Operational Demands Research Articles

Design, Configuration Synthesis, and Experimental Study of Side-Rolling Metamorphic Mechanism for Metal Additive Manufacturing

Arc additive manufacturing (AAM) has the advantages of fast deposition speed and good surfacing quality. It is a promising additive manufacturing (AM) method. However, arc additive manufacturing is difficult to use widely in industry due to its poor deformation, microstructure, and mechanical properties. Since the mechanical properties of materials can be greatly improved by rolling, a method for configuration synthesis of the side-rolling mechanism by using metamorphic mechanism theory is presented in this paper. Firstly, by analyzing the operational demands of the side-rolling mechanism, we obtained the motion cycle diagram for the metamorphic mechanism in addition to the corresponding equivalent resistance gradient matrix. Secondly, according to the motion cycle diagram and equivalent resistance gradient matrix of the metamorphic mechanism, the structure and constraint form of the metamorphic joints were established, and the relationship between the force variation and the structure and the constraint form of the metamorphic joints was also obtained. Then, the structures of all 12 corresponding constrained metamorphic mechanisms were synthesized. Ultimately, one among the twelve mechanisms was chosen as the side-rolling metamorphic mechanism. The topological transformation of its working configuration was examined. The results confirmed the feasibility and practicality of the proposed structural synthesis method in this study.

Collaborative Allocation Method of En-Route Network Resources Based on Stackelberg Game Model

To further enhance fairness in the allocation of en-route space–time resources in the collaborative trajectory selection program, a study on the plan preferences between air traffic control (ATC) and airlines in the selection process of the final plan is conducted based on the initial resource allocation plans, considering the roles of airlines in resources allocation decisions. By using Stackelberg game theory, a game model is established for the roles played by ATC and airlines in the process of selecting plans. Then, combining the overall consideration of ATC for all affected flights, the preferences of airlines for initial allocation plans are obtained, and the option range of selectable plans is narrowed down to determine the optimal allocation plan. The results of the example analysis show that the proposed model and method can effectively select the optimal allocation plan from the six initial allocation plans, select the trajectories and entry slots in the congestion areas for airlines that better meet the operation demand, and provide the decision basis with more preferences for ATC to select the final allocation plan. When ATC prefers the lowest overall delay cost, the delay cost of the selected optimal allocation plan is 267.7 min, which is 23.84% lower than the traditional RBS algorithm; when considering the preferences of the main base airline in East China, the delay cost of the selected optimal allocation plan is 287.7 min, which is 18.15% lower than the traditional RBS algorithm.

The Iowa Health Data Resource (IHDR): an innovative framework for transforming the clinical health data ecosystem.

This manuscript will be of interest to most Clinical and Translational Science Awards (CTSA) as they retool for the increasing emphasis on translational science from translational research. This effort is an extension of the EDW4R work that most CTSAs have done to deploy infrastructure and tools for researchers to access clinical data. The Iowa Health Data Resource (IHDR) is a strategic investment made by the University of Iowa to improve access to real-world health data. The goals of IHDR are to improve the speed of translational health research, to boost interdisciplinary collaboration, and to improve literacy about health data. The first objective toward this larger goal was to address gaps in data access, data literacy, lack of computational environments for processing Personal Health Information (PHI) and the lack of processes and expertise for creating transformative datasets. A three-pronged approach was taken to address the objective. The approach involves integration of an intercollegiate team of non-informatics faculty and staff, a data enclave for secure patient data analyses, and novel comprehensive datasets. To date, all five of the health science colleges (dentistry, medicine, nursing, pharmacy, and public health) have had at least one staff and one faculty member complete the two-month experiential learning curriculum. Over the first two years of this project, nine cohorts totaling 36 data liaisons have been trained, including 18 faculty and 18 staff. IHDR data enclave eliminated the need to duplicate computational infrastructure inside the hospital firewall which reduced infrastructure, hardware and human resource costs while leveraging the existing expertise embedded in the university research computing team. The creation of a process to develop and implement transformative datasets has resulted in the creation of seven domain specific datasets to date. The combination of people, process, and technology facilitates collaboration and interdisciplinary research in a secure environment using curated data sets. While other organizations have implemented individual components to address EDW4R operational demands, the IHDR combines multiple resources into a novel, comprehensive ecosystem IHDR enables scientists to use analysis tools with electronic patient data to accelerate time to science.

