The industry is changing: against a challenging backdrop with a ‘lower for longer’ economic forecast, Oil & Gas companies are turning to technology to modernise and improve their operations. This transformation has seen IT repositioned as a core business technology, drawn from a background support function to a crucial centre of value creation and innovation. This tectonic shift places IT leaders in a vital position within their organisation, ensuring existing assets and emerging technology are effectively harnessed to deliver tangible business outcomes.
Cost reduction is still the primary mandate for most organisations, with ongoing efforts to strip back overheads and address key areas of inefficiency to cope with tightening budgetary restraints. But while the pursuit of ‘more for less’ has become a fundamental necessity, it is important that the strategy employs sufficient safeguards to avoid stifling long term progress. Organisations need to retain the personnel, the skills and the tools to ensure they still have the capacity to innovate.
One of the most prevalent trends of recent years has been a concerted move towards greater automation. Organisations are increasingly incorporating sensors, robotics and live data feeds to enhanced remote operations. But this digitisation of process is not just taking place in far flung fields; across the operation, digital technologies are being applied to enable improved visibility and insight. And data analytics is increasingly being used to evaluate asset performance, and enhance predictability, forecasting and decision making.
Whilst operators have made strides to address inefficiencies and create faster, more agile processes, there are still several barriers to progress. Organisations need to adapt their structure, break down internal silos and allow more cohesive and collaborative engagement. This collaboration also needs to extend to the wider supply chain and external partners across the industry. Skills and leadership is also a key barrier to progress, while cultural inertia still poses a problem for the industry and needs to be tackled head-on if digital transformation ambitions are to be achieved.
This conference will bring together IT leaders from across the world for knowledge exchange, thought leadership and collaboration. Now in its 4th year, the conference has established itself as the must-attend event for IT leaders working in Oil & Gas. The programme will explore the use of Information Technology in driving tangible business benefits, with topics spanning: data analytics, cloud, cyber security, automation, leadership and culture.
4. Happy Birthday ICT Leaders…
If you've made it through the baby years, the terrible twos
and that dreaded threenager stage, and you're still
standing, …….Rumor has it that 4-year-olds leaving
toddlerhood and entering the preschool age are kind of
awesome. They live right there in that sweet spot where
they can talk and interact, and they might also listen to
reason. @sheknows.com
April 2017 4
6. Business Restructuring and Innovation
Enable innovation and business transformation
Joint Business and IT Cost Savings
Implement cost-savings initiatives and improve
business processes
Cost Savings within IT
Identify and prioritize opportunities to reduce IT costs
IT Procurement
Get the best pricing and terms
Difficulty
Value
External Procurement Review
Market Test – sub contract model
IT restructure
Shift Left
Contract redesign
2015 G&A -%30
Business Systems Projects
Offshore Bandwidth reduction
Source: Gartner (July 2011)
2016 G&A -%10
European Maximo consolidation
Evolution or same old cycle ?
8. Timing is Everything…
Industry
April 2017 8
Technology
People
Decommissioning and cessation of production
Mature fields
Cost Drivers
Transitions -m&a
Demographics
Global competition
Have seen and felt Digital Transformation-banking, betting, tv
Communication -facetime, messaging, skype, siri
My dad has stopped getting the Scotsman newspaper!!
The world has moved on, we haven't
Mobility – 17 years on from Nigeria
LTE/4G
Ex tablets/smartphones
Integration platforms
dev environmentsTechnology explosion – AI (Siri)
9. Transformational Technology
We focus on a lot of enablers, not Transformers.
-a Tsunami of Enablers.
April 2017 9
TECHNOLOGY+PEOPLE
+ VISION
= TRANSFORMATION
We have not Digitally Transformed,
we have DIGITISED
10. Digital Leaders…
We need to lead on Digital Transformation
The business case is key
Engage and manage the change
Hand in hand with the business
We need an UK industry response
April 2017 10
14. CGI Global 1000: Oil & Gas
14
Reduce the Run Invest in Change Grow Revenue
Industry Trends
Responding to revenue pressures
resulting from low oil price
76%
Assuring data privacy
protection/regulatory compliance
56%
Becoming digital organizations to meet
customer expectations
56%
Protecting through cybersecurity53%
Changing operational & business
models to drive operational excellence
29%
• The only industry where both Opex and Capex budgets have decreased
• Oil price pressure has industry focused on operational excellence and cost reduction
• Demand is increasing for operational agility to support asset re-alignment and data analytics to create new
business value
OpEx
(14.3%)
Decreased
CapEx
(10.7%)
Decreased
Business Priorities
Cost reduction and performance
improvement programs
75%
Optimize today’s operations64%
Harness the power of data analytics61%
Protect the organization as
cybersecurity risks mature
50%
