Design Of Systems Research Articles

CYBER-PHYSICAL SYSTEM FOR REMOTE MONITORING AND CONTROL OF A DRINKING WATER NETWORK

The fundamental aim of this paper is to investigate and explore the design and simulation methods for a cyber-physical system (CPS) for the monitoring and control of drinking water capture, treatment and distribution processes. This research aims to provide a deep understanding of the theoretical concepts underlying CPS systems, as well as a detailed analysis of the technologies and methodologies used in the design and simulation of these complex systems. Through this investigation, we aim to identify effective solutions for the management of drinking water resources, given the critical importance of water in modern society. By analyzing case studies and current practices, we will assess the advantages and challenges associated with implementing a CPS system in the real-world context of water supply infrastructure. This work not only aims to provide a theoretical perspective on the concept of CPS applied in the field of drinking water supply, but also to contribute to the development and implementation of practical solutions in order to ensure universal access to quality water and to protect the environment. Through academic approach and rigorous research, we hope to make a contribution to the field of mechatronic engineering and environmental technologies.

A MODEL FOR FORECASTING DEW POINT PRESSURE IN NIGER DELTA CONDENSATE RESERVOIRS USING CONSTANT VOLUME DEPLETION TESTS DATA

The dew point pressure plays a critical role in developing gas condensate reservoirs. It is a crucial factor used in fluid characterisation, performance estimation of gas reservoirs, and design of production systems. However, the composition of gas condensate fluid varies from one location to another, and thus, empirical correlations have been developed to determine the dew point pressure without conducting routine tests. As a result, correlations that are solely useful in the region where they were developed and analysed are developed. This work developed an empirical correlation using data from gas condensate wells in the Niger Delta using multiple linear regression. The model was developed using the Analytical Tools and techniques in Microsoft Excel. To develop the model, we utilised 63 data sets from the Constant Volume Depletion (CVD) experiment, which involved gas condensates from resources in the Niger Delta. The model comprises an empirical correlation that estimates the dew point pressure for gas condensate reservoirs. It uses compositional information of the fluid and reservoir temperature as input parameters. The correlation is derived through multiple regression analyses. Comparing the prediction accuracy of formulas based on the developed model and other conventional methods indicated that this model is more accurate than other statistical methods in predicting DPP within the Niger Delta region. This model can produce more accurate results than other estimation methods when working under limited field information and time constraints. The correlation developed in this process has a coefficient of determination of R2=0.869 and an Average Absolute Relative Error (AARE) of -2.0767%, along with a root mean square of 195279, indicating a high level of reliability in the proposed correlation.

'Accelerated' Deactivation of Carbon Nitride Photocatalyst for Solar Hydrogen Evolution.

Carbon nitride photocatalysts are among the most studied candidates for efficient solar hydrogen (H2) production due to their abundance of precursors, suitable bandgap, and visible light utilization. However, the polymeric nature of carbon nitride materials raises concerns regarding the self-decomposition during photocatalytic redox processes. Yet, the operational stability of carbon nitride photocatalysts for solar H2 production remains under-explored. Here we evaluate the photostability of carbon nitride photocatalysts with platinum (Pt) as the co-catalyst for solar H2 evolution and significant deactivation of this photocatalyst is observed under'accelerated' testing conditions. It is demonstrated that the detachment of the Pt co-catalyst on the surface of carbon nitride is the major reason for this deactivation, which can be attributed to a synergistic effect of photo-corrosion and mechanical stirring. The photo-corrosion weakens the interfacial bonding between carbon nitride and Pt co-catalyst, while continuous collisions from the mechanical stirring promote the detachment of co-catalysts from the surface of carbon nitride. These understandings provide insights into the rational design of photocatalysts and photocatalytic systems for improved operational stability.

Comparison of Machine Learning Approaches for Prediction of the Equivalent Alkane Carbon Number for Microemulsions Based on Molecular Properties.

