@article{zhang_zeng_starly_2021, title={Recurrent neural networks with long term temporal dependencies in machine tool wear diagnosis and prognosis}, volume={3}, ISSN={["2523-3971"]}, DOI={10.1007/s42452-021-04427-5}, abstractNote={Abstract Data-driven approaches for machine tool wear diagnosis and prognosis are gaining attention in the past few years. The goal of our study is to advance the adaptability, flexibility, prediction performance, and prediction horizon for online monitoring and prediction. This paper proposes the use of a recent deep learning method, based on Gated Recurrent Neural Network architecture, including Long Short Term Memory (LSTM), which try to captures long-term dependencies than regular Recurrent Neural Network method for modeling sequential data, and also the mechanism to realize the online diagnosis and prognosis and remaining useful life (RUL) prediction with indirect measurement collected during the manufacturing process. Existing models are usually tool-specific and can hardly be generalized to other scenarios such as for different tools or operating environments. Different from current methods, the proposed model requires no prior knowledge about the system and thus can be generalized to different scenarios and machine tools. With inherent memory units, the proposed model can also capture long-term dependencies while learning from sequential data such as those collected by condition monitoring sensors, which means it can be accommodated to machine tools with varying life and increase the prediction performance. To prove the validity of the proposed approach, we conducted multiple experiments on a milling machine cutting tool and applied the model for online diagnosis and RUL prediction. Without loss of generality, we incorporate a system transition function and system observation function into the neural net and trained it with signal data from a minimally intrusive vibration sensor. The experiment results showed that our LSTM-based model achieved the best overall accuracy among other methods, with a minimal Mean Square Error (MSE) for tool wear prediction and RUL prediction respectively.}, number={4}, journal={SN APPLIED SCIENCES}, author={Zhang, Jianlei and Zeng, Yukun and Starly, Binil}, year={2021}, month={Mar} }
 @article{nordberg_zhang_griffith_frank_starly_loboa_2017, title={Electrical Cell-Substrate Impedance Spectroscopy Can Monitor Age-Grouped Human Adipose Stem Cell Variability During Osteogenic Differentiation}, volume={6}, ISSN={["2157-6580"]}, DOI={10.5966/sctm.2015-0404}, abstractNote={Abstract Human adipose stem cells (hASCs) are an attractive cell source for bone tissue engineering applications. However, a critical issue to be addressed before widespread hASC clinical translation is the dramatic variability in proliferative capacity and osteogenic potential among hASCs isolated from different donors. The goal of this study was to test our hypothesis that electrical cell-substrate impedance spectroscopy (ECIS) could track complex bioimpedance patterns of hASCs throughout proliferation and osteogenic differentiation to better understand and predict variability among hASC populations. Superlots composed of hASCs from young (aged 24–36 years), middle-aged (aged 48–55 years), and elderly (aged 60–81 years) donors were seeded on gold electrode arrays. Complex impedance measurements were taken throughout proliferation and osteogenic differentiation. During osteogenic differentiation, four impedance phases were identified: increase, primary stabilization, drop phase, and secondary stabilization. Matrix deposition was first observed 48–96 hours after the impedance maximum, indicating, for the first time, that ECIS can identify morphological changes that correspond to late-stage osteogenic differentiation. The impedance maximum was observed at day 10.0 in young, day 6.1 in middle-aged, and day 1.3 in elderly hASCs, suggesting that hASCs from younger donors require a longer time to differentiate than do hASCs from older donors, but young hASCs proliferated more and accreted more calcium long-term. This is the first study to use ECIS to predict osteogenic potential of multiple hASC populations and to show that donor age may temporally control onset of osteogenesis. These findings could be critical for development of patient-specific bone tissue engineering and regenerative medicine therapies.}, number={2}, journal={STEM CELLS TRANSLATIONAL MEDICINE}, author={Nordberg, Rachel C. and Zhang, Jianlei and Griffith, Emily H. and Frank, Matthew W. and Starly, Binil and Loboa, Elizabeth G.}, year={2017}, month={Feb}, pages={502–511} }
 @article{zhang_starly_cai_cohen_lee_2017, title={Particle learning in online tool wear diagnosis and prognosis}, volume={28}, ISSN={["1526-6125"]}, DOI={10.1016/j.jmapro.2017.04.012}, abstractNote={Automated Tool condition monitoring is critical in intelligent manufacturing to improve both productivity and sustainability of manufacturing operations. Estimation of tool wear in real-time for critical machining operations can improve part quality and reduce scrap rates. This paper proposes a probabilistic method based on a Particle Learning (PL) approach by building a linear system transition function whose parameters are updated through online in-process observations of the machining process. By applying PL, the method helps to avoid developing a complex closed form formulation for a specific tool wear model. It increases the robustness of the algorithm and reduces the time complexity of computation. The application of the PL approach is tested using experiments performed on a milling machine. We have demonstrated one-step and two-step look ahead tool wear state prediction using online indirect measurements obtained from vibration signals. Additionally, the study also estimates remaining useful life (RUL) of the cutting tool inserts.}, journal={JOURNAL OF MANUFACTURING PROCESSES}, author={Zhang, Jianlei and Starly, Binil and Cai, Yi and Cohen, Paul H. and Lee, Yuan-Shin}, year={2017}, month={Aug}, pages={457–463} }