Restructuring through mergers,
acquisitions, diversifications
44%
IT Priorities
Embrace new IT delivery models56%
Digitize and automate business
processes
56%
Drive IT modernization50%
Protect through cybersecurity50%
Deliver the benefits of big data and
business insight
47%
Source: CGI Global 1000 (2016)
15. Devon To Sell Midland
Basin Assets
Exxon And BHP Consider
Major Divestment
Chevron To Sell Off $5
Billion In Asian Assets
Suncor Makes Third Acquisition
This Year, While Rest Of Big Oil
Is Selling
Colombia’s Ecopetrol Plans
$13B Investments By 2020
Statoil To Sell $96 Million In
US Shale Assets
Maersk Oil Well-Positioned To Do
Well As Standalone Business
Chesapeake Energy Quits Shale
Revolution Cradle
Anadarko Splashes US$2
Billion On Freeport Oil Assets
Shell Divests $1B
Canadian Oil Assets
Anadarko Exits Eagle Ford
DONG Quits Oil, Gas, Stays
With Wind Power
Shell Mulls Divestment Of
Norwegian Assets
Total, Shell Sell Oil
Assets In GabonShell Aims To Sell Stake In Danish
Offshore Oil, Gas Venture
The industry is rebalancing portfolios – and the shale revolution
continues
Marathon Oil Sells Canadian Oil
Sands Assets, Bets On Permian
ConocoPhillips Exits Most Canadian Operations
Sinopec Nearing Deal To Buy Chevron’s $1B
South African Assets
Petrobras Ordered To Restart
Asset Sale Program
Private Equity Hunting For Oil
& Gas Assets In South-East Asia
Shell About to Close Major
North Sea Asset Sale
BG and Shell shareholders vote in favour of the
recommended combination between Shell and BG
15
16. 16
Oil & Gas is seeking digital transformation that will optimize
the business
* Many CGI clients span multiple industry verticals and may be more advanced than peers. For the purposes here we have used
the predominant industry and average across all CGI clients
Consumer Intensive
Asset Intensive
Insurance
Oil & Gas
Healthcare
Transport & Logistics
Retail
Banking
Manufacturing
Government
Utilities
Communications
Risk & Investment Intensive
Business
urgency
Political
urgency
Investigate
to Understand
Source : CGI Global 1000 (2016)
17. CGI Agile Energy 360
Agile Energy 360 clients use all or any parts of the solution on the schedule that suits their
business.
17
IT Services
Business Services
Software
and
Solutions
CGI IP
3rd Party
Software
SaaS
PaaS
IaaS
Cloud Strategy
and Migration
Systems Integration
Cyber Security
Support
Full ITO
App & Infrastructure
Management
Reporting & Analytics
Oil and Gas BPO Services
• Accounting
• Land Administration
• Division Orders
• Production
• Document Management
Digital Transformation
IT Strategy
Internet of Things (IOT)
Business Process
Optimization
Vendor Management
18. Energy demand and transition drivers
18
• World demand for energy will
continue to increase
• Natural gas, the “bridge” fuel to
a renewable future
• Decarbonization” of energy
supply chain
EIA International Energy Outlook 2016, figure 1-1
EIA International Energy Outlook 2016, figure 3-1. World natural gas consumption,
2012-40 (trillion cubic feet)
EIA International Energy Outlook 2016, figure 1-5
19. Consumer demand for “energy as a service” will likely increase
19
Sources: “Energy as a Service”, RE Magazine, April 2016; “Millennials’ to Drive Future Value for
Energy Utilities”, T&D Magazine, July 2016
20. 20
The Future of Oil & Gas: Integrated Energy/CO2 Chains
20
Courtesy of Trigen Energy Projects Development
Managing complex integrated energy/resource systems for
optimum lifecycle value and lowest environmental impact
21. Opportunities for operational integration
• Natural gas can be:
• Transformed into electric power in the field or at centralized generation facilities
• This generates CO2, heat, and water. Each can be stored, transformed, transported, and/or
sold/traded
• Opportunities for EOR and/or sequestration
• Electric power can:
• Compress natural gas as a storage mechanism
• Split water into hydrogen and oxygen as a storage or production mechanism
• Compress natural gas into LNG
• Be produced on regasification of LNG
• Commitments for natural gas, electric power, and LNG can be satisfied either by internal
commitments or market contracts/trades
• Increased operational options create increased opportunities for profit
• Moving toward a real-time or right-time mode of operation will require greater
integration fueled by IT
21
22. Low Hydrocarbon Energy
Portfolio
Energy Supply Chain Optimization
Supply Chain Optimization
LNG
%
Interest Rate
Swap
$/€
Currency
Swap
$
Futures
$
Swaps
• Prices
• Tariffs
• Volumes
• Weather
• Location
• Supply
• Demand
• Currencies
• Trading Instruments
Credit Exposure Constraints
Market Exposure Constraints
Price Exposure Constraints
Position Limit Constraints
High Hydrocarbon Energy
Portfolio
Based on our:
• Asset classes
• Trading instrument expertise
• Markets
• Regulatory requirements
• Risk tolerance
• Strategic direction
• ….
What are our best trading strategy
options?
• Correlations
• Weights
• Optimization algorithms
• Metaheuristics
• Fitness functions
24. Summary – the future may be;
Big Furry Mammals (Integrated Value Chain) vs
Small Agile Dinosaurs (Niche Specialty Companies)
24
The current segmented value chain does not require integration across
the value chain.
Even in the current scenario integrated companies can benefit from
supply chain integration and optimization.
Renewable energy will increase in market share.
Climate factors and energy transition will require a much more integrated
mode of operation to manage the complete carbon cycle.