The chemical properties of oils are vital in the design of microemulsion systems. The hydrophilic-lipophilic difference equation used to predict microemulsions' phase behavior expresses the oils' physiochemical properties as the equivalent alkane carbon number (EACN). The experimental determination of EACN requires knowledge of the temperature dependence of the microemulsion system and the effects of different surfactant concentrations. Thus, the experimental determination is time-intensive and tedious, requiring days to months for proper separations. Furthermore, the experiments require high purity of chemicals because microemulsions are sensitive to impurities. Our work focuses on the quick and reliable predictions of the EACN with machine learning (ML) models. Due to the immaturity of ML chemical predictions, we compare three graph neural networks (GNNs) and a gradient-boosted tree algorithm, known as XGBoost. The GNNs use the molecular structures represented as simplified molecular-input line-entry system (SMILES) codes for the initial input, which allows us to assess whether geometry optimization is necessary for reliable results. The XGBoost model also begins with the SMILES representations of the molecules but uses molecular descriptors instead of geometry optimizations. The best model tested (crystal graph convolutional neural network with Merck molecular force field-94) has an error of 1.15 EACN units of the true EACN for unknown data with the errors skewed toward zero and an R2 score of 0.9.

Repurposing coal plants into thermal energy storage—a techno-economic assessment in the Indian context

Abstract Substantial coal phase out initiatives have been growing as the world mobilizes to meet the Paris climate goals. However, the stranded asset risk associated with this critical transition could fall disproportionately on Asian economies with younger coal fleets, like India. Here, we undertake plant-level techno-economic analysis to explore the value of installing commercially available, molten-salt thermal energy storage (TES) systems for repurposing existing coal power plants in the Indian context. We combine process simulation and an economic optimization model to evaluate design and operations of TES systems for a variety of technology assumptions, coal plant archetypes, and electricity price scenarios. Key drivers of economic viability identified include longer remaining plant lifetime, increasing peak TES temperature, lower TES energy capacity cost, co-production of waste heat for end-uses, and increasing temporal variability of electricity prices. The plant-level analysis was then extended to screen for the potential of TES retrofits within the coal power fleet in Uttar Pradesh, the most populous Indian state with a significant share of India’s coal capacity. Analysis for a single electricity price scenario indicates that over 82% of the coal units in the state can be retrofitted and recover the installed costs of TES retrofits, provided that fixed operating and maintenance costs are excluded. These results reinforce the opportunity for decision-makers to consider TES retrofits of coal plants into cost-effective grid decarbonization strategies.

Robustness Improved Method for Deadbeat Predictive Current Control of PMLSM with Segmented Stators

Permanent magnet linear synchronous motors (PMLSMs) with stator segmented structures are widely used in the design of high-power propulsion systems. However, due to the inherent delay and segmented structure of the systems, there are parameter disturbances in the inductance and flux linkage of the motors. This makes the deadbeat predictive current control (DPCC) algorithm for a current loop less robust in the control system, leading to a decrease in control performance. Compensation methods such as compensation by observer and online estimation of parameters, are problematic to apply in practice due to the difficulty of parameter adjustment and the high complexity of the algorithm. In this paper, a robustness-improved incremental DPCC (RII-DPCC) method—which uses incremental DPCC (I-DPCC) to eliminate flux linkage parameters—is proposed. The stability of the current loop was evaluated through zero-pole analysis of the discrete transfer function. Current feedforward was introduced to improve the stability of I-DPCC. The inductance stability range of I-DPCC was increased from 0.8–1.25 times to 0–2 times the actual value, and the theoretical stability range was increased more than 4 times, effectively improving the robustness of the predictive model to flux linkage and inductance parameters. Finally, the effectiveness of the proposed method was verified through numerical simulation and experiment.

