What to do next now we are working paperless
Contents Personal Info Introduction Lab Data and Systems Overview Next Steps Data Retention Business Intelligence Efficiency Improvement and added value for Customers
Personal Info
Corné Nous, Netherlands
Bachelor in Biochemistry (HLO)
28 Years at MSD/N.V. Organon, Oss, Netherlands
Started as Pharmaceutical Lab Assistant in 1987
8 Years experience in Pre-Clinical Analysis, Instrument Support and Chromatographic Method Development
25 Years in Lab Automation
Deployment Engineer, Technical&Functional Application Management, Project Management, Account Management, Architecture R&D Platforms
Contributor to Allotrope
Now starting own consultancy on Lab Automation Support
[email protected]
Introduction 3-4 decades of Lab Automation Lab automation evolved from data processing/calculation support to integrated systems to support product release, NDA filings and other reporting All data is now available electronically and often electronic only Focus on Pharma Product Development
Systems CDS = Chromatographic Data System
SDMS/ECM = Scientific Data Management System, Enterprise Content Manager EDMS = Electronic Document Management System ELN = Electronic Lab Notebook
LIMS = Laboratory Information Management System LIS = Logistic Information System SCADA = Supervisory Control And Data Acquisition
MES = Manufacturing Execution System Inventory Systems
Lab Data Storage & Processes Statistic Tools
EDMS
Query Tools
LIS
LIMS
Web Tools
MES
ELN CDS
SCADA
SDMS ECM
The Lab Data Life Cycle
Prepare for Analysis
Set-Up Instrument
Acquire & Process
Review & Approve
Report
Archive
Data Retention GxP, Legal Hold and other Regulations demand data storage for many years depending on the goal for which the data was generated Pharma companies apply 20-30 years or longer retention periods The originating systems have a lifecycle of 5-10 years Convert data to open standards to keep raw data files readable (AnIML, MzML, JCAMP-DX…) Data viewers available in SDMS/ECM systems, Open Source software or commercial software Viewers for Data in Standardized format and for Proprietary data formats Virtualization of instrument workstations/clients ensures full functionality (reprocess) Pharma take overs
Virtualization Process VMWare provides tools to image the workstation and make it available as a virtual machine This creates an OS independent solution to open the data in the proprietary format Ensure transfer of license keys to VM image (Microoft and proprietary software) Store image on media that is readable for many years
Other vendors? (e.g. Oracle Virtual Box)
Decission & Knowledge support Search on Structure, Substructure and Spectra to identify new impurities and degradation products Use chemistry and physics data to predict product stability Ensure Data is consistent: Meta Data definitions Data Stewards guard consequent use of meta data
Restrictions and in-efficiencies in the current environments Data are spread over several systems
Search or analytics are performed in a step-by-step approach Querying is slow Data is structured (Databases)/ non-structured (ELN/EDMS) Different Query tools apply Large amounts of data are stored on expensive SAN solutions Data structures end Meta data need to be re-defined if new applications of the data are required
Big Data approach Big Data approach can help in improving efficiency in data searches and analytics Combine company data (structured and non-structured) in one Data Lake and add external data (web) Search and perform analytics on the combined data
Big Data application on Lab Data Interact, Visualize
Analytics, Statistics, Data Mining
Data Lake
Structured & Non-Structured Data Lab Data Archive
Chemistry Databases
Spectral Libraries
External Scientific Sources
Structure Eludation
Structure Eludation Search Tools
Chemistry Databases
Spectral Libraries
Use data for process optimization Optimize Resource Planning
Spread resource utilization Retire surplus of instruments Plan Calibrations and Maintenance Group samples
Determine run time, costs and earnings per analysis Provide added value to customers Determine Trends Warn for overdue sample delivery dates
Summary In the 3-4 decades of Lab Automation data use has been transferred from data processing support to fully supported Lab processes. Data Archival and Retrieval have become mandatory requirements in respect to Regulations Special attention is required to keep raw-data files viewable Using Big Data technologies in combining data with other repositories, libraries and external content provide new capabilities and improved R&D productivity Data can be used to improve Lab processes