With the emergence of Industry 4.0, also known as the fourth industrial revolution, an increasing number of hardware and software sensors have been implemented in chemical production processes for monitoring key variables related to product quality and process safety. The accuracy of individual sensors can be easily impaired by a variety of factors. To improve process monitoring accuracy and reliability, a sensor fusion scheme based on Bayesian inference is proposed. The proposed method is capable of combining multi-rate sensor data and eliminating the spurious signals. The efficacy of the method has been verified using a process implemented at the Dow Chemical Company. The sensor fusion approach has improved the process monitoring reliability, quantified by the rates of correctly identified impurity alarms, as compared to the case of using an individual sensor.

Due to the complicated metabolism of mammalian cells, the corresponding dynamic mathematical models usually consist of large sets of differential and algebraic equations with a large number of parameters to be estimated. On the other hand, the measured data for estimating the model parameters are limited. Consequently, the parameter estimates may converge to a local minimum far from the optimal ones, especially when the initial guesses of the parameter values are poor. The methodology presented in this paper provides a systematic way for estimating parameters sequentially that generates better initial guesses for parameter estimation and improves the accuracy of the obtained metabolic model. The model parameters are first classified into four subsets of decreasing importance, based on the sensitivity of the model’s predictions on the parameters’ assumed values. The parameters in the most sensitive subset, typically a small fraction of the total, are estimated first. When estimating the remaining parameters with next most sensitive subset, the subsets of parameters with higher sensitivities are estimated again using their previously obtained optimal values as the initial guesses. The power of this sequential estimation approach is illustrated through a case study on the estimation of parameters in a dynamic model of CHO cell metabolism in fed-batch culture. We show that the sequential parameter estimation approach improves model accuracy and that using limited data to estimate low-sensitivity parameters can worsen model performance.

Due to environmental regulations continually reducing emission quantity allowed over time, there is a growing need for adaptable and feasible environmental monitoring, such as emission, wastewater quality, and air pollution monitoring, for the process industry (and surrounding communities). Alternative environmental monitoring and process monitoring technologies based on industrial internet of things (IIoT) and artificial intelligence (AI) enable the process industry to take a proactive approach toward the environment and asset integrity management. The monitoring devices can be deployed in a stationary or dynamic manner. In this study, the emerging trend and various applications of IIoT and advanced data analytics methodologies in environmental monitoring are reviewed. An example showing challenges and research needs in sensor placement is given. Future directions in technology, regulation, and application have been discussed as well.