Dual photoelectrode-drived Fe–Br rechargeable flow battery for solar energy conversion and storage: A cost-effective approach

The integrated design of solar energy conversion and storage systems has attracted increasing attention, and non-spontaneous redox reactions driven by dual photoelectrodes provide a potential solution to this issue. This study presents a solar rechargeable flow battery (SRFB) that combines dual photoelectrodes (BiVO4 or Mo–BiVO4 as photoanode, polyterthiophene (pTTh) as photocathode) with cost-effective redox pairs (Fe3+/Fe2+ and Br3−/Br−). The system charges under simulated solar illumination (100 mW∙cm−2, AM 1.5G) and releases stored energy controllably as electrical energy. Research indicates the Mo–BiVO4 photoanode and pTTh photocathode achieve an open-circuit voltage of 0.34 V and a short-circuit current density of 0.38 mA cm−2. Using a specially designed Fe3+/Fe2+-Br3−/Br− (Fe–Br) SRFB device made via 3D printing, a charging photocurrent of approximately 1.9 mA cm−2 is attained. In constant current discharge tests, an initial discharge voltage of 0.23 V is observed at 0.1 mA∙cm−2 within the 10 discharge cycles, demonstrating system stability and offering a viable solution for low-cost, large-scale solar energy storage and conversion.

Design of Cr-Based Molecular Electrocatalyst Systems for the CO2 Reduction Reaction.

ConspectusHuman influence on the climate system was recently summarized by the sixth Intergovernmental Panel on Climate Change (IPCC) Assessment Report, which noted that global surface temperatures have increased more rapidly in the last 50 years than in any other 50-year period in the last 2000 years. Elevated global surface temperatures have had detrimental impacts, including more frequent and intense extreme weather patterns like flooding, wildfires, and droughts. In order to limit greenhouse gas emissions, various climate change policies, like emissions trading schemes and carbon taxes, have been implemented in many countries. The most prevalent anthropogenic greenhouse gas emitted is carbon dioxide (CO2), which accounted for 80% of all U.S. greenhouse gas emissions in 2022. The reduction of CO2 through the use of homogeneous electrocatalysts generally follows a two-electron/two-proton pathway to produce either carbon monoxide (CO) with water (H2O) as a coproduct or formic acid (HCOOH). These reduced carbon species are relevant to industrial applications: the Fischer-Tropsch process uses CO and H2 to produce fuels and commodity chemicals, while HCOOH is an energy dense carrier for fuel cells and useful synthetic reagent. Electrochemically reducing CO2 to value-added products is a potential way to address its steadily increasing atmospheric concentrations while supplanting the use of nonrenewable petrochemical reserves through the generation of new carbon-based resources. The selective electrochemical reduction of CO2 (CO2RR) by homogeneous catalyst systems was initially achieved with late (and sometimes costly) transition metal active sites, leading the field to conclude that transition metal complexes based on metals earlier in the periodic table, like chromium (Cr), were nonprivileged for the CO2RR. However, metals early in the table have sufficient reducing power to mediate the CO2RR and therefore could be selective in the correct coordination environment. This Account describes our efforts to develop and optimize novel Cr-based CO2RR catalyst systems through redox-active ligand modification strategies and the use of redox mediators (RMs). RMs are redox-active molecules which can participate cocatalytically during an electrochemical reaction, transferring electrons─often accompanied by protons─to a catalytic active site. Through mechanistic and computational work, we have found that ligand-based redox activity is key to controlling the intrinsic selectivity of these Cr compounds for CO2 activation. Ligand-based redox activity is also essential for developing cocatalytic systems, since it enables through-space interactions with reduced RMs containing redox-active planar aromatic groups, allowing charge transfer to occur within the catalyst assembly. Following a summary of our work, we offer a perspective on the possibilities for future development of catalytic and cocatalytic systems with early transition metals for small molecule activation.

A Review: Development of a Synthetic Lactoferrin Biological System.