Integrated oil companies are best positioned to optimize the future
complex integrated value chain.
Smaller specialty companies may find niche value, if they remain agile.
25. Our commitment to you
We approach every engagement
with one objective in mind: to help
clients succeed
27. Transforming the data centre
Leaving Legacy Infrastructure Behind
James Sturrock
Senior Systems Engineer
james.sturrock@nutanix.com | @sturroj
28. Transforming your Data Centre - not a choice
https://www.ted.com/talks/malcolm_gladwell_on_spaghetti_sauce#t-1034286
29. IT Challenges
1. IT Budgets
2. Scale & Complexity of IT
3. User & Business Expectations
• No innovation in Infrastructure in the last 10 years
• Increase in “Shadow IT” and uncontrolled costs
Bridging the Gap
• Automation
• Simplification
• Predictability
• Performance
• Resilience
31. Dilemma of Bi-modal IT
• IT runs inside-out “Traditional first” vs. outside-in “Digital first”
• Backward-looking reporting vs. predictive data led analytics
Information and technology leadership
• IT efficiency vs. Value creation
• Reducing costs vs. Increasing revenue per £ of IT spent
Value Leadership
“Traditional
first”
“Digital
first”
Visible Valuable
Control Vision
• Run current IT shop vs. Become strategic
• Command and Control vs. Vision and Inspiration
• Traditional risk averse culture vs. value creation culture
People Leadership
Source: Gartner, “Flipping to Digital Leadership: The 2015 CIO Agenda”
33. Breaking the legacy mould
33
Virtualization
App App
Storage
Controller
Storage
Controller
Storage
Controller
Storage
Controller
Server Server
Storage
Controller
Storage
Controller
34. Bringing The Cloud To The Enterprise Datacenter
34
Fractional Consumption Invisible
Operations
Instant
Delivery
Frictionless
Tailored SLAs for
Every App
Balance Owning and
Renting
Data Access and
Governance
Choice and Freedom from
Lock-in
Control
Continuous
Innovation
35. Transforming the Enterprise Datacentre
35
Virtualization
App App
Integrated, scale-out compute and
storage
Virtualization
App App
Storage
Controller
Storage
Controller
Storage
Controller
Storage
Controller
Server Server
Storage
Controller
Storage
Controller
with built-in virtualization
and management
37. All Flash (and beyond) Reads
37
SSD
SSD
HDD
HDD
HDD
HDD
CVM VM VM
SCSI Controller
CPU
RAM
VM
Hypervisor
Hot Cold Compute
StorageI/O
• Virtualised SAN Controller
• Server BUS data performance
• Data remains local
Physical
Logical
Performance and availability
• Data is read locally
• Remote access only if data not locally present
Node
Hypervisor
Controller
VM
Storage
Node
Hypervisor
Controller
VM
Storage
Node
Guest
VM(s)
Hypervisor
Controller
VM
Storage
Node
Hypervisor
Controller
VM
Storage
38. Why Care About Data Locality?
38
0
10 0 0 0
20 0 0 0
30 0 0 0
40 0 0 0
50 0 0 0
60 0 0 0
ThroughputinMB/S Flash Network
SSD 10G
NVMe 40G
100G
3DXPoint
When Applications
predominantly access
data locally, NW
bandwidth demands are
lowered
39. Writes and guaranteed data resilience
39
Node
Hypervisor
Controller
VM
Storage
Node
Hypervisor
Controller
VM
Storage
Performance and availability
• Data is written locally
• Replicated to other nodes for high availability
• Data replicated across the cluster for high performance
Node
Guest
VM(s)
Hypervisor
Controller
VM
Storage
Node
Hypervisor
Controller
VM
Storage
48. Commercial Confidential 2017
Net-Defence >_ DLP: What is it?
“…aims to prevent the unauthorised transfer of classified
information from a computer or datacentre to the outside
world...”
“…a strategy for making sure that end users do not send
sensitive or critical information outside the corporate
network. The term is also used to describe tools that help
your IT Dept control what data end users can transfer.”
49. Commercial Confidential 2017
Net-Defence >_ DLP: A Case Study - Background
• In January 2017, a global aerospace firm reported a data breach
involving an employee emailing spreadsheet containing sensitive
information to an outside recipient. The spreadsheet, sent to provide
the employee's spouse with a formatting template, contained the
personal information of roughly 36,000 employees, including Social
Security numbers and dates of birth, in hidden columns.
• According to research by IBM and the Ponemon Institute in 2016,
the average cost of a data breach was estimated to be around
$158 per record, making the cost of this event around $5,700,000.
50. Commercial Confidential 2017
Net-Defence >_ DLP: A Case Study – The Response
“On January 9, 2017, we discovered that a company employee set an
email containing personal information of approximately 36,000
other employees to his non-company spouse on November 21, 2016.
During the company’s investigation, the employee stated that he
sent a spreadsheet with the personal information to his spouse for
help with a formatting issue. He did not realise there was sensitive
personal information included on the spreadsheet because that
information was contained in hidden columns.”
51. Commercial Confidential 2017
Net-Defence >_ DLP: A Case Study – What Went Wrong
• Was the employee aware of the dangers of sending the information outside
of the organisation?
