Institute of Industrial Engineering

In manufacturing systems with a job shop organization, queues between workstations create an intermittent process flow, allowing workers to schedule tasks entering the queue based on their needs and preferences. The resulting scheduling autonomy of individual workers often leads to inefficiencies in the overall production process due to the loss of control. Companies are therefore increasingly using algorithmic scheduling systems to assign task sequences to workers, thereby drastically reducing their autonomy and negatively affecting their job performance and well-being. This paper extends the existing flexible job shop scheduling problem by sequencing preferences (FJSPSP) to incorporate a human-centered perspective by predicting workers’ task sequencing decisions using learning-to-rank (LTR) methods. By learning workers’ individual task sequencing preferences, it becomes possible to predict the processing sequence based on task characteristics. The scheduling algorithm for the FJSPSP presented in the paper incorporates workers’ learned sequencing preferences as constraints. Considering workers’ learned task sequencing decisions, the FJSPSP optimizes only task assignments to maintain workers’ autonomy over task sequences. The contributions of this paper are fourfold, namely, (1) presenting an approach to elicit sequencing decision datasets from workers, (2) demonstrating the successful prediction of humans’ and an actual worker’s task sequencing decisions with LTR, (3) formulating the FJSPSP variant that integrates workers’ sequencing preferences as constraints and proving its effectiveness in a simulation study, and (4) consolidating these steps into an explainable artificial intelligence (XAI)- and LTR-enabled sociotechnical system design framework. The paper closes with a discussion of the overall methodology and future research perspectives. • Task sequencing decisions of workers are elicited through a simulation and IT systems. • Algorithms for predicting workers’ task sequencing decisions are compared. • Flexible job shop scheduling with sequencing preferences (FJSPSP) is proposed. • The FJSPSP jointly optimizes scheduling autonomy and overall scheduling performance. • A framework for sociotechnical system design and decision support is presented.

IEOM Society Publication Database: A dedicated hub for knowledge and insights in the field of Industrial Engineering and Operations Management. Explore research, articles, and resources that exemplify our commitment to advancing IE & OM globally, fostering critical thinking, and serving humanity.

Electrical muscle stimulation (EMS) can create compelling weight sensations in virtual reality by activating antagonist muscles, prompting users to exert effort against induced contractions. This experiment investigates personalizing EMS amplitudes to precisely make a 2 kg object feel 1 kg heavier. This task is challenging due to high individual variability in sensory, motor, and pain thresholds, as well as physical characteristics. We applied nine personalized EMS intensity levels to the triceps and extensor carpi ulnaris muscles of 75 participants, who then compared the perceived weight of the stimulated arm against a 3 kg reference. A machine learning model was trained on the collected data to predict the perceived weight difference from user-specific features. The best-performing XGBoost model achieved a classification accuracy of 77%. Feature analysis revealed that the applied EMS amplitude was the most influential predictor, supplemented by individual motor and pain thresholds, body mass index, and forearm length. This ML-based personalization is a significant advancement for creating adaptive haptic feedback, with applications in immersive VR, physical rehabilitation, and physiotherapy.

Wafer-level optics manufacturing has gained significant attention in recent years due to its potential to enable cost-effective production of miniaturized optical components for various applications. To meet the requirements of this manufacturing approach, tailored polymers compatible with the nanoimprint lithography (NIL) replication process have emerged as a versatile class of materials offering a range of properties necessary for the fabrication of high-quality optical devices. Development of these materials requires a careful balancing of all relevant properties including optical function, process constraints, (thermo-)mechanical properties and stability under reliability testing. DELO developed a modular system of materials which can be modified in this wide parameter space. This manuscript provides an overview of DELO's transparent and functional materials utilized in waferlevel optics manufacturing today, targeted at applications such as imaging & sensing, co-packaged optics or display and projection devices.

This research was carried out with the aim of investigating the effect of employees’ perception of artificial intelligence on the willingness to accept and implement artificial intelligence in knowledge-based companies. The statistical population of this research is the employees of knowledge-based companies across the Iran, because their number is unrestricted, 384 people were determined as the sample size and data was collected through an electronic questionnaire. The questions of the questionnaire were adapted from the article of Ahn and Chen 2022. In this research, Amos24 software was used to test hypotheses and model structural equations. The results showed that employees’ perceptions of AI (including 1: understanding the benefits of AI; 2: understanding the concerns of AI; 3: understanding the vision of AI; and 4: experience of using and ease of access to AI) have an impact on the willingness to adopt and implement AI in knowledge-based companies.