Lactoferrin is an iron-binding glycoprotein with antibacterial, antitumor, and immunomodulatory functions derived from milk and mucosal secretions. Lactoferrin is used in various products, such as infant formula milk powder, nutritional supplements, and cosmetics. Researchers have developed new technologies to produce lactoferrin because there are limitations in the separation and purification of lactoferrin from milk that cannot compensate for the market demand. Therefore, synthetic systems of lactoferrin have been developed with the development of genetic engineering, and the structure of lactoferrin expressed in heterologous systems is very similar to that of natural lactoferrin. The structure and functions of lactoferrin and the design and construction of synthetic lactoferrin biological systems, especially microbial synthetic systems, including prokaryotic and eukaryotic host-expression systems, are described. On the basis of these results, we summarize the challenges and solutions for constructing systems of high-yield lactoferrin. The development directions of recombinant lactoferrin are discussed in this review. Overall, the design and development of these synthetic biological systems have allowed us to explore the great potential of the industrial large-scale preparation of lactoferrin.

QGIS‐based support tools to simplify the complex design challenges of subsurface drainage systems

AbstractWe developed a comprehensive set of tools to simplify the design and layout of subsurface drainage systems. These tools, which utilize publicly available LiDAR (light detection and ranging) datasets, may also be used for other applications. The tool streamline operations include harmonizing geographic information system (GIS) layers into a single State Plane or Universal Transverse Mercator Coordinate System, clipping and thinning LiDAR data within a defined boundary, generating drainage networks with contour lines and laylines for surface flow visualization, identifying depressions and creating gridlines with user‐specified angles and spacings. To showcase their use, an illustrative example using a diverse dataset from the University of Illinois’ South Farm is provided. These user‐friendly tools, optimized for compatibility with all versions of QGIS3 (quantum geographic information system 3), are freely accessible for download or installation from the official QGIS plugin repository. A user manual with step‐by‐step instructions is available online on the Illinois Drainage Guide website.

Design Insights for Industrial CO2 Capture, Transport, and Storage Systems.

We present methods and insights for the design of CO2 capture, transport, and storage systems for industrial facilities with a case study focus on Louisiana. Our analytical framework includes (1) evaluating the scale and concentration of capturable CO2 emissions at individual facilities for the purpose of estimating the cost of CO2 capture retrofits that utilize various energy supply sources to meet parasitic demands; (2) screening to identify potential CO2 storage sites and estimate their capacities, injectivities, and costs; and (3) designing cost-minimized trucking or pipeline infrastructure connecting CO2 capture plants with storage sites, considering existing land uses, demographics, and a variety of social and environmental justice factors. Estimated levelized costs of capture at Louisiana's 190 industrial facilities range from below $50/tCO2 to above $500/tCO2, depending on facility-specific features. We identified 98 potential storage sites with storage costs ranging from $8 to $17/tCO2. We find that in most situations, pipelines are the least-costly mode of CO2 transport. When industrial facilities in a region share pipelines, aggregate pipeline mileage and average transport costs are dramatically lower than without sharing. Shared pipeline networks designed to avoid disadvantaged communities require right-of-way areas compared to those for networks that transect such communities, but result in 25% higher average per-tonne transport cost.

Multi-paradigm modelling and control of microgrid systems for better power stability in the Rockaways

The Rockaways Peninsula faces issues related to congestion and power outages during times of peak usage. Additionally, it is susceptible to disruptions caused by disasters such as hurricanes and storms. In this paper, we propose a new methodology that employs multi-paradigm modelling and control for the design and implementation of interconnected microgrid systems in the Rockaways. Microgrids are small-scale power networks that incorporate renewable energy technologies for power generation and distribution to enhance the control of energy supply and demand. Multi-paradigm modelling is employed to describe microgrids’ dynamic behavior more accurately by integrating system dynamics, agent-based modelling, as well as discrete event and continuous time simulation. We use agent-based models to describe the behavior of separate microgrid elements and the microgrid as a whole. Discrete event/continuous time simulation is used to analyze real-time operation of electrical parameters, such as voltage, current and frequency. Thus, the design, analysis and performance of microgrids are improved. Also, control strategies are used for the purpose of enabling the microgrids to operate effectively by responding to changes in power supply and demand and minimizing the effects of disturbances. The findings of this study demonstrate the feasibility and resilience benefits of incorporating multi-paradigm modelling and control in the design and management of microgrid systems in the Rockaways, which can result in the development of more durable, efficient, and sustainable energy systems in the region.

