《数字化和一次性使用——最终用户的观点.pdf》由会员分享,可在线阅读,更多相关《数字化和一次性使用——最终用户的观点.pdf(15页珍藏版)》请在三个皮匠报告上搜索。
1、Digitalization and Single-Use,an end users perspectiveC.MasyGSK July 11,2023Introduction How can AI&digitalization support use of Single Use in pharma processCase study of process improvement with input of materialCase study in the context of Single Use(SUT)Take awayAGENDACharlotte Masy,Christine De
2、 Herde,Youness Issaf,Patrick Seow,Marine Lepoutre,Etienne Michel,Amalia Trevisan,Carole Garnir are employees of the GSK group of companies.This work is sponsored by GlaxoSmithKline Biologicals SA.These timelines does not fit with pandemy acceleration neededAdvances and Challenges in Vaccine Developm
3、ent and Manufacture by Tony DAmore and Yan-ping Yang 2019 Bioprocessing International+immunological characterizationEVOLUTION OF VACCINES DEVELOPMENTAdvances and Challenges in Vaccine Development and Manufacture by Tony DAmore and Yan-ping Yang 2019 Bioprocessing InternationalIntroduction-What is Di
4、gitalization&Artificial Intelligence?What is Digitalization?Adaptation of a system,process,etc.to be operated with the use of computers and the internet 1 What is AI?Collection of multiple technologies that allow machines to detect,understand,act and learn either on their own or to augment human act
5、ivities.251 Oxford Languages website 2 Accenture Research website ImprovementModelDataIntroduction-How can AI&Digitalization support pharma process?6 There are several area in which AI can be a support:Discovery,Efficiency,Patient and Supply In this presentation:Focus on Discovery and efficiency of
6、processDISCOVERYEFFICIENCY of processSupport to PATIENTSUPPLY improvementAI supportHow can AI&Digitalization support pharma process?Discovery7Smart Cell Culture Systems:Integration of Sensors and Actuators into Microphysiological Systems by Mario M.Modena,Ketki Chawla,Patrick M.Misun,and Andreas Hie
7、rlemann 2018 ACS Chem Biol-13-1767-1784DISCOVERYEFFICIENCY of processSupport to PATIENTSUPPLY improvementAI support Technology that can support:Help analyze disease patterns best treatments Digital Twin including historical data on process designing and optimization of processSingle Use:development
8、of technologies allowing online monitoring(e.g.cell growth in bioreactor-Ultrasonic sensor Ovizio)How can AI support pharma process-Supply improvement8Smart Cell Culture Systems:Integration of Sensors and Actuators into Microphysiological Systems by Mario M.Modena,Ketki Chawla,Patrick M.Misun,and An
9、dreas Hierlemann 2018 ACS Chem Biol-13-1767-1784 What can AI do:Personnalize DiagnosticPredic epidemic outbreakimportance of SU supplyThis Photo by Unknown Author is licensed under CC BYDISCOVERYEFFICIENCY of processSupport to PATIENTSUPPLY improvementAI supportHow can AI/Digitalization support use
10、of SUT in process?-efficiency of process9Smart Cell Culture Systems:Integration of Sensors and Actuators into Microphysiological Systems by Mario M.Modena,Ketki Chawla,Patrick M.Misun,and Andreas Hierlemann 2018 ACS Chem Biol-13-1767-1784DISCOVERYEFFICIENCY of processSupport to PATIENTSUPPLY improve
11、mentAI supportInput materialProductionQC releaseDeliverySpecificationBatch numberCoA(test data)Drawing/BOMSupplier data-complianceEquipementsCPPSensor data(T/P)BR dataTest releasedataA standard process has a lot of data associated but not always available!Traditionally,improvement of process based o
12、n human/SME experienceDigitalization and Continued Process Verification(CPV)10DISCOVERYEFFICIENCY of processSupport to PATIENTSUPPLY improvementAI support Electronic Data Exchange:collect data automatically fromsupplier(CoA)Data are merged with process data and used for data analyticsMonitor raw mat
13、erial performance to control our process performance/variability and prediction Two Proof Of Concepts successfully developed with two suppliersFrom ChristineDe Herde,Youness Issaf,Patrick Seow Input material case study Team:C.De Herde,Y.Issaf,P.SeowInput materialProductionQC releaseDeliveryBatch num
14、berCoA(test data)Process Monitoring&analytics systemeDataconnectivity,securetransmission&data acceptanceSupplier Raw Material Data per ASTM standard-eDataDigitalization and Continued Process Verification(CPV)New way to extract data;Optical Character Recognition(OCR)11DISCOVERYEFFICIENCY of processSu
15、pport to PATIENTSUPPLY improvementAI supportFrom ChristineDe Herde,Youness Issaf,Patrick SeowThe data can be extracted automatically from the PDF copy of the CoA into a structured data format e.g.Excel to be analyzedData is unstructured(PDF file)1 PDF=1 raw material batchStructured data in ExcelDigi
16、talization and process improvementSingle Use Case studyTeam:C.Garnir,A.Trevisan,C.MasyDISCOVERYEFFICIENCY of processSupport to PATIENTSUPPLY improvementAI support Current work:exchange on Single Use drawing and Buid of Material(BOM)with supplier Supplier data exchange(compliance,e-questionnaire.)Fut
17、ure investigation:Machine Learning on based on SUS BOM/genealogy versus process yield?Batch numberCoA(test data)Drawing/BOMSupplier data-complianceInput materialProductionQC releaseDeliveryEquipementsCPPSensor data(T/P)BR dataYield investigationImprovementDigitalization and validation processSingle
18、Use Case study Team M.Lepoutre,E.Michel&C.MasyDISCOVERYEFFICIENCY of processSupport to PATIENTSUPPLY improvementAI support E&L heavy work in validation of Single Use Work:Automatization of E&L calculation per component base ondrawing,BOM,E&L data from supplier Automatic calculation based on process
19、and BOM speed up validation and changeInput materialProductionQC releaseDeliveryDrawing/BOMSupplier data-complianceEquipementsCPPAutomatic generationof Leachable Risk Assessment calculation Digitalization/AI does not necessarily require complex tools/software Digitalization/AI is key for Pharma process(optimization,)SUT Availability of data is a limiting factor need standards,tools,exchange Future is integration of processesTAKE AWAYThis Photo by Unknown Author is licensed under CC BY-SA-NC