• Were there assigned Data Owners responsible for overseeing custody of
this type of data?
• Were there adequate technical tools in place to detect and prevent the
sending of this data? If so, where did these fail?
• Were documents of this type protectively marked and backed by a data
classification policy?
52. Commercial Confidential 2017
Net-Defence >_ DLP: Consequences Of Data Loss
• Reputational harm & negative media attention
• Loss in customer confidence
• Loss of revenue
• Legal or regulator action
• Loss of Intellectual Property
• Exposing data subjects to increased data misuse risk: such as identity fraud
53. Commercial Confidential 2017
Net-Defence >_ DLP: Implementing A Strategy
Key Starting Points:
• Identification of data ingress and egress channels
• Mapping of data boundaries
• Assignment & Classification of Data Types
• Assigning Data Owners
• Building Effective Policy & Procedure
• End User Training & Awareness
• Auditing & Monitoring
• Technical Tooling
54. Commercial Confidential 2017
Net-Defence >_ DLP: Identifying Data Channels
Compile a list of all data ingress and egress points. These should include:
• Technical: Web, Email, IM, Cloud, 3rd Party Integrations, laptops, mobile
devices, storage media
• Non-technical: Postal Correspondence, Printed Media, Documentation,
Supplier Agreements, Contracts
• People: Visitors, Customer Facing Staff (including phone based, face to
face)
55. Commercial Confidential 2017
Net-Defence >_ DLP: Mapping Data Boundaries
Where data channels have been identified they should be mapped:
• Full Journey: The data journey from receipt to storage should be
mapped
• Record Touch Points: List all systems & services touched on by data
during its journey
• Audit: Audit those touch points identified above and record what data is
stored and where
• Visualise: Build data flow diagrams, these help to visualise boundaries
56. Commercial Confidential 2017
Net-Defence >_ DLP: Assignment Of Data Types
Split out those data types into logical categories:
• Printed: Hard to control/audit. High risk of leakage
• Data At Rest: Easy to control/audit. Variable risk of leakage
• Data In Transit: Ability to control/audit varies according to boundary.
Variable risk of leakage
• Data In Use: Hard to control/audit. Variable risk of leakage
57. Commercial Confidential 2017
Net-Defence >_ DLP: Classification of Data Types
Now you’re aware of what data you have, its type & journey its now time to
apply a classification to each. Considerations include:
• Business Value: How valuable is the data to the business
• Consequence of Loss: How would the business be affected in the event
of a breach
• Classification Scheme: Implement a simple, business wide protective
marking scheme (e.g. Confidential, Internal, Public)
• Classification & Retention Policy: Implement a single policy, this should
cover data retention (how long to keep, how to destroy, encryption)
• Training & Awareness: Ensure data users know their responsibilities
58. Commercial Confidential 2017
Net-Defence >_ DLP: Assigning Data Owners
All data should be ‘owned’ either at functional or job role level. For example:
• Legal: Contracts, IP, Supplier Agreements, Governance
• HR: Employee PII, Employee Financial Data, Recruitment Data
• Sales: Client Business Data, Internal Pricing, Client Quotes & Finances
• Finance: Payroll/Salary Data, Profit & Loss, Financial Statements
• IT: Network Diagrams, Configuration Data, Source Code, Passwords
59. Commercial Confidential 2017
Net-Defence >_ DLP: Policy & Procedure
Minimum policy & documentation set:
• Data Classification Policy
• Data Retention Policy & supporting management procedures
• Data Destruction Policy
• Access Management/Control Policy & Supporting Procedure
• Data & Equipment Acceptable Use Policy
• Documented Job Roles (linked to access control/management policies
above)
• Documented data owner responsibilities, reporting lines & escalation
paths
60. Commercial Confidential 2017
Net-Defence >_ DLP: User Training & Awareness
• Establish & embed baseline training into the starter/mover/leaver process
• Ensure refresher training is delivered on a regular basis
• Have data owners provide input on, and sign off of training and awareness
courses affecting their respective areas
• Benchmark the uptake of training & awareness sessions through regular
testing
• Assign training & awareness to an owner to ensure materials are updated
61. Commercial Confidential 2017
Net-Defence >_ DLP: Auditing & Monitoring
• Perform regular reviews of policy & procedure to ensure they remain
effective
• Perform regular audits of identified data types, classifications & owners
• Ensure all tooling & supporting systems are logging and auditing data
access, modification & deletion
• Record and benchmark user training sessions
• Perform regular ‘red team’ exercises to ensure data boundaries are guarded
and fit for purpose
• Apply ‘continuous improvement’ principles to your DLP management
strategy
62. Commercial Confidential 2017
Net-Defence >_ DLP: Auditing & Monitoring
The following ISO27001 clauses can assist when establishing a DLP auditing &
monitoring strategy:
• Monitoring & Measurement Results - clause 9.1
• Internal Audit Programme - clause 9.2
• Internal Audit Records - clause 9.2
• Management Review Records - clause 9.3
• Results of Corrective Actions - clause 10.1
• User activity, exceptions, security & event logs - clauses A12.4.2 &
A.12.4.3
63. Commercial Confidential 2017
Net-Defence >_ DLP: Evaluating Technical Tooling
The following points should be considered when evaluating DLP technical
tooling:
• Monitoring vs Prevention
• Centralised Management
• Backup & Storage Requirements
• Cloud or Self Hosted
• Ease of Integration
• Resources Required to Manage and Monitor
• Flexibility of Rulesets and support for custom rules
• Vendor Support
64. Commercial Confidential 2017
Net-Defence >_ DLP: Recommendations
Data Format: Printed
Control
• All staff should be made of their responsibilities throughout their
employment
• Do not leave copies of sensitive data unattended on desks, printers, fax
machines, copiers and other common access areas. Lock them away when
unattended
• Do not leave sensitive data visible/accessible to the public
• Shred sensitive paper records when no longer needed
65. Commercial Confidential 2017
Net-Defence >_ DLP: Recommendations
Data Format: Data in Transit
Control
• Sensitive Data should be sent and received from authorised personnel
inline with the Information Security Policy
• Devices that process sensitive data should be physically secured or locked
away when unattended
• Infrastructure assets that process sensitive data such as Networks, Systems,
Applications and Databases should be segregated and physical access
managed by controlling and restricting access to authorized personnel only
66. Commercial Confidential 2017
Net-Defence >_ DLP: Recommendations
Data Format: Data in Transit
Control
• Sensitive Data should be sent and received from authorised personnel in
line with the Information Security Policy
• Devices that process sensitive data should be physically secured or locked
away when unattended
• Infrastructure assets that process sensitive data such as Networks, Systems,
Applications and Databases should be segregated and physical access
managed by controlling and restricting access to authorized personnel only.