Changes in holding time depending liquid filling ratio, heat transfer parameters, and shape of the tank

This study simulates the self-pressurization of liquefied hydrogen using a three-element lumped model. We analyze the effects of liquid filling ratio, heat flux distribution, interface heat transfer parameters, and tank shape on pressure build-up, which can vary due to ship transportation conditions. Key findings are: lower liquid levels correlate with faster pressure build-up; the sensitivity of pressure build-up to changes in the heat flux ratio between vapor and liquid is higher during laden voyages compared to ballast voyages; The effect of heat absorption by vapor on pressure build-up is greater than that of BOG caused by heat transfer; and horizontal tanks exhibit the longest holding time. These findings enhance the understanding and prediction of pressure dynamics in liquid hydrogen transportation, aiding the design and optimization of systems for safe and efficient operation.

Unsteady nonlinear lifting line model for active gust load alleviation of airplanes

AbstractActive gust load alleviation is an important technology for designing future passenger airplanes to be lighter and thus more environmentally friendly. Unsteady Reynolds-averaged Navier–Stokes (URANS) simulations are typically used to accurately calculate gust loads, but because of their high computational cost, they can only be performed at a few selected operating points. In simpler potential theory models, stall is neglected, resulting in loss of accuracy. In this paper, a low-order unsteady aerodynamics wing model is presented, which is able to represent well compressible flow with stall. Furthermore, the model offers the possibility to modularly incorporate actuators, which allows the design and evaluation of active load alleviation systems. The model is based on a conventional unsteady 2D airfoil model including a dynamic stall model. The dynamic stall model requires viscous steady coefficients, e.g. from 2D steady RANS computations. This 2D airfoil model is coupled with a 3D steady-state lifting line model. The model is applied to the LEISA research airplane and extensively validated with URANS results. It performs well in calculating gust loads with and without simultaneous flap deflections, and provides significantly more accurate results in the case of stall than when stall is neglected.

Observed Impacts of Ground-Mounted Photovoltaic Systems on the Microclimate and Soil in an Arid Area of Gansu, China

Ground-mounted photovoltaic (GMPV) systems are a crucial component of photovoltaic (PV) applications, and their environmental impacts during large-scale development require thorough attention. This study conducted continuous observations at a GMPV plant in an arid region, employing a three-site comparative monitoring system to assess the environmental impact of both shaded and non-shaded areas within GMPV systems. The parameters measured included atmospheric temperature (AT), relative humidity (RH), soil temperature (ST), soil water content (SWC), and wind speed. The results revealed significant diurnal and seasonal variations in AT, with daytime warming and nighttime cooling ranging from 0.1 to 0.7 °C, with particularly large variations observed during high-temperature seasons. Shaded areas under the PV panels exhibited increased RH at night and decreased RH during the day, along with a cooling effect on ST, with a maximum reduction of 7 °C. SWC was higher in shaded areas during dry seasons but exhibited complex redistribution patterns during rainy seasons. Wind speed and direction were notably altered, demonstrating a corridor effect. These findings contrast with previous studies that only focused on the environmental assessment of non-shaded areas within PV systems and external areas using two-site monitoring. This study highlights the critical role of shaded areas in understanding the local environmental impacts of PV systems. This comprehensive approach offers deeper insights into how PV systems influence local meteorological and environmental conditions, suggesting that optimized design and placement of PV systems can enhance their ecological benefits and mitigate adverse environmental impacts in arid regions.

