© 2009 IBM Corporation
IBM Research
Food Security and Safety: Opportunities within the Advanced Technology Sector of IndustryGovernment-University-Industry Research Roundtable Meeting, February 4, 2009
Mary E. Helander, Ph.D.IBM T.J. Watson Research CenterBusiness Analytics and Mathematical Science DepartmentYorktown Heights, NY
2
IBM Research
© 2009 IBM Corporation
Background: IBM Research -3000 Researchers in 8 labs, 14 years of Patent Leadership
ChemistryComputer Science
ElectricalEngineering
Materials Science
Mathematical Sciences Physics
Service Science, Management &
Engineering
Behavioral Sciences
BusinessInnovation
TechnologyInnovation
Social Innovation
Demand Innovation
Science & Engineering
Business & Management
Social & Cognitive Sciences
Economics & Markets
IBM Research
Software Group
Systems & Technology
Group
Operations, Technology
Sales & Distribution
Global Services
Academia & Government
Clients, Alliances
Scientific / Technical
Community
3
IBM Research
© 2009 IBM Corporation
Food Security and Safety: Opportunities within the Advanced Technology Sector of Industry
Abstract
§ Are there some specific opportunities within the Advanced
Technology sector of Industry that can speed development of
solutions in Food Safety and Security? This is the question to be
addressed in this session.
4
IBM Research
© 2009 IBM Corporation
Food Security and Safety: Opportunities within the Advanced Technology Sector of Industry
Specific Opportunities
1. Supply chain management
2. Advanced analytics
3. IT infrastructure
5
IBM Research
© 2009 IBM Corporation
Food Security and Safety: Opportunities within the Advanced Technology Sector of Industry
Specific Opportunities
1. Supply chain management
2. Advanced analytics
3. IT infrastructure
6
Advanced supply chains allow the ability to track and trace entities through a multi-enterprise supply chain. Supply chains for food present unique challenges
| Complex networks of trading partners, including global sourcing
| Heterogeneous technological capabilities
| Non-homogeneous data
| Non-digital, incomplete, or unreliable data
| Disparate data sources
| Benefits not gained by trading partners who incur costs
| Difficult governance
u
u
v
v
w
w
x
x
y
y
z
{
|
7
DISTRIBUTION
§ Retailer Metro AG - RFID system for tracking consumer products from production through transport & warehousing to
sale and customer service
GOVERNMENT
§ UK Government Department for the Environment, Food & Rural Affairs – building Systems for Animal Movement and
Traceability using RFID and GPS
§ Japan - Waste Disposal Traceability Pilot
§ Government of Thailand - RFID pilot on Shrimp Traceability
AUTOMOTIVE
§ Honda - Infrastructure for Data/Process Integration including Traceability
§ Japan - Automotive Parts Traceability prototype
AGRICULTURE
§ Maple Leaf Foods - DNA traceability pilot for pork
§ Major beef farm & processor in US Midwest - RFID system for farm to fork traceability to meet new state regulations
and create a premium brand positioning
PHARMACEUTICAL
§ Japanese Ministry of Internal Affairs and Communications - Pharmaceutical Traceability
Learning from other industries can help speed solutions for food safety and security. These are a few examples of traceability projects for public and private sector clients across many different industries.
8
Gaining full compliance to a voluntary end-to–end traceability system depends on identifying the benefits for each participant in the Supply Chain
Technological Infrastructure & Common Business Processes
Sector/Commodity Specific Business Processes
Food Continuum/Supply ChainProducer/Grower
• Increased farm efficiency
• Individual animal/product value-added information from processor
• Increased yields –business analytics from feed, pesticides, processor
• Increased and secure access to global markets
• Risk mitigation and reduced liability
Processor
• Increased service offerings to clients
• More detailed understanding of input and throughput by client
• Increased quality control
• Risk mitigation and reduced liability
Distributor
• Increased productivity
• Improved inventory
• Improved shipping/ receiving accuracy
• Demand visibility and forecasting
• Decreased diversion expenses
• Risk mitigation and reduced liability
Retailer
• Increased productivity/supply chain optimization
• Improved inventory
• Improved shipping/ receiving accuracy
• Demand visibility and forecasting
• Refined client behaviour information
• More efficient marketing
• Risk mitigation and reduced liability
Customer
• Increased confidence in food supply
• Lower prices