• Data traversing public networks should be protected by SSL/TLS or a VPN
67. Commercial Confidential 2017
Net-Defence >_ DLP: Recommendations
Data Format: Data at Rest
Control
• Sensitive data should be stored only in authorised locations, with a valid
business reason and in line with the applicable security policy
• Physical access to assets that store sensitive data should be controlled and
restricted to authorised personnel only
• Sensitive Data at rest in authorised locations such as database servers
within customer or external networks should be encrypted
• Sensitive Data in Backup and storage should be encrypted
• Endpoints that are authorised to store sensitive data should be encrypted
68. Commercial Confidential 2017
Net-Defence >_ DLP: Recommendations
Data Format: Data in Use
Control
• Sensitive Data should only be accessed and used by authorised personnel in
line with the Information Security Policy
• Devices that access sensitive data should be secured or locked away when
not in use
• Infrastructure assets that are used to access sensitive data such as
Networks, Systems, Applications and Databases should be segregated and
physical access controlled and restricted to authorised personnel only
69. Commercial Confidential 2017
Net-Defence >_ DLP: Recommendations
Data Format: Removable Media
Control
• Portable/Removable Media should be used by authorised personnel based
on the approval from stakeholders in line with the information security
policy
• Portable/Removable media should be locked away when not in use or
unattended
• Portable/Removable media should be encrypted by default
• Portable/Removable media should never be taken off site without the
correct approval
72. 72
The next 25 minutes…
A bit about me…
The Five Step Migration Methodology
Case Studies – two different approaches
Handover? It’s up to you!
A bit about brightsolid…
74. 74
bright & Solid Product Roadmap
Longer-term data center planning
must be done in the context of the
enterprise's plans for application
and workload placement
relative to cloud computing
to ensure facility needs and
forecasts are realistic and rightsized
Gartner 2016 Strategic Roadmap for
Data Center Infrastructure
Innovation
is in our
hearts
Our Approach
Customer Market Driven
Technology Driven
Bi – Modal
Mode 1 – traditional
infrastructure
Mode 2 – cloud native
Try it on ourselves first!
Use business strategy — not technology — to drive infrastructure strategy
Successful IT organizations must meet digital business challenges
by adopting a bimodal approach to IT — a reliable Mode 1 that is
focused on safety and efficiency, and an agile Mode 2 that is
focused on flexibility and speed
Gartner 2016 Strategic Roadmap for Data Center Infrastructure
The Era of Managed Infrastructure
Services: Managed is the New
Normal – 62% of Cloud/Hosting
Infrastructure Spending Comes
Bundled with Value-Added Services
451 Research Hosting & Cloud
Study 2017
75. 75
The Five Step Migration Methodology
Collaborative Initiation
Communications pact
Collaboration on Design & Plan
Finding the right Partners
Procurement
Mitigation & best value
Designed Delivery
Drivers, timelines & risk appetite
Collaborative
Initiation
Partnering
Designed
Delivery
Handover? Develop
Handover?