CHARM 2.0: Composing Heterogeneous Accelerators for Deep Learning on Versal ACAP Architecture

Dense matrix multiply (MM) serves as one of the most heavily used kernels in deep learning applications. To cope with the high computation demands of these applications, heterogeneous architectures featuring both FPGA and dedicated ASIC accelerators have emerged as promising platforms. For example, the AMD/Xilinx Versal ACAP architecture combines general-purpose CPU cores and programmable logic with AI Engine processors optimized for AI/ML. An array of 400 AI Engine processors executing at 1 GHz can provide up to 6.4 TFLOPS performance for 32-bit floating-point (FP32) data. However, machine learning models often contain both large and small MM operations. While large MM operations can be parallelized efficiently across many cores, small MM operations typically cannot. We observe that executing some small MM layers from the BERT natural language processing model on a large, monolithic MM accelerator in Versal ACAP achieved less than 5% of the theoretical peak performance. Therefore, one key question arises: How can we design accelerators to fully use the abundant computation resources under limited communication bandwidth for end-to-end applications with multiple MM layers of diverse sizes? We identify the biggest system throughput bottleneck resulting from the mismatch between massive computation resources of one monolithic accelerator and the various MM layers of small sizes in the application. To resolve this problem, we propose the CHARM framework to compose multiple diverse MM accelerator architectures working concurrently on different layers within one application. CHARM includes analytical models which guide design space exploration to determine accelerator partitions and layer scheduling. To facilitate system designs, CHARM automatically generates code, enabling thorough onboard design verification. We deploy the CHARM framework on four different deep learning applications in FP32, INT16, and INT8 data types, including BERT, ViT, NCF, and MLP, on the AMD/Xilinx Versal ACAP VCK190 evaluation board. Our experiments show that we achieve 1.46 TFLOPS, 1.61 TFLOPS, 1.74 TFLOPS, and 2.94 TFLOPS inference throughput for BERT, ViT, NCF, and MLP in FP32 data type, respectively, which obtain 5.29 \(\times\) , 32.51 \(\times\) , 1.00 \(\times\) , and 1.00 \(\times\) throughput gains compared to one monolithic accelerator. CHARM achieves the maximum throughput of 1.91 TOPS, 1.18 TOPS, 4.06 TOPS, and 5.81 TOPS in the INT16 data type for the four applications. The maximum throughput achieved by CHARM in the INT8 data type is 3.65 TOPS, 1.28 TOPS, 10.19 TOPS, and 21.58 TOPS, respectively. We have open-sourced our tools, including detailed step-by-step guides to reproduce all the results presented in this paper and to enable other users to learn and leverage CHARM framework and tools in their end-to-end systems: https://github.com/arc-research-lab/CHARM .

Component, design, and prospects of self-consistent energy systems for transport infrastructures

ABSTRACT This paper introduces a Transportation Self-Consistent Energy System (TSCES) to meet the increasing electricity requirements of transport infrastructure in various traffic environments, including highway, railway, and waterway. The TSCES consists of four primary components: power source, load, energy storage, and interconnected microgrid. The power source component can be categorized into natural energy endowment and energy-electricity conversion components. Load components can be classified, quantified, and forecasted based on diverse transportation facilities, traffic volume, and importance levels. Energy storage components necessitate thorough consideration of the technical attributes of various energy storage methods, along with the functional requirements of the traffic environment, power sources, and loads. Interconnected microgrid components establish connections among all components, followed by configuration optimization and choosing appropriate network topological structures based on techno-economic conditions, ensuring that all components adhere to constraints imposed by transportation space, regulations, and policy indicators. Additionally, the design procedure and architectural patterns of TSCES are presented, which indicates that the TSCES has excellent adaptability and research application value in the field of transportation.