Dual time-scale state-coupled co-estimation of state of charge, state of health and remaining useful life for lithium-ion batteries via Deep Inter and Intra-Cycle Attention Network

The reliable and safe battery operations demand accurate estimation of battery states in the battery management system (BMS). Due to the complex electrochemical process, it is challenging to directly measure SOH and RUL of batteries for online applications. Moreover, the battery aging can strongly influence SOC estimation. The paper proposes a SOC, SOH, and RUL co-estimation method based on Deep Inter and Intra-Cycle Attention Network (DIICAN). From BMS data, the dual time-scale inter-cycle and intra-cycle features are automatically extracted with neural networks. The state degradation attention (SDA) unit is employed to recognize the state evolution pattern and evaluate battery degradation degree based on inter-cycle features. In addition, the extracted inter-cycle and intra-cycle features are simultaneously coupled for the SOC estimation to calibrate the capacity degradation. The mean absolute error (MAE) and root mean square error (RMSE) of the state of health (SOH) estimation are 0.30% and 0.43%, respectively. For the remaining useful life (RUL) estimation, the MAE is 178.5, and the RMSE is 230.3. With state coupling, the MAE of the state of charge (SOC) estimation decreases from 2.36% to 1.00%, and the RMSE decreases from 3.27% to 1.36% over the entire lifespan. In summary, the proposed method can not only accurately co-estimate SOH and RUL, but improve SOC estimation accuracy over the entire battery life.

Psychophysiology, Cognitive Function, and Musculoskeletal Status Holistically Explain Tactical Performance Readiness and Resilience.

Thompson, AG, Ramadan, JH, Alexander, JS, and Galster, SM. Psychophysiology, cognitive function, and musculoskeletal status holistically explain tactical performance readiness and resilience. J Strength Cond Res 37(12): 2443-2456, 2023-This study aimed to advance the techniques used in quantifying holistic readiness and resilience within military personnel. Tactical performers, instructors, and applied human performance scientists designed a weeklong competition to reflect realistic operational demands, test specific underlying performance constructs, and elucidate how modernized assessments could drive programmatic action. By placing first in their installation's local preliminary competition, 34 active-duty Marines earned the opportunity to compete in a series of 7 intense events for the title of champion. All inferential statistics were set to a p ≤ 0.05 level of significance. Morning heart rate variability identified top from bottom quartile finishers before a single competition event. By day 3, morning countermovement jump force production (normalized reactive strength index-modified) and cognitive psychomotor vigilance were significant indicators of performance resilience and final competition group rank. Heart rate variability also tracked performer readiness across time, identifying within-group and between-group differences among top, bottom, and field. Collectively, these holistic assessments proved significant markers of acute and chronic tactical performance capabilities. In summary, the incorporation of psychophysiological monitoring, cognitive performance testing, and musculoskeletal force plate evaluations could help inform selection and support needs, drive workload or recovery modulation, and provide critical metrics for evaluating training efficacy and operational readiness. Defense organizations should consider routinely incorporating and actioning similar holistic status monitoring strategies in training and operational settings. Moreover, leveraging other tactical competitions may provide key opportunities for advancing the standard of practice through additional scientific investigation.