It’s up to you
We’re your support team
Development Opportunities
Opportunities for service improvements
Account Management
79. 79
Martin Currie Migration
Project requirements:
Virtually risk free
Roll-back options
Stop dead date – contract ended –
lines and migration
Cyber Essential Plus in place
Prince 2 – Auditable
Planning & Risk Mitigation
80. 80
Martin Currie Migration
Our Approach
Plan, Mitigate, Plan, Mitigate
In-depth design – customer
workshops
Identify market leader physical
migration partner
Work with incumbent to ensure
smooth transition of network &
services
Planning & Risk Mitigation
81. 81
Key Milestones – Martin Currie
Milestone Start Date End Date
1 Service Discovery and Planning 24 Feb 15 30 Mar 15
2 Data lines ordered 01 May 15 03 Sep 15
3 WAN Testing and acceptance 04 Sep 15 10 Sep 15
4 Production Site Migration 25 Sep 15 26 Sep 15
5 Production Site Testing 26 Sep 15 26 Sep 15
6 Service Commencement (Production Site live) 26 Sep 15
7
CC Exit Plan: Formal notice that CC can terminate agreed services in Leeds
(Point of no return for production)
08 Oct 15
8 DR Site Migration 13 Nov 15 14 Nov 15
9 DR Site Testing 14 Nov 15 14 Nov 15
10 DR Site Live 14 Nov 15
11 Long Stop Date, Production 19 Nov 15
12 Long Stop Date, DR 02 Dec 15
13
CC Exit Plan: Formal notice that CC can terminate all remaining services. (Point
of no return for DR)
30 Dec 15
83. 83
Aberdeen City Council Migration
Project requirements:
The impossible with an immoveable date
Out of their incumbent supplier
Substantial financial penalties
Minimized down time – weekend windows
No downtime – 08:00 Monday – 17:00 Friday
Storage requirement
Commercial and Timeline drivers…..
What is the art of the possible?
84. 84
Aberdeen City Council Migration
Our Approach:
Plan, Test, Do – asap
Identify optimal way to transfer data and
services
Reliance on brightsolid expertise
Mutual trust required
Open, honest transparent
Collaboration – talk talk talk
Commercial and Timeline drivers…..
What is the art of the possible?
85. 85
Key Milestones – Aberdeen City Council
6 weeks to deliver the migration
Milestone Duration Start Date End Date
ACC Project 38 days 27/11/15 06/01/16
1 DR re-located to Aberdeen DC for use as Prod 4 days 27/11/15 30/12/15
2 Prod Phase 1 VMs & Kit to Aberdeen DC 5 days 03/12/15 07/12/16
3 Prod Phase 2 VMs & Kit to Aberdeen DC 5 days 10/12/15 14/12/15
4 Prod Phase 3 VMs & Kit Aberdeen DC 5 days 17/12/15 21/12/15
5 Prod Phase 4 to Aberdeen DC 8 days 23/12/15 31/12/15
6 DR Phase 2 to Dundee DC 1.5 days 05/01/16 06/01/16
VDE Issue Resolution:
92. TIME LINE OF DEVELOPMENT
• Balloon warfare in 1845
• WWI & II development and the introduction of the term drone for autonomus
flight
• Hobbyists in a field with the patients of a saint
• The military use of drones in modern warfare
93. Intelligent software to control flight, helping
autonomous flight
Better coordination between return to home
feature and anti collision
Ability to repeat flight paths months apart
106. • Rapid deployment with clean up
team
• Transponders for tracking the slick
• Constant monitoring
• Clean up coordination
• Multispectral cameras for picking
up the oil against the black of the
North Sea easier
107.
108. WILLIAMS F1 COLLABORATIONS
• Battery development. Increase safety and reliability, greater endurance.
• Design a drone utilising the aerodynamic engineering skills within a world leading engineering
company. Enabling us to fly in that 25-40knt window.
111. Machine Learning and Vision Applications
Eyad Elyan
School of Computing Science and Digital Media
Robert Gordon University
Oil & Gas ICT Leader 2017
April 19, 2017
112. 1 Humans & Machines
2 Challenges &
Opportunities Oil and Gas
Data
Opportunities
3 Background
Learning from Data
Past and Present
Examples
4 RGU
Computing
Research
Industry-Partnerships
113. Algorithms vs Humans
A bat and a ball cost $1.10.
The bat costs one dollar more than the ball.
How much does the ball cost?1
Answer
The ball cost 10 cents ✗
ball cost 5 cents ✔
1
Thinking Fast and Slow by Daniel Kahneman
114. Algorithmic Solution
1 bat + ball = 1.10
2 1 + ball = bat
3 ball = bat −1
4 substitue in 1
5 bat + (bat − 1) = 1.10
6 2bat = 2.10
bat = 2.10
7 2
8 bat = 1.05
9 ball= bat − 1 = 0.05
115.
116. The Invisible Gorilla
Imagine you watch a video in which twoteams in white and black shirts pass balls around.
You areasked to count the number of passes made by the people in white shirts. During
this, a gorilla strolls into the middle of the action and faces the camera, then leaves,
spending 9 seconds on the screen.Would you see the gorilla?
In an experiment at Harvard, half of the people who watched the video
missed the gorilla!!
"This experiment reveals twothings: that wearemissing a lot of what goes on around
us, and that we have no idea that we are missing so much"2
2
Christopher Chabris,http://www.theinvisiblegorilla.com
120. Challenges
Different data modalities (text, images, notes, sensors, ..)
A need for moreintelligent ways to utilise and make senseof such legacy of data Real-
time monitoring and predictions
Large volumes of data needs to be digitised and intelligently processed
121. Opportunities
It is possible to digitise data and make senseof it (its happening) Can
we digitise and replicate human expertise?
123. Machine Learning
Machine Learning gives computers the ability to learn without being explicitly programmed (Ar
Observations (past examples) areused to train computers to perform certain tasks such as pred
Spam detection Fraud detection
Give a customer a loan?