• Greater availability of products and services
Secure Transportation/Secure Trade Lanes
Government
• Improved public safety
• Increased competitiveness
• International trade
• Risk mitigation
• Reduced compensation
Infrastructure• Interoperable
• Globally compliant
• Flexible/ scaleable
Demand/Information Flow
Farm Input/Supplier
• Increased productivity/supply chain optimization
• Improved inventory
• Improved shipping/ receiving accuracy
• Demand visibility and forecasting
• Refined client behaviour information
• More efficient marketing
• Risk mitigation and reduced liability
9
Various technologies exist, e.g. sensors and actuators for data capture. Enabling efficient and effective food safety and security solutions is still a logistical challenge
FARMFARM Slaughterhouse/Processor
Slaughterhouse/Processor
MarketsMarkets
Local Database
Local Database
Local Database
Central Repository Database
Qu
ery
Animal Testing
Animal Testing
Local Database
Query Query
Local Database
FARMFARM Slaughterhouse/Processor
Slaughterhouse/Processor
MarketsMarkets
Local Database
Local Database
Local Database
Central Repository Database
Qu
ery
Animal Testing
Animal Testing
Local Database
Animal Testing
Animal Testing
Local Database
Query Query
Local Database
serial comms
Bar Code reader
serial comms
Internet IP
J2SE
Bar Code tag
<Animal Location Record><Land Parcel Bar Code/><Gov Gateway Credentials/><Ear Tag Bar Code/><Date and Time/>
</Animal Location>
<Animal Location Record><Land Parcel Bar Code/><Land Parcel Geometry/><Gov Gateway Credentials/><Customer ID/><Ear Tag Bar Code/><Animal ID/><Date and Time/>
</Animal Location>
Livestock Register Services
Enterprise Service Bus
Corporate Repositories
internet
Field 1
Field 2
Field 3
Personalised Barcode Pamphlet serial comms
Bar Code reader
serial comms
Internet IP
J2SE
Bar Code tag
<Animal Location Record><Land Parcel Bar Code/><Gov Gateway Credentials/><Ear Tag Bar Code/><Date and Time/>
</Animal Location>
<Animal Location Record><Land Parcel Bar Code/><Land Parcel Geometry/><Gov Gateway Credentials/><Customer ID/><Ear Tag Bar Code/><Animal ID/><Date and Time/>
</Animal Location>
Livestock Register Services
Enterprise Service Bus
Corporate Repositories
internet
Field 1
Field 2
Field 3
Personalised Barcode Pamphlet
Bluetooth
GPS unit
RFID reader
Bluetooth
GPRS
J2ME
RFID tag
<Animal Location Record><Spatial Position/><SIM Card ID/><RFID Tag ID/><Date and Time/>
</Animal Location>
<Animal Location Record><Spatial Position/><Land Parcel Geometry/><SIM Card ID/><Customer ID/><RFID Tag ID/><Animal ID/><Date and Time/>
</Animal Location>
Livestock Register Services
Enterprise Service Bus
Corporate Repositories
internet
Bluetooth
GPS unit
RFID reader
Bluetooth
GPRS
J2ME
RFID tag
<Animal Location Record><Spatial Position/><SIM Card ID/><RFID Tag ID/><Date and Time/>
</Animal Location>
<Animal Location Record><Spatial Position/><Land Parcel Geometry/><SIM Card ID/><Customer ID/><RFID Tag ID/><Animal ID/><Date and Time/>
</Animal Location>
Livestock Register Services
Enterprise Service Bus
Corporate Repositories
internet Stroke Reco ServerStroke Reco Server
Animal verification via RFID taggingBarcode based near real time animal movement data capture
GPS based real time animal movement data capture Digital Pens and Workplace Forms used in traditionally low tech environments
| 10 © Copyright IBM Corporation 2009
Data security maintained via encryption, restricted password access, etc…
Secure, standards-based queries enabled
A track and trace system that supports food safety and security must capture, structure and integrate data on product a) movements, b) attribute changes, and c)processing activities from across and within the supply chain
Example: Beef - Each company maintains its own product information and record of transactions, making that information available on a permission basis to stakeholders
Rapid communication of essential data facilitated through open-standard software
and adoption of industry ID standards
Antibiotics
Logistics Logistics
Fertilizers Packaging Ingredients
Logistics
Ingredients
Logistics
Data Data Data Data DataData
Virtual Traceability System
Logistics Logistics Logistics LogisticsLogistics
Grocery Store/
RestaurantCorn Farmer
Cattle Rancher
Distribution Center
Beef Processor
CP Manufacturer
Transaction & Historical Data
FirewallFirewallFirewallFirewall
FirewallFirewall
Fir
ew
all
Track and trace products and risks within the four walls to isolate and prevent issues
11
IBM Research
© 2009 IBM Corporation
