Cha Yong Jong, Akshay Mittal, Geordi Tristan, Vanessa Noller, Hui Ling Chan, Yongkai Goh, Eunice Wan Qi Yeap, Srinivas Reddy Dubbaka, Harsha Rao Nagesh and Shin Yee Wong*,
{"title":"ANFIS-Driven Machine Learning Automated Platform for Cooling Crystallization Process Development","authors":"Cha Yong Jong, Akshay Mittal, Geordi Tristan, Vanessa Noller, Hui Ling Chan, Yongkai Goh, Eunice Wan Qi Yeap, Srinivas Reddy Dubbaka, Harsha Rao Nagesh and Shin Yee Wong*, ","doi":"10.1021/acs.oprd.3c00505","DOIUrl":null,"url":null,"abstract":"<p >Manual crystallization trials have historically posed significant challenges, demanding substantial expertise for process development and often offering unpredictable outcomes. This study addresses these difficulties by introducing an automated system that alleviates the need for manual iterations and intuitive deductions. The system leverages machine learning algorithms capable of learning from high-quality data to discern patterns and recommend optimal actions for subsequent runs. The automation process commences with a direct chord length (DCL) control system, generating system-specific training data via universal crystallization rules. After that, the automation process will progress into a machine learning iteration loop using adaptive neuro-fuzzy inference system (ANFIS) models. In this iteration loop, multiple models will be built (with accumulative historical data) and deployed to the crystallization process until predefined exit criteria are met or a maximum of five iterative cycles are reached. Results from the two campaigns are presented. It is evident that the automated crystallization platform with machine learning’s ability can confidently explore the operational space, proposing credible processing conditions that yield desirable process outcomes.</p>","PeriodicalId":55,"journal":{"name":"Organic Process Research & Development","volume":null,"pages":null},"PeriodicalIF":3.1000,"publicationDate":"2024-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Organic Process Research & Development","FirstCategoryId":"92","ListUrlMain":"https://pubs.acs.org/doi/10.1021/acs.oprd.3c00505","RegionNum":3,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CHEMISTRY, APPLIED","Score":null,"Total":0}
引用次数: 0
Abstract
Manual crystallization trials have historically posed significant challenges, demanding substantial expertise for process development and often offering unpredictable outcomes. This study addresses these difficulties by introducing an automated system that alleviates the need for manual iterations and intuitive deductions. The system leverages machine learning algorithms capable of learning from high-quality data to discern patterns and recommend optimal actions for subsequent runs. The automation process commences with a direct chord length (DCL) control system, generating system-specific training data via universal crystallization rules. After that, the automation process will progress into a machine learning iteration loop using adaptive neuro-fuzzy inference system (ANFIS) models. In this iteration loop, multiple models will be built (with accumulative historical data) and deployed to the crystallization process until predefined exit criteria are met or a maximum of five iterative cycles are reached. Results from the two campaigns are presented. It is evident that the automated crystallization platform with machine learning’s ability can confidently explore the operational space, proposing credible processing conditions that yield desirable process outcomes.
期刊介绍:
The journal Organic Process Research & Development serves as a communication tool between industrial chemists and chemists working in universities and research institutes. As such, it reports original work from the broad field of industrial process chemistry but also presents academic results that are relevant, or potentially relevant, to industrial applications. Process chemistry is the science that enables the safe, environmentally benign and ultimately economical manufacturing of organic compounds that are required in larger amounts to help address the needs of society. Consequently, the Journal encompasses every aspect of organic chemistry, including all aspects of catalysis, synthetic methodology development and synthetic strategy exploration, but also includes aspects from analytical and solid-state chemistry and chemical engineering, such as work-up tools,process safety, or flow-chemistry. The goal of development and optimization of chemical reactions and processes is their transfer to a larger scale; original work describing such studies and the actual implementation on scale is highly relevant to the journal. However, studies on new developments from either industry, research institutes or academia that have not yet been demonstrated on scale, but where an industrial utility can be expected and where the study has addressed important prerequisites for a scale-up and has given confidence into the reliability and practicality of the chemistry, also serve the mission of OPR&D as a communication tool between the different contributors to the field.