technology innovations driving transformation in regulatory · validated data - pharmaceuticals...
TRANSCRIPT
Copyright ©2020 Version 1. All rights reserved.1
“The Regulatory Analytics: Stories of Technology Innovations Driving Transformation in Regulatory
Compliance”
January 29th | Republic of Work, Cork
Copyright ©2019 Version 1. All rights reserved.2
Lorcan Malone | Chief Executive
INSPIRE EVENT with Wednesday, 29th January 2020
Copyright ©2019 Version 1. All rights reserved.3
Questions – Slido.com #Inspire2020
Alan Crowley – Head of Delivery
Version 1
Copyright ©2020 Version 1. All rights reserved.4
Speakers
Karl O’ConnellHead on Integration
ICBF
Michael PhelanBusiness Technology Leader in Data Science
Johnson & Johnson
David Hawe Data Scientist
Cork Institute of Technology
Copyright ©2019 Version 1. All rights reserved.5
Decision Science
Michael Phelan – Business
Technology Leader in Data Science
Johnson & Johnson
CONFIDENTIAL AND PROPRIETARY
SUPPLY CHAIN
BUSINESS
TECHNOLOGY
Labels, Flow & DNA
Observations from a Regulated Industry
LABELS
7
DECISION SCIENCE
AI (ML / NN / RL)BI / DashboardsSoftware Eng / APIs
Developing intuitive
Decision Support Solutions
helping decision makers take
Action from Data driven Insights
many career
opportunities
in AI
Data
Scientist
Data
Engineer
Data
Architect
Business
Analyst
Software
Engineer
Solutions
Architect
Citizen
Data
Scientist
B u s i n e s s
O p p o r t u n i t y
AI LABELS
FLOW
SCBT – Research Design & Engineering
DECISION FLOW - Solution Components
11
DATA
ETL, Storage
COMPUTE
Cognitive Services,
ML Models, AI, Simulation
CONSUME
API, RPA, Visualise
SCBT – Research Design & Engineering
DECISION FLOW - Customer Experience
12
RA
W T
EX
T
SQL
noSQL
Translate to English
Intent, Entity &
Sentiment Analysis
NLP
pre-processing
ML Classification
Phone
Fax
Move to folder /
label email
Dashboard
RPA
Speech
to Text
Analogue
to Digital
Attachments
Data is unlike any other
asset
It never depletes
It never wears out
Same data used across unlimited
use-cases at near zero marginal cost
Bill Schmarzo, CTO Hitachi Vantara
SCBT – Research Design & Engineering
Adopting an AGILE APP-Based Approach
14
PARTNER
Active customer
engagement & ownership
PRODUCTION
• Asking the right question?
• Quality impacts (CSV)?
• Business support / ROI
QUESTION
• PoC App based solution
(1 Question = 1 App)
• Limited data sources
• no automation or integration
• 3-4 sprints
SCBT – Research Design & Engineering
Agile Decision ScienceWhat is the Question?
Failure is good!
What’s the business
impact?
DNA
DNA
SCBT – Research Design & Engineering
What is a Data Attribute?
Example: 1 Excel Workbook = 1 ERP
17
• Worksheets = ERP tables – (e.g., vbak, vbup, vbkd, vbuk, vbap etc.)
• Columns of each sheet = EPR Attributes– (e.g., abdis, abhob, abhod, abhov, etc.)
• Rows of data = potential information!
• Identify the data attributes that explain an organisation!
Tables
Attributes
SCBT – Research Design & Engineering
FLOW and REUSE of an Attribute
18
Example: data initially collected for ERP (L4) & MES (L3) systems
PROFILE: type of data, dispersion, ranges, categorical etc.
BI: individual or blended attributes (rule based business logic)
AI: model features – individual or synthetically blended combinations
Curation: NLP semantic mapping with internal / external data sources
Explore non-linear relationships (e.g. MES product flow in Graph DB’s)RPA
vbak.aedat
vbak.abdis
Ensure most
important attributes
are always captured
accurately
SCBT – Research Design & Engineering
Active customer engagement
& collaboration
What is the Question?
Solutions Focused
App – Based, Agile Releases
Create Reusable Flows
Copyright ©2019 Version 1. All rights reserved.22
Management of Validated Data
David Hawe – Data Scientist
Cork Institute of Technology
mathematics.cit.ie
Management of Validated Data
David Hawe [email protected]
Department of Mathematics @CIT
• .
