Operator Workload Research Articles

Analysis of workload and long rest periods on mobile rice grain milling operator at Sidolelono Pleret Community Bantul

Noise and vibration are the hallmarks of mobile rice grain milling. The difference in working time between operators, the number of grains milled in one day, and looking for customers activity can be a factors that affect the operator’s workload. This study aims to determine the workload experienced, identify the factors that affects the workload, and determine the long rest periods required by the mobile rice grain milling operator. The workload experienced by operators measured by the cardiovascular load (CVL), the level of caloric needs, and NASA-TLX questionnaire. Factors considered in determining operator workload are age, body mass index (BMI), body temperature, noise, ambient temperature, humidity, and vibration. The long rest periods are calculated based on the operator’s heart rate converted into energy consumption. Based on CVL, operators experience no fatigue while from the level of caloric needs and mental workload, operators experience a moderate workload. The partial t-test for one operator group indicates that there is an effect of BMI and ambient temperature on the level of caloric needs. The regression test for two operators group showed that body temperature had an effect of 63,9% on CVL. The long rest periods applied by the operator is sufficient.

Higher social cue utilisation improves communication, reduces perceived workload, and improves performance amongst ad hoc dyads in simulated rail control

The objective was to examine whether ad hoc dyads with different collective social cue utilisation would record differences in performance and perceptions of workload during a simulated rail control task that incorporated distinct levels of demand. The frequency of two types of communicative statements was also examined as mediating factors: closing the loop and informative responding. A quasi-experimental design was employed using 40 dyadic teams. The results indicated that ad hoc teams whose members comprised higher social cue utilisation recorded relatively faster response times and perceived lower levels of workload, and engaged in a greater frequency of communicative statements that involved ‘closing the loop’ and ‘informative responses’. Social cue utilisation also exerted an indirect effect on perceived workload through informative responding. The outcomes have theoretical implications for models of ad hoc team performance, and practical implications for the selection and training of teams that operate on an ad hoc basis. Practitioner summary: This study indicates that, in the context of simulation tasks requiring teamwork, higher social cue utilisation amongst ad hoc team members is associated with communication, workload perception, and overall performance.

Relationship between Workload Performance and Job Satisfaction

In Pakistan’s public health care delivery system, charge nurses hold a very challenging position to perform their workload. They have to work very hard to accomplish nursing and non-nursing care tasks which are imposed on them by the system. Overstretching of workload deprives them from concentration which is badly needed for their performance and this creates dissatisfaction which negatively impact on the quality of nursing care. As a result, this study intends to analyses the relationship between workload performance (WLP) and job satisfaction (JS). This correlation study involved 105 charge nurses in Nishtar Medical College and Hospital Multan Pakistan recruited by convenience sampling. Nurses’ WLP were collected by using self-developed instrument, and JS were collected by using modified Spector 1985. The collected data were analyze descriptively (mean, SD, frequencies, percentage) and inferentially (Pearson’s correlation). The results suggested that nurses’ WLP (average time consumption) in the morning (93.83%) and evening shift (95.63%) were higher compared with night shift (70.69%). Additionally, the proportion of time consumed in the morning and evening shift were higher on nursing care activities (Morning = 57.10%, Evening =52.1%) rather than non-nursing care activities. Oppositely, in the night shifts nurses consumed more than half of their time in non-nursing care activities (55. 66%). Charge nurses observed on moderate level of job satisfaction (mean = 38.6, SD = 5.42). There was no statistically significant correlation between nurses’ job satisfaction and workload performance (p =.137). The findings conclude nurses’ high workload and moderate level of job satisfaction and no statistically significant correlation between both. The results suggested that, it is important for hospital management to adopt some better strategies in order to improve WLP and JS.

Woodpecker+: Customized workload performance evaluation based on data characteristics

Woodpecker+: Customized workload performance evaluation based on data characteristics

Side Channel Attacks on GPUs

Graphics Processing Units (GPUs) are commonly integrated with computing devices to enhance the performance and capabilities of graphical workloads. In addition, they are increasingly being integrated in data centers and clouds such that they can be used to accelerate data intensive workloads. Under a number of scenarios the GPU can be shared between multiple applications at a fine granularity allowing a spy application to monitor side channels and attempt to infer the behavior of the victim. For example, OpenGL and WebGL send workloads to the GPU at the granularity of a frame, allowing an attacker to interleave the use of the GPU to measure the side-effects of the victim computation through performance counters or other resource tracking APIs. We demonstrate the vulnerability by implementing three end-to-end attacks. We show that an OpenGL or CUDA based spy can fingerprint websites accurately (attack I), track user activities within the website, and even infer the keystroke timings for a password text box (attack II) with high accuracy. The third attack demonstrates how a CUDA spy application can derive the internal parameters of a neural network model being used by another CUDA application on the cloud. To counter these attacks, the paper suggests mitigations based on limiting the rate of the calls, or limiting the granularity of the returned information.

