Anomaly Detection and Inter-Sensor Transfer Learning on Smart Manufacturing Datasets.

Smart manufacturing systems are considered the next generation of manufacturing applications. One important goal of the smart manufacturing system is to rapidly detect and anticipate failures to reduce maintenance cost and minimize machine downtime. This often boils down to detecting anomalies within the sensor data acquired from the system which has different characteristics with respect to the operating point of the environment or machines, such as, the RPM of the motor. In this paper, we analyze four datasets from sensors deployed in manufacturing testbeds. We detect the level of defect for each sensor data leveraging deep learning techniques. We also evaluate the performance of several traditional and ML-based forecasting models for predicting the time series of sensor data. We show that careful selection of training data by aggregating multiple predictive RPM values is beneficial. Then, considering the sparse data from one kind of sensor, we perform transfer learning from a high data rate sensor to perform defect type classification. We release our manufacturing database corpus (4 datasets) and codes for anomaly detection and defect type classification for the community to build on it. Taken together, we show that predictive failure classification can be achieved, paving the way for predictive maintenance.

A Discrete Artificial Bee Colony Algorithm for Multiobjective Disassembly Line Balancing of End-of-Life Products.

Disassembly lines are the most effective way to address large-scale value recovery from end-of-life (EOL) products. Disassembly line balancing (DLB) greatly affects the economics and throughput of EOL product processing. Complete disassembly is generally not suitable for disassembly enterprises; most often, the maximum profit is realized through partial disassembly. Thus, this article proposes a partial disassembly method and establishes a new DLB model that addresses both economic benefits and environmental impacts. The objective of the model is to maximize the effectiveness of workers, increase profit, reduce energy consumption, and balance the loads of workers. Moreover, the model considers the impact of disassembly face and tool changes on the disassembly process. A discrete multiobjective artificial bee colony (MOABC) algorithm is developed, and it takes the precedence constraints into account to obtain the Pareto solutions. The MOABC algorithm is applied to the disassembly lines of two real-world EOL products, including those of an LCD TV and a refrigerator. Experiments show that the performance of the MOABC algorithm is better than those of five well-known multiobjective algorithms. The proposed model and method can provide multiple disassembly schemes for decision makers of disassembly enterprises based on their preferences.

Life Cycle Assessment of Neodymium-Iron-Boron Magnet-to-Magnet Recycling for Electric Vehicle Motors.

Neodymium-iron-boron (NdFeB) magnets offer the strongest magnetic field per unit volume, and thus, are widely used in clean energy applications such as electric vehicle motors. However, rare earth elements (REEs), which are the key materials for creating NdFeB magnets, have been subject to significant supply uncertainty in the past decade. NdFeB magnet-to-magnet recycling has recently emerged as a promising strategy to mitigate this supply risk. This paper assesses the environmental footprint of NdFeB magnet-to-magnet recycling by directly measuring the environmental inputs and outputs from relevant industries and compares the results with production from "virgin" materials, using life cycle assessments. It was found that magnet-to-magnet recycling lowers environmental impacts by 64-96%, depending on the specific impact categories under investigation. With magnet-to-magnet recycling, key processes that contribute 77-95% of the total impacts were identified to be (1) hydrogen mixing and milling (13-52%), (2) sintering and annealing (6-24%), and (3) electroplating (6-75%). The inputs from industrial sphere that play key roles in creating these impacts were electricity (24-93% of the total impact) and nickel (5-75%) for coating. Therefore, alternative energy sources such as wind and hydroelectric power are suggested to further reduce the overall environmental footprint of NdFeB magnet-to-magnet recycling.

Experimental and analytical efforts to characterize cutting fluid mist formation and behavior in machining.

The use of cutting fluids in machining operations is being carefully scrutinized by industry for several reasons, including its overall cost in the manufacturing process and its impact on worker health. Given the concerns associated with the use of cutting fluids, a number of experimental and analytical research efforts are being conducted to gain an understanding of the role of these fluids in various machining processes. The knowledge gained by this research will aid in the development and implementation of strategies to reduce or eliminate the negative effects of cutting fluids, while maintaining their beneficial role. This article presents the results of designed experiments focused on determining the significant variables that influence air quality during turning operations, as well as characterize the aerosol emissions associated with wet and dry turning. Air quality is characterized by measuring the mass concentration and particle size distribution of the dust and mist created during a set of machining experiments. The relative importance of vaporization/condensation and atomization as mist-generating mechanisms is also explored. The experiments revealed that spindle speed has a dominating effect on both mist mass concentration and aerodynamic particle size. Analytical models are presented that predict the average droplet size of the mist generated by atomization and are used to investigate droplet size trends for various cutting fluids and machining parameters. The results predicted by the models are consistent with the expected trends.

Sustainability science and engineering: the emergence of a new metadiscipline.

A case is made for growth of a new metadiscipline of sustainability science and engineering. This new field integrates industrial, social, and environmental processes in a global context. The skills required for this higher level discipline represent a metadisciplinary endeavor, combining information and insights across multiple disciplines and perspectives with the common goal of achieving a desired balance among economic, environmental, and societal objectives. Skills and capabilities that are required to support the new metadiscipline are summarized. Examples of integrative projects are discussed in the areas of sustainability metrics and integration of industrial, societal, and environmental impacts. It is clear that a focus on green engineering that employs pollution prevention and industrial ecology alone are not sufficient to achieve sustainability, because even systems with efficient material and energy use can overwhelm the carrying capacity of a region or lead to other socially unacceptable outcomes. To meet the educational and human resource needs required for this new discipline, the technological and environmental awareness of society must be elevated and a sufficient and diverse pool of human talent must be attracted to this discipline.