Data Lifecycle
Data Lifecycle
Data Lifecycle
Data Lifecycle
“Data validation … ensures the correspondence of the final (published) data with a number of quality characteristics.”
(Simon 2013)
Validated Data - Finance
“Data validation … ensures the correspondence of the final (published) data with a number of quality characteristics.”
(Simon 2013)
• Some Examples:• Name: \18• Date of Birth: 14/01/1888• Salutation and gender
Validated Data - Finance
https://govzilla.com/blog/2019/05/pharma-medical-devices-data-integrity-breaking-down-keywords-and-citation-trends-from-the-fda/
Validated Data - Pharmaceuticals
Validated Data - Pharmaceuticals
“without complete, accurate, reliable, or retrievable raw data about the HPLC system’s qualification, you lacked complete assurance that the system was operating as intended” (FDA issued 483 in 2015)
ScientificData
MgmtSystem
Data Lake/Warehouse
LaboratoryInformationMgmtSystem
Standalone Instruments
Electronic Laboratory Notebooks
Distributed Control
System/ Historian
Enterprise Resource Planning
Paper/Paper on
glass
Other Systems
The boring stuff:• Accurate - Legible - Contemporaneous - Original
Attributable - Complete - Consistent - Enduring Available when needed
• Data in systems should be at least as accurate as those recorded by paper means.
• Reviewed & Verified• When a mistake is realised correct the recorded
data. This is traceable in audit logs.• Backup/Archive
Validated Data - Pharmaceuticals
• Interestingly, there is massive opportunities for streamlining reporting and analysis in terms of how systems are used.
Validated Data - Pharmaceuticals
} Reports &Analysis(Risk Based)
Future Strategy
ScientificData
MgmtSystem
Data Lake/Warehouse
LaboratoryInformationMgmtSystem
Standalone Instruments
Electronic Laboratory Notebooks
Distributed Control
System/ Historian
Enterprise Resource Planning
Paper/Paper on
glass
Other Systems
• Quality data for ML and AI • Static models for subsequent real-time use.
• Analysis• Domain expertise• Documentation (ensure reproducibility)• Resources• Auditing
• Inappropriate samples• Gender bias
Validated Data - Analysis
• Deviations/Investigations• Annual Reviews/Continuous Validation• Process Analytical Technology – the unused data• Never miss a case
Uses of Validated Data
• MSc in Data Science and Analytics• HDip in Data Science and Analytics• MSc in Artificial Intelligence• MSc in Cybersecurity• MSc in Cloud Computing• Certificate in Biopharmaceutical Processing • Higher Diploma in Science in Cloud Computing• Certificate in Automation & Control Systems• Certificate in SCADA & Automation Systems• Certificate in Industrial Automation• Certificate in Intelligent Manufacturing Systems
Relevant Programs @CIT
[1] Guideline on process validation for finished products - information and data to be provided in regulatory submissions. European Medicines Agency
Report, 21 November 2016, EMA/ CHMP/CVMP/QWP/BWP/70278/2012-Rev1,Corr.1
[2] Methodology for data validation 1.0 Revised edition June 2016 EssnetValidat Foundation Marco Di Zio, Nadežda Fursova, Tjalling Gelsema, Sarah Gießing, Ugo Guarnera, Jūratė Petrauskienė, Lucas Quensel-von Kalben, Mauro Scanu, K.O. ten Bosch, Mark van der Loo, Katrin Walsdorfer [3] UNECE 2013 Glossary of terms on statistical data editing
http://www1.unece.org/stat/platform/display/kbase/Glossary [4] Simon A., (2013a) Definition of validation levels and other related
concepts v01307. Working document. Available at https://webgate.ec.europa.eu/fpfis/mwikis/essvalidserv/images/3/30/Eurostat_-_definition_validation_levels_and_other_related_concepts_v01307.doc
[5] 09 June 2010 EMA/INS/GCP/454280/2010 GCP Inspectors Working Group (GCP IWG) Reflection paper on expectations for electronic source data and data transcribed to electronic data collection tools in clinical trials
References
Copyright ©2019 Version 1. All rights reserved.38
Blockchain
Karl O’Connell – Head of Integration
The Irish Cattle Breeding Federation
Reducing Carbon Emissionsand
Finding the Tastiest Steak
AgTech-It’s in our DNA
Our Vision“Empowering sustainable farming through collaboration and
excellence in genetics and big data solutions”
AgTech-It’s in our DNA
About ICBF
Next generation of animals are more environmentally and economically sustainable than the last.
World-leading (research => implementation).
• 2nd in world to launch dairy genomics.