Tall buildings with different structural systems can be exposed to explosive loading, and a proper understanding of their structural response under blast loads is crucial for robust structural design. This study investigates the structural performance of selected tall structural systems - shear wall frame, outrigger, and tube-in-tube systems - under different blast load effects. Nonlinear finite element analysis was performed using commercial package ETABS with blast loads applied as time-history functions for various charge weights and standoff distances. Structural responses were compared in terms of displacement and inter-story drift. Results showed that the tube-in-tube system performed relatively better in terms of displacement and inter-story drift as standoff distance increased and charge weight was constant. When explosive charge weights increased, and standoff distance was constant, the tube-in-tube system displayed better performance with respect to displacement, whereas the outrigger system was the optimum with reference to inter-story drift. The overall study indicates that selecting a tall structural system for better resilience against blast loading depends on specific aspects of structural behavior, identified as the required criteria.

The transition to sustainable energy is paramount for addressing climate change, and low-carbon technologies play a pivotal role in this shift in the USA. The prime objective of this research paper was to apply the capabilities of machine learning in an examination of America's low-carbon technology trading. With powerful analysis tools, we attempted to detect trends in exporting and importing, estimate the contribution of such technology to the economy, and estimate the effectiveness of supporting policies. The scope of our activity was U.S. low-carbon technology trade, both its imports and its exports. Examining a rich dataset including volumes of trade, technological categories, and economic factors, we try to unveil deeper trends driving this new sector. The dataset for analysis in such a case involved in-depth information drawn from a range of reliable sources, including U.S. trade reports, economic statistics, and global databases for sustainability. Trade volumes, in terms of value and quantity of low-carbon technology exported and imported, form one of the key variables in such a dataset. There was extensive information about carbon emissions, providing an analysis of the impact on terms of the environment through such technology, and policy incentives, in terms of government actions for encouragement of low-carbon alternatives. In selecting machine learning models for examining low-carbon technology trade, three candidates—Logistic Regression, Support Vector Machines (SVM), and K-Nearest Neighbor (KNN)—stood out for their particular strengths. In terms of accuracy, the SVM model is the top scorer, closely followed by KNN, while Logistic Regression takes a considerable drop, indicating its relatively lower predictive capability. Precision measurements also rank similarly, with SVM and KNN recording high precision values, suggesting that they are reliable in predicting true positives. Recall scores also indicate the strength of SVM and KNN in recalling all instances, while the Logistic Regression model records lower recall, particularly in predicting the class. Finally, the F1 score, being the trade-off between precision and recall, further reinforces the superior performance of SVM and KNN, as both models record high scores, with Logistic Regression lagging. To enhance the U.S. position in the global low-carbon technology market, a multi-faceted approach must be taken. Firstly, efforts must be made to drive innovation through increased investment in research and development (R&D). For business firms and investors, the transition to a low-carbon economy presents a plethora of market opportunities in the low-carbon technology sector. U.S. firms can leverage growing consumer demand for green products by developing product lines to cater to renewable energy systems, energy-efficient appliances, and electric vehicles. For investors, an understanding of the dynamics of the low-carbon technology market is essential for risk management through predictive economic modeling. It is necessary to synchronize trade policy with U.S. and global carbon reduction objectives to foster a sustainable economy.

The existing tunnel detection structural analysis methods have certain problems in fault identification, health assessment and positioning accuracy. In particular, when processing complex tunnel structure data, traditional methods are difficult to adapt to diverse fault types and real-time data changes. To this end, this paper introduces neural network technology, aiming to solve the shortcomings of traditional methods in tunnel detection through deep learning models and improve the accuracy and efficiency of fault detection. This paper adopts a recurrent neural network (RNN) model and combines various sensor data from the tunnel for structural analysis. In the model design, the input layer receives various sensor data such as temperature, pressure, crack width, settlement displacement, etc. from the sensor, and outputs the tunnel health assessment results after multi-layer neural network processing. The model uses the ReLU activation function to optimize nonlinear feature extraction, adjusts network weights through the back-propagation algorithm, and accurately identifies fault types such as cracks, settlement, and deformation. In fault location, through the relationship between network learning and sensor data, the model can locate the fault location and output accurate location results. The experimental results show that in the diagnosis of "crack" faults, the accuracy of the neural network is 90%, much higher than the 75% based on the threshold algorithm; in the diagnosis of "settlement" and "deformation" faults, the accuracy of the neural network is also about 7% and 12% higher than the threshold algorithm. In addition, the accuracy of fault location is also significantly improved. The model can locate faults within an error range of ±1 meter and has a high spatial resolution.