In the Industry 4.0 era, the chemical industry is embracing the broad adoption of Artificial Intelligence (AI) and Machine Learning (ML) methods. Machine learning techniques have been used in drug discovery, molecular property prediction, catalyst inverse design, etc. Material science research currently focuses on mixtures of molecules such as polymers, detergents and personal care products. Properties of mixed molecules are challenging to predict using experimental and theoretical techniques developed for pure molecules. For example, the rheology of mixed molecules is not equal to the averaged effect of each molecule. Complex materials bring new dimensions to the table. It is too expensive to test all the possible combinations experimentally or using ab initio theories. Machine learning combined with fundamental science could provide a guide to better design experiments, use limited experimental data to build fundamental models or data-driven models.
The chemical industry has a long history of recording manufacturing data. How to best use the data and run future production more efficiently with fewer disruptions is the billion-dollar research question today. Machine learning and related techniques play an important role here. Manufacturing data include structured data such as sensor data (time series in fixed frequency) and operation control settings, as well as unstructured data such as documented trouble-shooting or related events. Use of all this information requires the design of a machine learning system that can deal with text and numeric data, images, time series, and physics models.
Lastly, forecasting plays an important role in today’s supply-chain optimization. Knowing when and where demands will change in advance, and getting prepared for rare but severe situations is critical to manage risks and maximize return. In recent years, black swan incidents became more frequent. Supply-chain needs to quickly adapt to many unprecedented changes. Adaptive forecasting will play a more important role today.
Forecasting in the supply chain includes methods and systems to predict customer’s behavior. For example, what will drive customers to fill up or empty their inventories, how will holiday shopping spikes affect the demand for raw material. Forecasting is also related to how to use real-time information to make quick adjustments. For instance, will a disaster or severe weather in one location slow down the production or shipment? With today’s emerging techniques in artificial intelligence and industry 4.0, forecasting-related research also takes advantage of machine learning and big data to bring new possibilities to supply-chain optimization.
This Research Topic aims to highlight state-of-the-art AI research in R&D, Manufacturing, and Supply Chain. The overarching theme is to highlight how these functions are collaborating to speed up the product development cycle and how the industry is operating towards breakthrough performances in safety, reliability, and sustainability.
We invite the submission of Original Research, Review, Mini Review, Perspective articles on themes including, but not limited to:
• Integration of R&D research into scale-up and operations in Manufacturing
• Integration of Manufacturing operations into Supply Chain planning
• Property prediction for complex materials using machine learning approaches
• Machine learning combined with fundamental models in new materials discovery
• AI-guided design of experiments in material development
• Inverse-modeling for material design
• Recent advances in predictive maintenance
• Recent advances in fault detection and diagnosis
• Machine learning applications of diverse data streams (such as text, images, videos) in Manufacturing – design of optimal input features
• Hybrid modeling applications in Manufacturing
• Robust logistics system to help balance supply-demand
• AI-powered forecasting research in different levels of global supply-chain
• Text mining in supply-chain forecasting.
Topic Editors Dr. Bo Shuang and Dr. Leo Chiang are affiliated with the company "Dow Chemical Company". The other Topic Editors declare no conflict of interest.
In the Industry 4.0 era, the chemical industry is embracing the broad adoption of Artificial Intelligence (AI) and Machine Learning (ML) methods. Machine learning techniques have been used in drug discovery, molecular property prediction, catalyst inverse design, etc. Material science research currently focuses on mixtures of molecules such as polymers, detergents and personal care products. Properties of mixed molecules are challenging to predict using experimental and theoretical techniques developed for pure molecules. For example, the rheology of mixed molecules is not equal to the averaged effect of each molecule. Complex materials bring new dimensions to the table. It is too expensive to test all the possible combinations experimentally or using ab initio theories. Machine learning combined with fundamental science could provide a guide to better design experiments, use limited experimental data to build fundamental models or data-driven models.
The chemical industry has a long history of recording manufacturing data. How to best use the data and run future production more efficiently with fewer disruptions is the billion-dollar research question today. Machine learning and related techniques play an important role here. Manufacturing data include structured data such as sensor data (time series in fixed frequency) and operation control settings, as well as unstructured data such as documented trouble-shooting or related events. Use of all this information requires the design of a machine learning system that can deal with text and numeric data, images, time series, and physics models.
Lastly, forecasting plays an important role in today’s supply-chain optimization. Knowing when and where demands will change in advance, and getting prepared for rare but severe situations is critical to manage risks and maximize return. In recent years, black swan incidents became more frequent. Supply-chain needs to quickly adapt to many unprecedented changes. Adaptive forecasting will play a more important role today.
Forecasting in the supply chain includes methods and systems to predict customer’s behavior. For example, what will drive customers to fill up or empty their inventories, how will holiday shopping spikes affect the demand for raw material. Forecasting is also related to how to use real-time information to make quick adjustments. For instance, will a disaster or severe weather in one location slow down the production or shipment? With today’s emerging techniques in artificial intelligence and industry 4.0, forecasting-related research also takes advantage of machine learning and big data to bring new possibilities to supply-chain optimization.
This Research Topic aims to highlight state-of-the-art AI research in R&D, Manufacturing, and Supply Chain. The overarching theme is to highlight how these functions are collaborating to speed up the product development cycle and how the industry is operating towards breakthrough performances in safety, reliability, and sustainability.
We invite the submission of Original Research, Review, Mini Review, Perspective articles on themes including, but not limited to:
• Integration of R&D research into scale-up and operations in Manufacturing
• Integration of Manufacturing operations into Supply Chain planning
• Property prediction for complex materials using machine learning approaches
• Machine learning combined with fundamental models in new materials discovery
• AI-guided design of experiments in material development
• Inverse-modeling for material design
• Recent advances in predictive maintenance
• Recent advances in fault detection and diagnosis
• Machine learning applications of diverse data streams (such as text, images, videos) in Manufacturing – design of optimal input features
• Hybrid modeling applications in Manufacturing
• Robust logistics system to help balance supply-demand
• AI-powered forecasting research in different levels of global supply-chain
• Text mining in supply-chain forecasting.
Topic Editors Dr. Bo Shuang and Dr. Leo Chiang are affiliated with the company "Dow Chemical Company". The other Topic Editors declare no conflict of interest.