• Beef Genomics => largest cattle genomics project globally.
Carbon EfficientGenetic Gain
Sustainable Farming
AgTech-It’s in our DNA
Overview• Grass based seasonal
• 0.95m beef cows
• 66,000 herds
• Avg herd size: 14 cows
• Export 90% beef
• Grass based seasonal
• 1.5m dairy cows
• 17,000 herds
• Avg herd size: 80 cows
• Export 90% of milk produced
Dairy Beef
AgTech-It’s in our DNA
Yearly Data Figures
2.4 Million
BVD Records
102kHerds
7.7 MillionFarm Movements
850kAI Records
1Oracle Exadata Platform
6.5 MillionLive Animals
2.2 Million
Slaughter Records
2.4 Million
Animal Births
4.8 Million
Milk Recording
Records
AgTech-It’s in our DNA
Genomics
Non-Genomic Index
Genomic Index
AgTech-It’s in our DNA
Genotype Data SetsBovine Genome
• Contains 3 billion nucleotides
• Scientists have discovered 40 million SNPs
3 billion rows of data per animal
Ireland Genotypes
380k animals per year on a 54k SNP Chip
First full genome sequence cost $300 million
Today costs = $1,000
AgTech-It’s in our DNA
Genotype Data Sets
1,800,000,000
8,600,000,000
18,300,000,000
62,700,000,000 95,000,000,
000
Start Jan-15 Mar-16 Mar-18 Aug-19
Genotype Records
AgTech-It’s in our DNA
Farm of the FutureIoT & Big Data
• IoT devices in agriculture will reach 75M in 2020
• Growing 20% annually
Geofencing & tracking
• Smart sensors
• Smart tractors
Blockchain
• Traceability
• Consumer Feedback
Climate & Environment
• Low carbon cows
• Sustainable farming
AgTech-It’s in our DNA
Breeding for Tasty Steak
Important to the consumer
• Tenderness
• Juiciness
• Taste
• Traceability
• Environmental Impact
AgTech-It’s in our DNA
Breeding for Tasty Steak (Meat Eating Quality)
• Sensory testing with trained panellists to generate breeding values.
• Partnerships with Eolas, Teagasc, Industry through MTI.
• Already being used in a “structured way” by Meat & Livestock Australia (MSA grading).
• Can we use consumer data in a more “unstructured” way?
Can consumer pick the difference?How does the consumer compare with trained panellist?
AgTech: It’s in our DNA
Our partnership with ICBF
IT Delivery Partner for 15 years
Considered an extension of the in-house team
Main Duties:
✓ Install & Support of core Oracle Database (Exadata / ZFS)
✓ Cloud service configuration and management
✓ Bespoke system and app development
✓ Provide support and advice across breadth of our services
AgTech: It’s in our DNA
How does Blockchain impact ICBF?
Enable Farm to Fork traceability using Blockchain technology to include
additional parties involved in the supply chain.
Consume
r
ICBF Industry▪ Enables surgical like recalls
▪ Further enhances Ireland’s already stellar agri image, adding value to the industry as a whole
▪ In line with strategic mission
▪ Enabler of current programs:-> Meat tasting => crowd
source!-> GHG Emissions = increased
consumer awareness
▪ Increased understanding, visibility, trust & transparency
▪ Informed purchase decision (buy local / buy green / buy tasty.
AgTech: It’s in our DNA
Is there a precedent?
Wyoming Beef Chain: https://beefchain.com
AgTech: It’s in our DNA
Project Scope & Objectives
Scope:
• Technical Proof of Value limited to:
• Single participant (ICBF)
• Mocked external participants
Objectives:
• Help ICBF better understand Blockchain technology
• Technical understanding
• Business impact understanding
• Industry Engagement Tool
Run by Version 1 Innovation Labs
AgTech: It’s in our DNA
Solution Overview
API Gateway
Consumer
ICBF
Future Participants
• Standards
• Suppliers
• Distributors
• Supermarkets
Mobile App Blockchain Network
AgTech: It’s in our DNA
Mobile App
AgTech: It’s in our DNA
Outcome• “Proof of Value” (PoV) solution developed
• Solution provides a real understanding of what Blockchain means to ICBF
• Blockchain demonstrates real value in Supply Chain use cases
• Solution attracted international attention and was presented at OOW 2019 in San Francisco and
featured on Forbes.com
Next Steps
• PoV being extended and implemented initially within ICBF
• Industry conversations ongoing
Thank you for listening
#WeBreedInnovation