Rancang Bangun Sistem Informasi Manajemen Pengabdian Masyarakat Berbasis Web di Politeknik Elektronika Negeri Surabaya

Community service activities as mentioned in Undang-Undang of Indonesia State No. 20 of 2003 concerning the National Education system, article 20 is one of the mandates of the Tri Dharma of Higher Education, including the Politeknik Elektronika Negeri Surabaya (PENS). Management of community service in the Research and Community Service Center (P3M) is still manual in collecting, storing and processing service data. The system raises several problems including delays, inaccuracies and excessive workload for administrators and operators and stakeholders in processing and accessing data of community service. Therefore it is necessary to make an application for a multiplatform-based information system to handle transactions and management of various community service data. This application handles the collection of proposal data, progress reports and final reports and handles proposal selection transactions, monitoring and evaluating progress reports, as well as seminar results, including the plotting of service reviewers. This system uses PHP framework and MySQL database management system. The test results show that all application functions are running well and make it easy for all service stakeholders to manage community service activities.

A Closed-Loop Controller to Ensure Performance and Temperature Constraints for Dynamic Applications

To secure correct system operation, a plethora of Reliability, Availability and Serviceability (RAS) techniques have been deployed by circuit designers. RAS mechanisms however, come with the cost of extra clock cycles. In addition, a wide variety of dynamic workloads and different input conditions often constitute preemptive dependability techniques hard to implement. To this end, we focus on a realistic case study of a closed-loop controller that mitigates performance variation with a reactive response. This concept has been discussed but was only illustrated on small benchmarks. In particular, the extension of the approach to manage performance of dynamic workloads on a target platform has not been shown earlier. We compare our scheme against the version of a Linux CPU frequency governor in terms of timing response and energy consumption. Finally, we move forward and suggest a new flavor of our controller to efficiently manage processor temperature. Again, the concept is illustrated with a realistic case study and compared to a modern temperature manager.

원자력발전소 기동 및 정지 운전을 위한 순환 신경망 기반 인공지능 프레임워크 개발

In order to reduce operator workload from startup and shutdown operations for existing Nuclear Power Plants (NPPs), it is necessary to develop an automation system based on deep learning, the leading approach in current Artificial Intelligence (AI) technology. From existing research, it is challenging to develop an automation system using conventional machine learning for startup and shutdown operation since the automation system needs to be able to handle many instances of both monitoring and control variables in NPPs. Deep learning is able to simulate a variety of operating actions based on the experience of each operator. In this study, an AI framework for an automation system for startup operation in NPPs has been developed using a Recurrent Neural Network (RNN), which is a robust deep learning method for time series analysis. A feasibility study for an AI framework for the automation system is conducted using a Compact Nuclear Simulator (CNS) based on Westinghouse three-loop NPPs. The target scenario for the feasibility study is operation under bubble creation conditions in a pressurizer under startup.

원자력발전소 기동 및 정지 운전을 위한 순환 신경망 기반 인공지능 프레임워크 개발

In order to reduce operator workload from startup and shutdown operations for existing Nuclear Power Plants (NPPs), it is necessary to develop an automation system based on deep learning, the leading approach in current Artificial Intelligence (AI) technology. From existing research, it is challenging to develop an automation system using conventional machine learning for startup and shutdown operation since the automation system needs to be able to handle many instances of both monitoring and control variables in NPPs. Deep learning is able to simulate a variety of operating actions based on the experience of each operator. In this study, an AI framework for an automation system for startup operation in NPPs has been developed using a Recurrent Neural Network (RNN), which is a robust deep learning method for time series analysis. A feasibility study for an AI framework for the automation system is conducted using a Compact Nuclear Simulator (CNS) based on Westinghouse three-loop NPPs. The target scenario for the feasibility study is operation under bubble creation conditions in a pressurizer under startup.

