.

A context-aware multi-agent reasoning based intelligent assistive formalism

In recent years, the rapid proliferation of intelligent smart resource-bounded devices and autonomous software technologies has significantly influenced human life and made human life much smarter, more efficient, flexible, comfortable, and secure but device dependent. These intelligent devices are equipped with the feature of context awareness that often run in a highly decentralized environment with minimal and/or without human intervention. As these systems are often distributed in nature, they exhibit complex adaptive behavior by exchanging contextualized information among different resource-bounded devices and/or systems. However, due to the limitation of resources (computational memory), there are still many challenges for a real-time deployment. In this paper, we present a resource-bounded multi-agent reasoning-based context-aware model that assists users intelligently to take the right action autonomously whenever and wherever needed. The core emphasis is given to the lightweight, efficient reasoning mechanism to manipulate rapidly triggering contextual information with the least utilization of computational resources. We develop algorithms along with the complexity factors to evaluate the system’s efficacy in computational time and memory utilization. To suitably formalize the specification and model the proposed formalism using UPPAAL model checker, we develop the case study of a context-aware intelligent assistant along with its rule-based reasoning strategies and priorities for the execution of prioritized tasks according to the customized needs of the user and verify the correctness properties.

Overview of Software Agent Platforms Available in 2023

Agent-based computing remains an active field of research with the goal of building (semi-)autonomous software for dynamic ecosystems. Today, this task should be realized using dedicated, specialized frameworks. Over almost 40 years, multiple agent platforms have been developed. While many of them have been “abandoned”, others remain active, and new ones are constantly being released. This contribution presents a historical perspective on the domain and an up-to-date review of the existing agent platforms. It aims to serve as a reference point for anyone interested in developing agent systems. Therefore, the main characteristics of the included agent platforms are summarized, and selected links to projects where they have been used are provided. Furthermore, the described platforms are divided into general-purpose platforms and those targeting specific application domains. The focus of the contribution is on platforms that can be judged as being under active development. Information about “historical platforms” and platforms with an unclear status is included in a dedicated website accompanying this work.

A stochastic process of software fault detection and correction for business operations

Automatic software fault detection and repair is made possible by autonomic software recovery. By incorporating this function, the software will run more efficiently and aggressively while requiring much less time and resources for maintenance. The focus of this article is on a suggested automated approach to Software Fault Detection and Recovery. The Software Fault Detection and Recovery (SFDR) procedure identifies when a software component has been damaged due to a fault and then restores the damaged component so that the software can continue functioning normally. During the design process, the SFDR is examined and created independently from the intended program. The proposed technique was implemented into an application that demonstrates the SFDR's performance and effectiveness to guarantee its practicality in real-world scenarios. The results of this experiment were encouraging. Results from experiments and comparisons to prior works show that the proposed methodology is successful.

Angiogenesis and activated fibroblast in a model of premature ventricular contraction-induced cardiomyopathy.

Premature ventricular contractions (PVCs) are the most common ventricular arrhythmia in the elderly, and when frequent, can result in PVC-induced cardiomyopathy (PVC-CM). PVC-CM canine model showed eccentric hypertrophy and fibrosis in the left ventricle (LV). The underlying structural and molecular mechanism behind this LV remodeling is unknown. In situ structural adaptations of the LV tissue and the activation of the VEGF/ERK1/2 signaling pathway were studied in the PVC-CM model. Pacemakers were implanted in healthy canines where PVCs were enabled only in the PVC-CM group (50% of PVC burden, 200-220-ms coupling interval) for 12 weeks. Fixed LV longitudinal- and cross-sections were stained with anti-α-SMA antibody, a marker of activated fibroblast (myofibroblasts) and smooth muscle cells (blood vessels), and WGA-AF633 (a marker of cell membranes). Confocal images were analyzed using artificial intelligence (AI)-powered autonomous software, AIVIA. Western blot (WB) analysis was used to study the pro-angiogenic signaling pathway and proteins associated with fibroblast activation. In situ analysis showed a significant increase in the total area and number of blood vessels, along with an increase in activated fibroblasts in PVC-CM LV sections. Also, the PVC group showed a 30% increase in the number of activated fibroblasts per cardiomyocyte compared with the sham group. The endothelial marker, eNOS, and IL-1β, a cytokine involved in fibroblast activation, were elevated in the PVC-CM group. Additionally, the VEGF-A, VEGF-B, and VEGF receptor 2 were upregulated in PVC-CM vs Sham; downstream ERK1/2 was hyperphosphorylated and overexpressed, and MMP9 (marker of fibroblast activation linked to this pathway) was elevated in PVC-CM vs Sham. In conclusion, activated fibroblasts and angiogenesis are a feature of the cardiac LV hypertrophic remodeling in PVC-CM.

Genesis-DB: a database for autonomous laboratory systems.

Artificial intelligence (AI)-driven laboratory automation-combining robotic labware and autonomous software agents-is a powerful trend in modern biology. We developed Genesis-DB, a database system designed to support AI-driven autonomous laboratories by providing software agents access to large quantities of structured domain information. In addition, we present a new ontology for modeling data and metadata from autonomously performed yeast microchemostat cultivations in the framework of the Genesis robot scientist system. We show an example of how Genesis-DB enables the research life cycle by modeling yeast gene regulation, guiding future hypotheses generation and design of experiments. Genesis-DB supports AI-driven discovery through automated reasoning and its design is portable, generic, and easily extensible to other AI-driven molecular biology laboratory data and beyond. Genesis-DB code and installation instructions are available at the GitHub repository https://github.com/TW-Genesis/genesis-database-system.git. The database use case demo code and data are also available through GitHub (https://github.com/TW-Genesis/genesis-database-demo.git). The ontology can be downloaded here: https://github.com/TW-Genesis/genesis-ontology/releases/download/v0.0.23/genesis.owl. The ontology term descriptions (including mappings to existing ontologies) and maintenance standard operating procedures can be found at: https://github.com/TW-Genesis/genesis-ontology.

Distributed real-time production control for resilient manufacturing systems

To enable resilient manufacturing, production control systems must be able to adapt quickly and flexibly to unexpected changes. Motivated by Industrie 4.0, multi-agent systems offer a promising approach to meet these requirements. This paper presents a method for distributed production control that enables production entities to make optimal decisions regarding capacity utilization and the compliance of delivery dates in real-time despite disruptive events. Based on a multi-agent Markov Decision Process, production entities are modeled as autonomous software agents that can solve Flexible Job-Shop Problems through message passing. The performance of the proposed method is evaluated in a simulated production environment.

RLReC: Towards Reinforcement Learning-based Dynamic Design of Reconfiguration Control

The design of Reconfigurable Control Systems (RCS) is a tremendous task since that their increasing need for reconfiguration imposes the anticipation of all potential runtime changes at design time, which is in most cases infeasible. To deal with the high dynamicity of RCS at design-time we propose a novel design approach, RLReC (Reinforcement Learning-based Reconfiguration Control) allowing for the improvement of reconfiguration knowledge through online (runtime) exploration. This approach combines the benefits of meta-modelling with Reinforcement Learning (RL) principals for reconfiguration knowledge modelling in order to handle early and effective exploration of various design choices. In this work, the reconfiguration controller is an autonomous software agent designed as a RL agent (RLRA) that handles both offline (exploitation) and online (exploration) learning through interactions with the controlled system at runtime. In particular, the RLRA has to select reconfiguration rules that were not selected before, which is known as exploration. The main aim of this proposal is to cope with design-time unplanned changes by giving the RLRA the ability of improving its knowledge through online learning feedback. We illustrate the use of our contributions with a running example from the manufacturing domain.