Dealing with abnormalities and deviations to enhance resilience in engineering Assets: A critical review from human factors and decision-making perspectives under complex operational contexts

With the growing scale of industrial demands, complexities, and uncertainties around asset engineering and operations due to advanced technology utilization, digitalization, sustainability, new operating models, etc., the sensitive role of abnormalities and deviations towards human safety, systems security, reliability and resilience of engineering assets and industrial systems are becoming even more significant for modern industrial sectors as well as societies in general. In these contexts, the abilities of operators to capture and sense-make early signals that emerge from engineering assets and systems need more attention since it enables them to enhance critical situation awareness (SA) during complex operations. This calls for proactive solutions that can integrate core data with operator knowledge using suitable logical approaches, particularly in a period where there is growing recognition that asset data can provide strong support for engineering and operational decisions in demanding contexts. Based on an ongoing research project, this paper sheds light on abnormalities and deviations; two specific attributes that should be better understood. The purpose is to explore how to capitalize them at very early sense-making stages to enhance situation awareness and thus resilience of dynamic and complex engineering assets and systems. Through a critical review of the current state of knowledge, together with industrial observations, this paper studies these core concepts in detail with due attention to the critical need of so-called priory contextual knowledge and hybrid contextual decision solutions. This R&D work explores proactive possibilities to mitigate inherent potentials for unwanted events and incidents to enhance resilience in the era of digital twins and cyber-physical systems, where complex technologies and operational demands generate new conditions for asset performance.

Systems dynamic model of wear evolution for sheaves used in oil and gas hoisting operations

The global offshore drilling contracted fleet consisted of 378 jack-ups, 68 semisubmersibles and 73 drillships, according to Westwood Global Energy Group, as of September 2023. In most regions of the world, rig activity has picked up from increased operator demand and pushed the marketed utilization for jack-ups from 76% to 85%, semis from 60% to 80% and drillships from 82% to 90%. The utilization of a drilling rig depends on various factors, mainly drilling demand, rig capability and specifications, and rig efficiency and reliability. Worn sheave grooves can pull down the efficiency of a drive by 8% accelerate the wearing of ropes reduce rope life, increase maintenance costs, and the need for more frequent rope replacements. Wear depth for sheaves is typically measured as the depth of wear in the groove of the sheave, which indicates the extent to which the sheave has worn down over time. Research on surface wear of wire rope caused by the contact between the wire rope and the sheave has rarely been carried out. Upgrading Archard’s Wear model is needed to provide a better estimation of wear depth for sheaves. Moreover, the wear coefficient shall be determined. Therefore, the purpose of this paper is to model the wear evolution caused by the sliding contact between the wire rope and the sheave. To achieve this purpose, case sheaves were purposefully selected, studied and modelled using both the analytical and simulation modelling approaches.

Fueling the seaport of the future: Investments in low‐carbon energy technologies for operational resilience in seaport multi‐energy systems

AbstractThe ability to withstand and recover from disruptions is essential for seaport energy systems, and in light of the growing push for decarbonization, incorporating clean energy sources has become increasingly imperative to ensure resilience. This paper proposes a resilience enhancement planning strategy for a seaport multi‐energy system that integrates various energy modalities and sources, including heating, cooling, hydrogen, solar, and wind power. The planning strategy aims to ensure the reliable operation of the system during contingency events, such as power outages, equipment failures, or extreme weather incidents. The proposed optimization model is designed as a mixed‐integer nonlinear programming formulation, in which McCormick inequalities and other linearization techniques are utilized to tackle the model nonlinearities. The model allocates fuel cell electric trucks (FCETs), renewable energy sources, hydrogen refueling stations, and remote control switches such that the system resilience is enhanced while incorporating natural‐gas‐powered combined cooling, heating, and power system to minimize the operation and unserved demand costs. The model considers various factors such as the availability of renewable energy sources, the demand for heating, cooling, electricity, and hydrogen, the operation of remote control switches to help system reconfiguration, the travel behaviour of FCETs, and the power output of FCETs via vehicle‐to‐grid interface. The numerical results demonstrate that the proposed strategy can significantly improve the resilience of the seaport multi‐energy system and reduce the risk of service disruptions during contingency scenarios.