Automating distributed tiered storage management in cluster computing

Data-intensive platforms such as Hadoop and Spark are routinely used to process massive amounts of data residing on distributed file systems like HDFS. Increasing memory sizes and new hardware technologies (e.g., NVRAM, SSDs) have recently led to the introduction of storage tiering in such settings. However, users are now burdened with the additional complexity of managing the multiple storage tiers and the data residing on them while trying to optimize their workloads. In this paper, we develop a general framework for automatically moving data across the available storage tiers in distributed file systems. Moreover, we employ machine learning for tracking and predicting file access patterns, which we use to decide when and which data to move up or down the storage tiers for increasing system performance. Our approach uses incremental learning to dynamically refine the models with new file accesses, allowing them to naturally adjust and adapt to workload changes over time. Our extensive evaluation using realistic workloads derived from Facebook and CMU traces compares our approach with several other policies and showcases significant benefits in terms of both workload performance and cluster efficiency.

Optimal Performance-Aware Cooling on Enterprise Servers

Datacenters house massive databases and applications to provide business decision support and cloud services where commercial success is contingent on timely responses. The servers that these tasks run on dissipate a lot of power, requiring equally powerful cooling systems to maintain a safe and efficient temperature level. In a typical enterprise server, server chassis fans can generate vibrations that are powerful enough to degrade the performance of data-intensive workloads. Our methodology measures and reproduces real-life vibrations on a rack server to evaluate the performance of different hard disks. Effective hardware management relies on an accurate understanding of these devices and their interactions to mitigate any performance degradation and meet thermal constraints. While current strategies focus on managing processing resources, at the expense of more data-dependent workloads, this paper approaches server efficiency by targeting the cooling-performance relationship in conjunction with other dependencies between power, thermal, and cooling. We extract a model from common database benchmarks based on expected resource utilization and corresponding cooling needs, while considering these mechanical disturbances. Our proposed strategy uses convex optimization to maintain thermal constraints at all times, while reducing the energy consumption of a server by 65% compared to basic proportional-integral-derivative controllers, or by 19% in comparison to advanced hardware management techniques proposed in literature.

Evaluating Resource Centric Behavior of Workloads and Performance Analysis in CMPs due to Shared Resources

Multicore systems are achieving new dimensions to meet the pace between processor technology and high processing demands. These chip multiprocessors (CMPs) share some on- chip and off-chip resources to achieve higher performance. With the increase in core integration, the performance of workloads is dependent on the allocation of resources. Though the CMPs elevate performance, but the challenge imposed due to subtle interactions of several applications contending for resources leads to performance degradation by several magnitudes. In this direction the work performs a two-fold evaluation of application programs. The applications running on CMP cores depict distinct behavior towards consumption of shared resources. It characterizes the resource centric nature of SPEC CPU2006 benchmarks based on their resource consumption behavior. Secondly the work aims to evaluate the effect of inter-core interference on the performance of application programs based on the characterization obtained and the potential contention caused on performance due to corunners. Lastly we place significant remarks on the impact on performance due to resource sharing and its implication on resource contention.

Effects of Performance and Task Duration on Mental Workload during Working Memory Task

N-back is a working memory (WM) task to study mental workload on the prefrontal cortex (PFC). We assume that the subject’s performance and changes in mental workload over time depends on the length of the experiment. The performance of the participant can change positively due to the participant’s learning process or negatively because of objective mental fatigue and/or sleepiness. In this pilot study, we examined the PFC activation of 23 healthy subjects while they performed an N-back task with two different levels of task difficulty (2-, and 3-back). The hemodynamic responses were analyzed along with the behavioral data (correct answers). A comparison was done between the hemodynamic activation and behavioral data between the two different task levels and between the beginning and end of the 3-back task. Our results show that there is a significant difference between the two task levels, which is due to the difference in task complication. In addition, a significant difference was seen between the beginning and end of the 3-back task in both behavioral data and hemodynamics due to the subject’s learning process throughout the experiment.