...
124. Formal Definition
A dataset A with m instances x1, x2, ..., xm, where each instance xi is defined by an n
features as xi =(xi 1, xi 2, ..., xin).
A =
...
x11 x1nx12 ...,
x22 ...,
... ... ... ...
xm1 ... ..., xmn
y1
...
, Y =
..
..
ym
(1)
Learn a function h(x) that maps an instance xi ∈A to a class yj ∈Y .
127. 63 Years ago
Paul Meehl published his book ‘Disturbing little book’ and in oneof his studies he
comparedhuman experts performance against simple algorithms based on some observed
data on 20 different medical cases
In each of these 20 cases, the simple algorithms outperformedthe well-informed
human experts in the domain (medical cases)
Paul E. Meehl, Clinical Versus Statistical Prediction: A Theoretical Analysis and a Review of the Evidence Minneapolis, MN: University of
Minnesota Press,1954)
133. Deep Blue, 1997
IBM Deep Blue was the first computer to
beat chess champion Kasparov in 1997
134. Netflix, 2006 - 2009
"To qualify for the
$1,000,000 Grand
Prize, the accuracy of your
submitted predictions on the
qualifying set must be at least 10%
better than the accuracy Cinematch
can achieve on the same training
data set at the start of the Contest."
[source:http://www.netflixprize.com/rules.html]
135. Target, 2012
"How Target Figured Out A Teen Girl Was Pregnant Before Her Father Did", [source:
http://www.forbes.com/]
136. AlphaGo, 2016
Breakthrough in machine learning / deep learning whenGoogle DeepMind’s
AlphaGo computer program won against Lee Sedol
[Image source:https://gogameguru.com/younggils-pro-go-videos-deepmind-alphago-vs-lee-sedol-game-4/]
137. Cancer Detection, 2017
Google AI Just Beat Human at Detecting Cancer (89% vs 73% humans accuracy) 3
3
https://www.fool.com/investing/2017/04/04/google-ai-just-beat-human-pathologists-at-
detectin.aspx
139. Ray Kurzweil: Predicted that a computer would beat a human in chess and self-
driving cars (happened, happening)
"fundamental measures of information technology follow predictable and
exponential trajectories."4
Ray Kurzweil
4
http://uk.businessinsider.com/ray-kurzweil-law-of-accelerating-returns-2015-5?r=US&IR=T
148. Ensemble Learning
Combining a number of classifiers to vote towards the winning class has been
thoroughly investigated by machine learning and data mining communities.
149. RF: State-of-the-art Classifier
179 classifiers
121 datasets (the whole UCI repository at the time of the experiment) Random
Forest was the first ranked, followed by SVM with Gaussian kernel
Reference
Fernandez-Delgado, M., Cernadas, E., Barro, S., & Amorim, D. (2014). Do weneed
hundreds of classifiers to solve real worldclassification problems?. The Journal of Machine
Learning Research, 15(1), 3133-3181.
150. Research Focus
Eyad Elyan, Mohamed Medhat Gaber, A genetic algorithm approach to optimising random forests applied to class engineered
data, Information Sciences, Volume 384, April2017, Pages 220-234, ISSN 0020-0255,
http://dx.doi.org/10.1016/j.ins.2016.08.007
Ahmed Hussein, Mohamed Medhat Gaber, Eyad Elyan, and Chrisina Jayne. 2017. Imitation Learning: A Survey of Learning
Methods. A C M Comput. Surv. 50, 2, Article 21(April 2017), 35 pages. DOI:https://doi.org/10.1145/3054912
Ahmed Hussain, Eyad Elyan, Mohamed Gaber, Chrisina Jayne, "Deep Reward Shaping from Demonstrations", to-appear in
2017International Joint Conference on Neural Networks ( I J C N N 2017)
Eyad Elyan and Mohamed M. Gaber. A fine-grained random forests using class decomposition: an application to medical
diagnosis. Neural Computing and Applications, 27(8):2279-2288,2016,doi:10.1007/s00521-015-2064-z
Barrow, E., Eastwood, M., Jayne, C.(2016), Selective Dropout for Deep Neural Networks. ICONIP (3) 2016: 519-528
Barrow, E., Jayne, C., Eastwood, M., (2015), Deep Dropout Artificial Neural Networks for Recognising Digits and Characters in
Natural Images. ICONIP (4) 2015: 29-37
168. Mining and Visualising Oilfield Data
(InnovateUK)
(Data Lab Innovation Centre)
Process structured and unstructured oilfield data Text
processing / NLP using Deep Learning Building
predictive models for intelligent well-planning Visualsation
tools
169. Way Forward
The amount of data available is challenging the human brain and the state of the art
technologies
Hardware and learning algorithms are improving at exponential rates
Machine learning provides a unique opportunity to uncover hidden knowledge,
improve existing practices, etc. . .