Structural Integrity of Turbine Stator Blades Using Different Super Alloys with Internal Cooling at Fluid Temperature Range of 600 K – 700 K

The importance of turbines in power generation cannot be overstated. While the failure in stationary plants can lead to downtime and high repair costs, its failure in mobile plants like the jet engines can be catastrophic with attendant loss of lives. Hence, by all possible means, the prevention of turbine failure is a necessity, and a very good means of doing this is with the use of super-alloys. Super-alloys are tailored to withstand the demands of turbine operations especially stress and elevated temperature and pressure. The blades are thus, manufactured from super-alloys, and of prominence are the Nickel-based super-alloys. The performance of five different super-alloys: (DS) GTD 111, Ti-6Al-4V, Inconel 718, CMSX-4, and Nimonic 80A was simulated using COMSOL MultiPhysics 5.5 at cooling air temperature range of 600 K – 700 K. The mode of cooling employed in the study is only internal cooling. With the developed stress percentage of the yield stress value and the stator blade displacement at the operating conditions as the criteria of performance, super-alloy Ti-6Al-4V faired as the best material for the stator blade.

Robust transformer-based anomaly detection for nuclear power data using maximum correntropy criterion

Due to increasing operational security demands, digital and intelligent condition monitoring of nuclear power plants is becoming more significant. However, establishing an accurate and effective anomaly detection model is still challenging. This is mainly because of data characteristics of nuclear power data, including the lack of clear class labels combined with frequent interference from outliers and anomalies. In this paper, we introduce a Transformer-based unsupervised model for anomaly detection of nuclear power data, a modified loss function based on the maximum correntropy criterion (MCC) is applied in the model training to improve the robustness. Experimental results on simulation datasets demonstrate that the proposed Trans-MCC model achieves equivalent or superior detection performance to the baseline models, and the use of the MCC loss function is proven can obviously alleviate the negative effect of outliers and anomalies in the training procedure, the F1 score is improved by up to 0.31 compared to Trans-MSE on a specific dataset. Further studies on genuine nuclear power data have verified the model's capability to detect anomalies at an earlier stage, which is significant to condition monitoring.

Comparison of the cumulative energy demand of BEV and FCEV in their long-term operation

The paper presents a method of using the theory of cumulative energy demand to assess this demand in long-term operation of vehicles with a mileage forecast of up to 350,000 km. Based on the results of operational "consumption" of energy and taking into account the energy "costs" of obtaining it, a comparison of the currently popular BEV’s (Battery Electric Vehicles) and FCEV’s (Fuel Cell Electric Vehicles) was presented. The question arises how much energy must be used to propel vehicles in their natural operation. After calculating with using of available data, the answer is – by operating of FCEV’s in average two time more electricity is need as by operating of BEV’s.

Trajectory planning and control of large robotic excavators based on inclination-displacement mapping

This paper proposes a trajectory planning and control method for heavy-duty manipulators based on inclination-displacement mapping. In this method, inclination sensors are employed to detect the excavator's orientation. By establishing a mapping relationship between joint inclination and cylinder displacement, the cylinder displacement is indirectly derived. Subsequently, considering the typical excavation operation process, the spatial trajectory of the excavator bucket is planned. The displacement of the driving cylinder is determined using kinematic inverse solutions and cubic polynomials. The cylinder displacement is then controlled using a position-velocity (PV) strategy. This method is tested on a 95-ton large excavator, and the results demonstrate several advantages, including easy installation, low maintenance requirements, smooth trajectory planning, and high tracking accuracy. These qualities make it suitable for the demands of intelligent operations in large excavators. Thus, this paper offers valuable insights for the intelligent transformation of heavy excavators.