Food Security and Safety: Opportunities within the Advanced Technology Sector of Industry
Specific Opportunities
1. Supply chain management
2. Advanced analytics
3. IT infrastructure
12
IBM Research
© 2009 IBM Corporation
Opportunities for data analytic technologies in food safety and security solutions
§ Data Mining– The process of extracting hidden patterns from data
– Increasingly important tool as the volume of data increases
§ Predictive Modeling– The process of trying to best predict the probability of an
outcome
§ Risk Analysis– The process of identifying and assessing factors that
jeopardize the success of a goal
§ Statistical Analysis and Forecasting– The mathematical science of collection, analysis,
interpretation, explanation, and presentation of data, and estimating unknown quantities
Antibiotics
Logistics Logistics
Fertilizers Packaging Ingredients
Logistics
Ingredients
Logistics
Data Data Data Data DataData
Virtual Traceability System
Logistics Logistics Logistics LogisticsLogistics
Grocery Store/
RestaurantCorn Farmer
Cattle Rancher
Distribution Center
Beef Processor
CP Manufacturer
Transaction & Historical Data
FirewallFirewallFirewallFirewall
FirewallFirewall
Fir
ew
all
Track and trace products and risks within the four walls to isolate and prevent issues
Antibiotics
Logistics Logistics
Fertilizers Packaging Ingredients
Logistics
Ingredients
Logistics
Data Data Data Data DataData
Virtual Traceability System
Logistics Logistics Logistics LogisticsLogistics
Grocery Store/
RestaurantCorn Farmer
Cattle Rancher
Distribution Center
Beef Processor
CP Manufacturer
Transaction & Historical Data
FirewallFirewallFirewallFirewall
FirewallFirewall
Fir
ew
all
Track and trace products and risks within the four walls to isolate and prevent issues
What conditions existed prior to a food safety event?
Based on identified patterns, predict the probability of a future food safety event
Make decisions on risk mitigation. For example, product recall prior to a food safety event
Support impact analysis by forecasting demand under different hazard scenarios
13
IBM Research
© 2009 IBM Corporation
Risk models appear to be particularly relevant for food safety and security solutions
§ Statistical and machine learning models, used to discover key risk indicators and characterize likelihood and impact of risks based on historical data;
§ Simulation models, which are (usually) data-driven representations of a system facilitated by sampling from specified probability distributions.
§ Stochastic optimization models, where at least one of the variables involves uncertainty, and is assumed to follow a particular probability distribution
Ref: Ray, Apte, McAuliffe, Deleris and Cope, Harnessing Uncertainty: The Future of Risk Analytics, IBM Research Report, 2008
A layered view of the enterprise that maps key resources supporting business processes, and the causes of failures affecting these resources
14
IBM Research
© 2009 IBM Corporation
Why “advanced data analytics” is a technology to consider in food safety and security solutions?
§ Methods deal with large volumes of dataand address computational complexity
§ Unstructured data may be leveraged for improving predictive accuracy and insights
§ Approaches consider missing data and uncertainty
§ Even more powerful when combined with visualization tools
§ Advanced data analytics are the key for moving food safety and security from reactive to preventative
Stack graph Line graph Bar chart Scatterplot
US Map World map Block histogram Bubble chart
Pie chart Treemap (2 types) Stack graph for categories Network diagram
Stack graph Line graph Bar chart Scatterplot
US Map World map Block histogram Bubble chart
Pie chart Treemap (2 types) Stack graph for categories Network diagram
http://manyeyes.alphaworks.ibm.com/manyeyes/
15
IBM Research
© 2009 IBM Corporation
In system design for safety, the highest priorities are assigned to hazard prevention. Advanced data analytics help to move in this direction.
TimeHazard event
occurs
x
DamageMinimization
DamageMinimization
A hazard event has occurred. Minimize the damage.
HazardControlHazardControl
A hazard event has occurred. Mitigate the effects.
HazardReduction
HazardReduction
Minimize the probability of future hazard events occurring.
HazardElimination
HazardElimination
Complete elimination of the possibility for future hazard events.
* Reference: N. Leveson's adaption in Safeware: System Safety and Computers, Addison-Wesley, 1995 of the safety precedence described by W. Hammer, Handbook of System and Product Safety. Prentice-Hall, Inc. Englewood Cliffs, NJ, 1972.