Collaboration between academia and industries provide great opportunities to test state-
of the art research findings against real-world challenging problems
212. We supply integrated
products and services
Design and engineering support
Equipment sale and rental
Operations and maintenance
management
Inspection and integrity
management
Training and competence
assessment
Spare parts supply
In these
specialist areas
To customers in
these sectors
Drilling
(Rig owners)
Upstream facility
(E&P operators, EPC’s)
Marine
(Marine contractors)
Cable and pipe lay
Lifting and mechanical handling
Fluid power
Renewables
What we do
213. Global coverage with a local focus
Sparrows Americas
Sparrows Europe
Sparrows Africa
Sparrows MEICAP
Centres of engineering excellence
Strategic partnerships
Abbeville,
Broussard,
Houma &
Slidell, USA
Houston,
USA
Mexico
St John’s,
Canada
Trinidad &
Tobago
Macae,
Brazil
Aberdeen,
UK
Great
Yarmouth
& Norwich, UK
Netherlands
Kazakhstan
Saudi
Arabia
Qatar
Abu Dhabi &
Dubai, UAE
Mumbai,
India
Malaysia Singapore
Batam,
Indonesia
Jakarta,
Indonesia
Perth,
Australia
Malongo
& Luanda,
Angola
CongoGabon
Cameroon
Lagos,
Nigeria
China
214. Sparrows IT Case Study - 2 YearsAgo
• Failed ERP Project & Outsource Initiatives
• Over promised/ Under Delivered
• Significant lack of trust in IT enablement
• Existing application portfolio had been neglected
• Oil Industry Downturn was taking full effect
• A tsunami of demand
• Reduce Costs
• Reduce IT Team
• A Compelling Event
• Fertile Ground for new approach
• Cloud past the “tipping point”
214
215. Sparrows IT Case Study - IT Strategy
• To improve Business Performance through an improved Application Portfolio
• IT as Core Competency
• Embrace & go beyond shadow IT
• Let experts talk to experts
• Re-shape the IT Team
• Substantially reduce effort on Infrastructure & Support
• Focus on Analysis, Project Delivery & Integration
• IT team has to add value or get out of the way
• Cloud as an Opportunity
• 100% Commit
215
216. • Infrastructure (as a Service)
• No core systems or data remains on premise
Sparrows Case Study - Our Current Landscape
Legacy Applications
217. • Platform (as a Service)
Sparrows Case Study - Our Current Landscape
217
• SIMS (Bus Mgt)
• Intranet
• HR Service Centre
• Legal/Commercial Library
• Forecasting
• Budgeting
• Manifesting
• Mobilisation
218. • Software (as a Service)
Sparrows Case Study - Our Current Landscape
218
219. Are we on target?
• Cost Reduction
• Business teams are leaner – but with better tools
• IT Team is smaller
• IT budget is flat - bigger application portfolio
• Zero capex on servers / storage/ backup/ etc.
• A transformed IT Estate
• All key data and services now in the cloud
• Now Highly resilient/scalable
• We have dramatically reduced effort on Infrastructure & Support
• Greater Agility
• Speed of evaluation & deployment
• Cloud Applications allow experts to talk to experts.
• That gives greater ownership which is key to successful adoption
• Pain points
• Licensing models still immature
• Some software vendors don’t know how to run their own applications
• Geography & bandwidth are limiting factors
• Skype PBX
219
224. Putting data at the heart
of every decision.
Transforming the UKCS by
unlocking the wealth of
information hidden in the data.
225. • Exploration cycle is longer than the time
remaining to develop
• Decommissioning is upon us
• Can’t wait for the oil price to recover
• The next generation will demand it!
Need to act now and deploy the best, most advanced & most integrated solutions
226. • Well logs
• Well core and samples
• Seismic; pre & post stack
• Mapped interpretations
• Engineering data / drawings
• Structured / unstructured
• Reports
• Core / sample descriptions
• Photographs
• Rock samples
227. • Well temp, pressure, flowrates, phase
• DTS / DAS data volumes
• Valve performance monitoring
• Rotating equipment
• Vessel inspection video
• Pipeline corrosion / erosion tracking
• Chemical treatment monitoring
• Video surveillance & measurement
• Supply chain logistics
228. • Well integrity & performance
• Plant inspection & maintenance
• Planned shut-down optimisation
• Production information
• Sub-surface modelling & visualisation
• Supply chain optimisation
230. Leverage the scale of data & information across the UKCS
• Define key business challenges
and opportunities
• Learn from other industries –
airlines/airports, banking/finance,
automotive/manufacturing
• Identify optimisation opportunities
• Encourage collaboration – leverage data
across the basin – expose the data
• Use multiple & latest analytic, machine
learning and cognitive techniques
• Leverage both industry and academic
resource, knowledge and skills
• Consider the human implications
• Drive automation and digital innovation
through acceleration
• Deploy solutions nimbly and at scale
231. Applying data to transform the way we work
• Evaluating Industry needs
• Progressing scoping
• Developing business cases
• Understanding University & Industry offering
• Identifying Performance Gaps
• Learning from other industry parallels
232. Applying data to transform the way we work
Changing the
way we work And how it could apply
to the industry
Reduce
Costs
Find more
barrels
Become more
Efficient
Make better
Decisions
233. Help transform
the future of the
UKCS!
Help us define
significant industry
challenges requiring
new digital solutions
Collaborate by
match funding in
cash or in-kind
Share your
innovative ideas
for new digital
solutions
Join the team!