Hydrogen carriers for zero-emission ship propulsion using PEM fuel cells: an evaluation

Green hydrogen combined with PEM fuel cell systems is a viable option to meet the demand for alternative maritime fuels. However, hydrogen storage faces challenges, including low volumetric density, fire and explosion risks and transport challenges. We assessed over fifteen hydrogen carriers based on their maritime performance characteristics to determine their suitability for shipboard use. Evaluation criteria included energy density, locally zero-emission, circularity of process, safety, dehydrogenation process, logistic availability and handling. Thus, excluding ammonia and methanol because of these constraints, we found that borohydrides, liquid organic hydrogen carriers and ammoniaborane are the most promising hydrogen carriers to use on ships with PEM fuel cells. Borohydrides, specifically sodium borohydride, have high energy densities but face regeneration issues. The liquid organic hydrogen carrier dibenzyltoluene has a lower energy density but exhibits easy hydrogenation and good handling. Given varying operational demands, we developed a framework to assess the suitability of hydrogen carriers for use in different ship categories. Evaluating the three types of hydrogen carriers, using our framework and considering current practices, shows that these are viable options for almost all ship types. Thus, we have identified three types of hydrogen carriers, which should be the focus of future research.

Analysis of demand and cost trends of telecommunications market operators in Russia

The analysis of the functions of demand and costs of mobile operators leading in the telecommunications market is carried out. Methods used in the study: mathematical modeling, regression analysis, optimization. The Russian telecommunications market is an oligopoly, which makes it possible to analyze the dynamics of the development of the three dominant firms (MTS, Megafon and VimpelCom), since they are the ones that influence the market to a greater extent. Various models of demand and costs are considered in accordance with the dynamics of the development of the telecommunications market. Analysis of a number of weighted average voice communication prices depending on the total volume of voice traffic of services showed the presence of linear and power-law models of the inverse demand function in 2002-2021. The functions of the costs of mobile operators are considered and the direct dependence of costs on the volume of voice traffic is revealed. Cost functions and demand functions have a linear and power-law character in different periods of operator development, which indicates different types of scale expansion effects. The regularity of the influence of price dynamics on market growth is established: a power-law downward trend in prices stimulated intensive market growth, a linear downward trend contributed to smooth growth, and an increasing trend corresponding to the Giffen effect caused stagnation; it is proved that the concentration of the telecommunications market mainly determines the change in the market price of communication services; the principles of the impact of the scale-up effect on the change in the company's market share are defined: the constant return on scale-up can lead to both an increase and a narrowing of the operator's market share, depending on whether the company increases the pace of investment or reduces; the positive effect of scale-up, as a rule, causes an increase in the operator's market share, and the negative effect leads to a reduction in market share/ The conducted research can be used for further analysis of the actual data and trends of demand functions and costs functions of telecommunications market companies.

Nonflammable cosolvent enables methyl acetate-based electrolyte for 4.6 V-class lithium-ion batteries operating at −60 °C

The operational demands of electronic devices in extreme environments present a significant challenge for lithium-ion batteries (LIBs) to function effectively at low temperatures. To address this challenge, methyl acetate (MA) has been considered as a potential electrolyte solvent. However, MA-based electrolytes show poor compatibility with graphite anodes and high flammability, which affects battery cycle life and brings safety risks. Herein, an advanced non-flammable MA-based diluted highly concentrated electrolyte (DHCE) is designed incorporating ethoxy (pentafluoro) cyclotriphosphazene (PFPN) as a multifunctional diluent. The MA-based DHCE forms an anion-dominated solvation structure with improved thermodynamic stability. Moreover, PFPN synergizes with anions to form a protective LiF-rich solid electrolyte interphase (SEI), which effectively prevents continuous solvent decomposition. More importantly, PFPN can quench the free radical reaction and stop the spread of flames. This electrolyte enables LiNi0.65Co0.15Mn0.2O2 (NCM65)/graphite full cell to maintain 94.6 % capacity retention after 300 cycles at 1 C and −20 °C. Moreover, it allows NCM65/graphite full cell to operate at a high voltage of 4.6 V and −60 °C. This work demonstrates the importance of co-solvents for electrolyte design and provides new routes for the development of LIBs operating in harsh conditions.