The effect of repeated full immersion simulation training in ureterorenoscopy on mental workload of novice operators

BackgroundDifficult surgical procedures may result in a higher mental workload, leading to increased fatigue and subsequent errors. This study was aimed to investigate the effect of repeated simulation training in ureterorenoscopy in a high-fidelity setting on the performance and mental workload of novice operators.MethodsMedical students voluntarily participated in the present simulation study. After a didactic and video-based lecture, they underwent simulation training involving a renal stone case, including a rigid cystoscope component (task 1, performing a WHO checklist, assembling a scope, and insertion of a guide-wire and an access sheath after examining the bladder) and a flexible ureterorenoscope component (task 2, retrieving a stone located in the upper calyx using a basket after inspecting the upper, middle, and lower calyx). Training was performed in a mock operating theater. Technical skills were assessed by one author (an experienced urologist) onsite using an Objective Structured Assessment of Technical Skills (OSATS) score at each training session. The mental workload was subjectively evaluated by the National Aeronautics and Space Administration Task Load Index (NASA-TLX) questionnaire after each training session.ResultsSeventeen students completed a minimum of 6 training sessions (male: female = 10: 7, median age of 22) over a median of 21 days (range, 10–32). In both tasks 1 and 2, the OSATS score improved over the 6 sessions with evidence of plateauing (MANOVA model, task 1: p < 0.0001, task 2: p < 0.0001). In contrast, the NASA-TLX score persistently decreased without plateauing (task 1: p = 0.0005, task 2: p = 0.0028).ConclusionsUnder repeated simulation training in ureterorenoscopy in a high-fidelity setting, participants showed a continual decrease of the mental workload, while the improvement of technical skills reached a plateau over the 6 sessions. Our study showed the important benefit of simulation training to reduce the mental workload by repeated scenario training before actual clinical practice.

Construction of ICU nursing workload performance appraisal index based on information system

Objective To develop a reasonable performance evaluation index system for ICU nursing workload based on the hospital information system and the structure-process-result model. Methods The preliminary indicators of ICU nurses' performance appraisal system were constructed by the research group. Delphi method was used for expert consultation to evaluate the importance of the indicators of the performance appraisal system. After summarizing, sorting out and modifying the expert opinions, the performance appraisal index system of ICU nurses was obtained. SPSS 19.0 software is used to describe the importance of experts' evaluation of indicators at all levels, such as mean, standard deviation, coordination coefficient of experts' opinions and degree of coordination. Results The effective recovery rate of expert consultation papers was 100%. In the first round of inquiry, the average of importance assignment was 2.05-5.00, the average of rationality assignment was 2.13-5.00, and the average of importance assignment in the second round was 4.57-5.00. The average of rationality assignment was 4.71-5.00. The coefficient of variation of all levels of indicators in the first round was 0.02-0.18, the coefficient of variation in the second round was 0.02-0.12, which was less than 0.2, the positive coefficient of experts was 100%, and the authoritative coefficient of experts in the second round was 0.86 and 0.88. After two rounds of Delphi expert consultation, the ICU nurses' performance appraisal index system was finally formed, including three first-level indicators and 37 second-level indicators. Conclusion This study is based on the information system, according to the structure-process-result model, using the expert consultation method to construct the ICU nursing workload performance assessment index system, and establish the weight of each index, the research results are helpful to further improve the performance appraisal system of ICU nursing staff and improve the scientific management level of ICU. Key words: Intensive care; Nursing workload; Performance appraisal; Information system; Index system

Spectral-based convolutional neural network without multiple spatial-frequency domain switchings

Recent researches have shown that spectral representation provides a significant speed-up in the massive computation workload of convolution operations in the inference (feed-forward) algorithm of Convolutional Neural Networks (CNNs). This approach results in reducing the computational complexity of the classification task, which makes spectral-based CNN suitable for implementation on embedded platform that typically has constrained resources. However, a major challenge in this approach is that the mathematical formulation of a nonlinear activation function in spectral (frequency) domain is currently not available; hence, computation of the activation functions in each layer has to be performed in the spatial domain. This results in several spatial-frequency domain switchings that are computationally very costly, and as such, it would be advantageous to strictly stay in the frequency domain. Hence, in this work, a novel Spectral Rectified Linear Unit (SReLU) for the activation function is proposed, that makes it possible for the computations to remain in the frequency domain, and therefore avoids the multiple compute-intensive domain transformations. To further optimize the classification speed of the network, an efficient spectral-based CNN model is presented that uses only the lower frequency components by way of fusing the convolutional and sub-sampling layers. Additionally, we provide and utilize a frequency domain equivalent of the conventional batch normalization layer that results in improving the accuracy of the network. Experimental results indicate that the proposed spectral-based CNN model achieves up to 17.02 × and 3.45 × faster classification speed (without considerable accuracy loss) on AT&T face recognition and MNIST digit/fashion classification datasets, respectively, as compared to the equivalent models in the spatial domain, hence outperforming conventional approaches significantly.