PR
IOR
ITY
Traceability enables improved reaction given a food safety hazard has occurred
Advanced data analytic methods take large volumes of data from various sources, including from traceability solutions, to enable prediction and avoidance
16
IBM Research
© 2009 IBM Corporation
Food Security and Safety: Opportunities within the Advanced Technology Sector of Industry
Specific Opportunities
1. Supply chain management
2. Advanced analytics
3. IT infrastructure
| 17 © Copyright IBM Corporation 2009
Since the constituent database systems often must remain autonomous, a federated database system is an alternative to the (sometimes daunting) task of merging together several disparate databases
Antibiotics
Logistics Logistics
Fertilizers Packaging Ingredients
Logistics
Ingredients
Logistics
Data Data Data Data DataData
Logistics Logistics Logistics LogisticsLogistics
Grocery Store/
Restaurant
Corn FarmerCattle
RancherDistribution
CenterBeef
Processor
CP Manufacturer
Federated data base
Data to Smart Decisions - 18©2009 IBM Corporation
2005 2006 2007 2008 2009 2010 20110
200
400
600
800
1,000
1,200
1,400
1,600
1,800
Ex
ab
yte
s
DVD,RFID,
Digital TV,MP3 players,
Digital cameras,Camera phones, VoIP,
Medical imaging, Laptops,Datacenter applications, Games,
Satellite images, GPS, ATMs, Scanners,Sensors, Digital radio, DLP theaters, Telematics,
Peer-to-peer, Email, Instant messaging, Videoconferencing,CAD/CAM, Toys, Industrial machines, Security systems, Appliances
TenfoldGrowth in
Five Years!
TenfoldGrowth in
Five Years!
PhysicalRDFdocuments
2+ billion
RDF triples
Over 10,000ontologies
10M documents401M triples
Real-worldRDF Applications
12/07 04/08
Data arises from many sources (instrumentation, automation, on-line communities). Managing and preparing the data for use is a necessary first step in data-driven decision making.
Gather Data
Data cleansingSearchingFeature extraction
Semantic linking and extractionStream processingCrowd computing
Data to Smart Decisions
Define ProblemAct, Monitor,
Learn
DecideAnalyzeGather data
The number of semantically tagged documents and data sets is growing dramatically, improving data gathering capabilities.
The volume of digital data is exploding§80% of new data growth is unstructured§Data and metadata quality varies§Better approaches to finding relevant data are critical
Cloud - 19©2009 IBM Corporation
Flexible pricingFlexible pricing
Rapid provisioningRapid provisioning
Cloud Computing is a model of shared network-delivered services, both public and private, in which the user sees only the service, and need not worry about the implementation or infrastructure
InfrastructureServices
PlatformServices
ApplicationServices
BusinessServices
PeopleServices
Built on radically scalable, manageable, virtualized IT resources
Built on radically scalable, manageable, virtualized IT resources
Service layers separated by clean APIs, enabling
composition.
Service layers separated by clean APIs, enabling
composition.
Important roles for both public and private
clouds.
Important roles for both public and private
clouds.
Consumable web-delivered services
requiring no installation, minimal setup
Consumable web-delivered services
requiring no installation, minimal setup
Elastic scalingElastic scaling
Advanced virtualization
Advanced virtualization
Standard Internet technologies
Standard Internet technologies
Cloud
20
IBM Research
© 2009 IBM Corporation
Other recommendations to speed development of food safety and security solutions
1. Form real G-U-I teams to study the problem, identify gaps, and prioritize solution development approacheso Combine domain experts with technology, commercial and regulation expertise
o Take a ”system” view that recognizes multi-perspective, multi-objective aspects
2. Find ways to propose / contribute toward the economic stimulus packageso Mainly, financial stabilization, but includes spending in 2 other key areas
3. Take advantage of existing experiences w/ G-U-I relationships, e.g.o Joint research programs
o Programs for interns and summer students
o Research mentorship
o Workshops, various outreach
4. Learn from other industrieso Retail, Government, Automotive, Agriculture, Pharmaceutical, …
Creating Innovative Solutions With IBM Research © 2009 IBM Corporation21
IBM Research
© 2002 IBM Corporation
Example: Value of IBM’s FOAK Program
§ Links Research strategic initiatives to real client challenges
Validate market requirements
Test market readiness
§ Accelerate delivery of new technologies to the market
Enhance core technologies
Create new offerings
Enable On Demand Innovation Services (ODIS) engagements
§ Provide headlights into emerging market opportunitiesUncover new markets and growth opportunities
§ Gain valuable experience and thought leadershipSkills and knowledge transfer
§ Facilitate solution sales
Proof points for reuse
§ Create mindshareReferences and differentiation
22
Economic Stimulus:Government Spending Prioritized in Three Key Areas
New energy sourcesSustainabilityIndustry change Research
Intelligent transportationShared servicesCustoms, ports, borders
InfrastructureSmart gride-healthSchools modernizationBroadband
Cross sector stimulus
Loan managementRisk assessment and management
Financial stabilization and reform
Regulatory systemsTroubled assetsSocial safety net systems
Focus
Government stimulus
Program