A flow-rate-aware data-driven model of vanadium redox flow battery based on gated recurrent unit neural network

The vanadium redox flow battery (VRB) system involves complex multi-physical and multi-timescale interactions, where the electrolyte flow rate plays a pivotal role in both static and dynamic performance. Traditionally, fixed flow rates have been employed for operational convenience. However, in today’s highly dynamic energy market environment, adjusting flow rates based on operating conditions can provide significant advantages for improving VRB energy conversion efficiency and cost-effectiveness. Unfortunately, incorporating the electrolyte flow rate into conventional multi-physical models is overly complex for VRB management and control systems, as real-time operations demand low-computational and low-complexity models for onboard functionalities. This paper introduces a novel data-driven approach that integrates flow rates into VRB modeling, enhancing data processing capabilities and prediction accuracy of VRB behaviors. The proposed model adopts a gated recurrent unit (GRU) neural network as its fundamental framework, exhibiting exceptional proficiency in capturing VRB’s nonlinear voltage segments. The GRU network structure is carefully designed to optimize the predictive ability of the model, with flow rate considered as a crucial input parameter to account for its influence on VRB behavior. Model refinement involves analyzing well-designed simulation results obtained during VRB operations under various flow rates. Laboratory experiments were also designed and conducted, covering different conditions of currents and flow rates to validate the proposed data-driven modeling method. Comparative analyses were performed against several state-of-the-art algorithms, including equivalent circuit models and other data-driven models, demonstrating the superiority of the proposed GRU-based VRB model considering flow rates. Thanks to the GRU’s outstanding capability in processing time series data, the proposed model delivers impressively accurate terminal voltage predictions with a low error margin of no more than 0.023 V (1.3%) under wide operating ranges. These results indicate the efficacy and robustness of the proposed approach, highlighting the novelty and significance of accounting for flow rates in accurate VRB modeling for management and control system design.

Fatigue risk assessment of a Helicopter Emergency Medical Service crew working a 24/7 shift pattern: results of a prospective service evaluation

BackgroundThe work of Helicopter Emergency Medical Services (HEMS) teams crosses the boundaries of several high-risk occupations including medicine, aviation, and transport. Working conditions can be challenging and operational demands requires a 24-h rota, resulting in disruption of the normal circadian rhythm. HEMS crews are therefore prone to both mental and physical fatigue. As fatigue in medical providers is linked to poor cognitive performance, degradation of psychomotor skills and error, this study aimed to explore the existence of predictable patterns of crew-fatigue in a HEMS service.MethodsHEMS medical crew members working a 3-on 3-off forward rotating rota with a 5-week shift cycle were asked to do psychomotor vigilance tests (PVT) as an objective measure of fatigue. PVT testing was undertaken at the start, mid- and at the end of every shift during a full 5-week shift cycle. In addition, they were asked to score subjective tiredness with the Samn-Perelli Fatigue Scale (SPFS), and to keep a Transport Fatigue Assessment shift log, wherein they noted shift characteristics potentially related to fatigue. Primary outcome of interest was defined as the change in PVT and SPFS scores over time.ResultsMean baseline resting PVT in milliseconds at the start of the study period was 427 [390–464]. There was an overall trend towards higher PVT-scores with shift progression mean [95% CI] PVT at the start of shifts 447 [433–460]; halfway through the shift 452 [440–463]; end of the shift 459 [444–475], p = 0.10), whereas SPFS scores remained constant. Within a 5 week forward-rotating cycle, an overall trend towards a gradual increase in both average PVT (from 436 [238–454] to 460 [371–527, p = 0.68] ms;) and SPFS (from 2.9 [2.6–3.2] to 3.6 [3.1–4.0], p = 0.38) was observed, although significant interindividual variation was present. Reported SPFS scores ≥ 4 (moderate fatigue) were mainly related to workload (number of jobs) and transport mode (car-based shifts).ConclusionAn overall trend towards a decline in psychomotor vigilance and an increase in self-reported tiredness was found for HEMS crew over a 5-week shift cycle. Using a bespoke predictive fatigue tool on a day-to-day basis could increase fatigue awareness and provide a framework to which relevant mitigating options can be applied.