Individual-Specific Classification of Mental Workload Levels Via an Ensemble Heterogeneous Extreme Learning Machine for EEG Modeling

In a human–machine cooperation system, assessing the mental workload (MW) of the human operator is quite crucial to maintaining safe operation conditions. Among various MW indicators, electroencephalography (EEG) signals are particularly attractive because of their high temporal resolution and sensitivity to the occupation of working memory. However, the individual difference of the EEG feature distribution may impair the machine-learning based MW classifier. In this paper, we employed a fast-training neural network, extreme learning machine (ELM), as the basis to build an individual-specific classifier ensemble to recognize binary MW. To improve the diversity of the classification committee, heterogeneous member classifiers were adopted by fusing multiple ELMs and Bayesian models. Specifically, a deep network structure was applied in each weak model aiming at finding informative EEG feature representations. The structure of hyper-parameters of the proposed heterogeneous ensemble ELM (HE-ELM) was then identified and then its performance was compared against several competitive MW classifiers. We found that the HE-ELM model was superior for improving the individual-specific accuracy of MW assessments.

Accurate Classification for Automatic Vehicle-Type Recognition Based on Ensemble Classifiers

In this paper, a real-world problem of the vehicle-type classification for automatic toll collection (ATC) is considered. This problem is very challenging because any loss of accuracy even of the order of 1% quickly turns into a significant economic loss. To deal with such a problem, many companies currently use optical sensors (OSs) and human observers to correct the classification errors. Herein, a novel vehicle classification method is proposed. It consists in regularizing the problem using one camera to obtain vehicle class probabilities using a set of convolutional neural networks (CNNs) and, then, uses the Gradient boosting-based classifier to fuse the continuous class probabilities with the discrete class labels obtained from the OS. The method is evaluated on a real-world dataset collected from the toll collection points of the VINCI Autoroutes French network. The results show that it performs significantly better than the existing ATC system and, hence, will vastly reduce the workload of human operators.

Press Start to Play: Classifying Multi-Robot Operators and Predicting Their Strategies through a Videogame

One of the active challenges in multi-robot missions is related to managing operator workload and situational awareness. Currently, the operators are trained to use interfaces, but in the near future this can be turned inside out: the interfaces will adapt to operators so as to facilitate their tasks. To this end, the interfaces should manage models of operators and adapt the information to their states and preferences. This work proposes a videogame-based approach to classify operator behavior and predict their actions in order to improve teleoperated multi-robot missions. First, groups of operators are generated according to their strategies by means of clustering algorithms. Second, the operators’ strategies are predicted, taking into account their models. Multiple information sources and modeling methods are used to determine the approach that maximizes the mission goal. The results demonstrate that predictions based on previous data from single operators increase the probability of success in teleoperated multi-robot missions by 19%, whereas predictions based on operator clusters increase this probability of success by 28%.

A Renewable Energy Driven Approach for Computational Sprinting

Computational Sprinting, which allows a chip to exceed its power and thermal limits temporarily by turning on all processor cores and absorbing the extra heat dissipation with certain phase-changing materials, has proven to be an effective way to boost the computing performance for bursty workloads. However, extra power available for sprinting is constrained by existing power distribution infrastructures. Using batteries alone to provide the additional power to achieve performance target not only limits the effectiveness of sprinting, but also negatively impacts the lifetime of the batteries. Leveraging renewable power supply in a green data center provides an opportunity to make full use of Computational Sprinting. However, the intermittent nature of renewable energy, along with limited cooling capacity, makes it very challenging. In this paper, we propose GreenSprint, a renewable energy driven approach that enables a data center to boost its computing performance efficiently by conducting computational sprinting under the intermittent and time-varying nature of renewable energy supply. Three basic strategies are designed to determine the core count and frequency level for sprinting based on current power supply. Furthermore, we propose a Hybrid strategy that combines reinforcement learning to dynamically determine the optimal server setting, targeting at both the power provision safety and the quality of service. In consideration of practical cooling conditions, we also present a thermal-aware sprinting strategy Hybrid-T. Finally, we build an experimental prototype to evaluate GreenSprint on a cluster of 10 servers with a simulated solar power generator. The results show that renewable energy by itself can sustain different duration lengths of sprinting when its supply is sufficient and can improve performance by up to 4.8x for representative interactive applications. We also show the effectiveness of core-count and frequency scaling in the presence of varied renewable power and limited battery energy.