Experimental study on a solar thermoelectric power generation system under non-uniform irradiation

Non-uniform irradiation caused by high light concentration significantly affects the performance of the solar thermoelectric power generation system, but the research remains at the theoretical stage. This paper establishes an experimental setup for the solar thermoelectric system and conducts a comprehensive experimental study of the system operating under non-uniform irradiation. The temperature, power and efficiency of the systems are tested for different input powers, cooling methods and structures, and irradiation uniformity. The results indicate that the output power of the solar thermoelectric system slightly increases and then decreases as the irradiation uniformity increases. The system’s output power is minimized when the irradiation uniformity approaches 100%. The optimum irradiation uniformity of the solar thermoelectric system is 47.4% and the maximum efficiency is 1.495%, which is an improvement of 7.6% compared to the efficiency of 98.4% irradiation uniformity. The performance of the solar thermoelectric system can be further improved by using a double-layer TE structure. At a concentration ratio of only 12.4, the double-layer system connected in series achieved a maximum efficiency of 2.06%, which is 0.55% higher than the single-layer system. The importance of optimizing irradiation uniformity to improve system performance is experimentally revealed. These results are important for clarifying the operating principle of the solar thermoelectric system under non-uniform irradiation and can guide the subsequent design of high-efficiency solar thermoelectric systems.

Flame transfer functions of asymmetric conical and V-shaped flames

Transfer functions of asymmetric flames subjected to incident flow perturbations are explored using the classical kinematic model, the G-equation, parameterized by flame tip angle α, flame-base incline angle β, and eccentricity e. With perturbations of various frequencies, we analyze perturbed asymmetric flames and derive analytical formulations of Flame Transfer Functions (FTF) for various flame conditions. It is observed that the gain of the asymmetric conical flame is increased compared to the symmetric case. Meanwhile, the amplification effect of the V-shaped flame is considerably suppressed as the flame becomes asymmetric. That means with a proper asymmetric geometry setting, the V-shaped flame no longer tends to amplify the perturbation and in other words becomes less susceptible to combustion instability. These findings have significant value for the future design and optimization of propulsion systems, where improving flame stability and reducing thermoacoustic instabilities are critical. Subsequently, a numerical study is conducted to verify the FTF results and examine the response of asymmetric flames under various perturbation levels. It has been demonstrated that the asymmetric conical flame is minimally affected by the amplitude of these perturbations, and the asymmetric V-shaped flame is highly sensitive to the convective amplitudes.

Towards Trustworthy AI-Enabled Decision Support Systems: Validation of the Multisource AI Scorecard Table (MAST)

The Multisource AI Scorecard Table (MAST) is a checklist tool to inform the design and evaluation of trustworthy AI systems based on the U.S. Intelligence Community’s analytic tradecraft standards. In this study, we investigate whether MAST can be used to differentiate between high and low trustworthy AI-enabled decision support systems (AI-DSSs). Evaluating trust in AI-DSSs poses challenges to researchers and practitioners. These challenges include identifying the components, capabilities, and potential of these systems, many of which are based on the complex deep learning algorithms that drive DSS performance and preclude complete manual inspection. Using MAST, we developed two interactive AI-DSS testbeds. One emulated an identity-verification task in security screening, and another emulated a text-summarization system to aid in an investigative task. Each testbed had one version designed to reach low MAST ratings, and another designed to reach high MAST ratings. We hypothesized that MAST ratings would be positively related to the trust ratings of these systems. A total of 177 subject-matter experts were recruited to interact with and evaluate these systems. Results generally show higher MAST ratings for the high-MAST compared to the low-MAST groups, and that measures of trust perception are highly correlated with the MAST ratings. We conclude that MAST can be a useful tool for designing and evaluating systems that will engender trust perceptions, including for AI-DSS that may be used to support visual screening or text summarization tasks. However, higher MAST ratings may not translate to higher joint performance, and the connection between MAST and appropriate trust or trustworthiness remains an open question.