Solving the Inter-Terminal Truck Routing Problem for Delay Minimization Using Simulated Annealing with Normalized Exploration Rate

The growth in containerized shipping has led to the expansion of seaports, resulting in the emergence of multiple terminals. While multi-terminal systems increase port capacity, they also pose significant challenges to container transportation, particularly in inter-terminal movements. Consequently, the transportation delay of containers in inter-terminal operations demands crucial attention, as it can adversely affect the efficiency and service levels of seaports. To minimize the total transportation delays of the inter-terminal truck routing problem (ITTRP), we introduce simulated annealing with normalized acceptance rate (SANE). SANE improves the exploration capability of simulated annealing (SA) by dynamic rescaling of the transportation delay objective to modify the acceptance probability. To validate the quality of solutions provided by SANE, we have developed a mathematical model that provides a set of linear formulations for ITTRP constraints, avoiding the known set-partitioning alternative. Experimental results showed that for small-scale ITTRP instances, SANE achieved a solution close to the optimal. In larger instances with 100–120 orders, SANE found feasible suboptimal solutions within 15–21 seconds, which is unattainable using the exact solver. Further comparison with baselines indicates that SANE provides considerable improvements compared to both SA and Tabu search in terms of the objective value.

Toward automated instructor pilots in legacy Air Force systems: Physiology-based flight difficulty classification via machine learning

The United States Air Force (USAF) is struggling to train enough pilots to meet operational requirements. Technology has advanced rapidly over the last 70 years but USAF pilot training has not. Modern operational requirements demand a change and, for this reason, USAF senior leadership has advocated for innovation. The automation of instructor and evaluator pilots in select bottlenecks (e.g., simulators) is one such measure. However, to implement this vision, numerous technical issues must be mitigated. Accurate classification of flight difficulty is a foundational problem underpinning many of these technical issues, which requires either the acquisition of new systems or the development of new procedures. Therefore, given this need and the costly nature of purchasing new equipment, physiological-based classification of flight difficulty is our focus herein. Leveraging multimodal data from a designed experiment of pilots landing a simulated aircraft, we develop a high-quality machine learning pipeline for classifying flight difficulty, called the Multi-Modal Functional-based Decision Support System (MMF-DSS). MMF-DSS distills a tabular set of features from our multimodal and functional data through the use of functional principal component analysis, summary statistics, and BorutaSHAP. In this manner, information is derived from the time-series data via the generation of hundreds of features, of which a small subset having the most predictive capability is discerned. Four full factorial designs are used to perform hyperparameter tuning on a set of classifiers. In so doing, a superlative technique is identified. Impacts on executive decision making are examined as well as associated policymaking implications. Alternative classifiers are considered for use within our pipeline that trade predictive accuracy for cost efficiency, and recommendations for choosing among these alternatives is provided.

A Closer Look at Bearing Fault Classification Approaches

Rolling bearing fault diagnosis has garnered increased attention in recent years owing to its presence in rotating machinery across various industries, and an ever increasing demand for efficient operations. Prompt detection and accurate prediction of bearing failures can help reduce the likelihood of unexpected machine downtime and enhance maintenance schedule, averting lost productivity. More recent technological advances have enabled monitoring the health of these assets at scale using a variety of sensors, and predicting the failures using modern machine learning (ML) approaches. Deep neural network approaches that has demonstrated superior performance across computer vision and natural language processing tasks are increasingly considered for failure prediction using vibration sensors. Vibration data has been collected using accelerated run-to-failure of overloaded bearings, or by introducing known failure in bearings, under a variety of operating conditions like rotating speed, load on the bearing, type of bearing fault, and data acquisition frequency. However, in using deep neural networks or other ML models for predicting failure using vibration data, there is a lack of consistency in the choice of how these approaches are evaluated, what portion of data from run-to-failure experiments are considered for training failure prediction models, or even how these problems are formulated as machine learning classification problems. An understanding of the impact of these choices is important to reliably develop models, and deploy them in practical settings. In this work, we demonstrate the significance of these choices on the performance of the models using publicly-available vibration datasets and discuss the implications of these choices for the models to be useful in real world scenarios.