Post on 26-Jun-2020
Confidential. Not to be copied, distributed, or reproduced without prior approval. © 2018 Baker Hughes, a GE company, LLC - All rights reserved.
Demystifying Digital Transformation for O&GBy leveraging AI at the heart of decision making
Rolando J. GabarronPrincipal Digital Product ManagerBaker Hughes, a GE Company
Confidential. Not to be copied, distributed, or reproduced without prior approval.
The information presented is intended to be an outline of general product direction and it should not be relied on in making a purchasing decision.
The information on the roadmap is for information purposes only, and may not be incorporated into any contract and is not a commitment, promise or legal
obligation to deliver any material, code, or functionality. The development, release, and timing of any features or functionality
described for our products remains at our sole discretion.
Safe Harbor Disclaimer
2
Confidential. Not to be copied, distributed, or reproduced without prior approval.
In some circles, it’s called Industry 4.0. In others, the preferred term is Industrial Internet ofThings (IIoT). In whatever terms you choose to describe it, the Digital Transformation ofIndustries is impacting the oil and gas industry alongside other major businesses inmanufacturing, healthcare, and aviation. In fact, according to the World Economic Forum, thedigital transformation of the Oil and Gas industry represents up to $1 trillion of economicvalue at stake through digitalization for Oil and Gas firms by 2025.
Baker Hughes, a GE company has a unique digital story, with BHGE Digital serving as thesoftware business creating and deploying cloud-based software solutions for oil and gascustomers. But what exactly does that mean for our customers?
Demystifying Digital Transformation
3
Confidential. Not to be copied, distributed, or reproduced without prior approval.
Industry 4.0 & Industrial Internet of Things (IIoT)
4
Industry 1.0Industry 2.0
Industry 3.0Industry 4.0
Confidential. Not to be copied, distributed, or reproduced without prior approval.
Oil & Gas Time Elapsed
5
Digital Transformation
Confidential. Not to be copied, distributed, or reproduced without prior approval.
a) Digital transformation is not a single departmental or workgroup idea. It is a company-wide initiative basedon collaboration between business and IT leaders who recognize that technology is critical to achievingbusiness goals.
b) Digital transformation is not the implementation of any one product or service. Digital transformation is thestrategic use—over time--of digital technology to drive nimble innovation, to find better ways of working, toadopt inventive business models, and to deliver engaging new customer experiences.
c) Digital transformation is not the public cloud. The public cloud is a big part of digital transformation, to besure, but digital transformation is more than that: it is a strategic integration of on-premise softwaresystems, and public, private and hybrid cloud technology—all enabled, managed and secured in a software-defined data center leveraging converged and hyper-converged infrastructure.
d) Digital transformation is not just about solutions. Digital transformation is about helping customers achievebusiness outcomes. To achieve true digital transformation, partners must be plugged into customers businessoutcome needs—and ready, willing and able to deliver on those needs in near real time.
What Digital Transformation Is ….. and Is Not (Source Channel Futures - Vmware Guest Blogger Apr 03, 2017)
6
Confidential. Not to be copied, distributed, or reproduced without prior approval.
The Internet of things (IoT) is the network of physical devices, vehicles, home appliances, and other itemsembedded with electronics, software, sensors, actuators, and connectivity which enables these things toconnect, collect and exchange data, creating opportunities for more direct integration of the physical world intocomputer-based systems, resulting in efficiency improvements, economic benefits, and reduced humanexertions. (From Wikipedia, the free encyclopedia).
The Industrial Internet of Things, or IIoT, is the use of internet of things technologies to enhance manufacturingand industrial processes. Also known as the industrial internet or Industrie 4.0, IIoT incorporates machinelearning and big data technologies to harness the sensor data, machine-to-machine (M2M) communication andautomation technologies that have existed in industrial settings for years.
IoT and IIoT, both have the same main character of availability, intelligent and connected devices. The onlydifference between those two is their general usages. While IoT is most commonly used for consumer usage,IIoT is used for industrial purpose such as manufacturing, supply chain monitor and management system.
Industrial Internet of Things (IIoT)
7
Confidential. Not to be copied, distributed, or reproduced without prior approval.
What is unique about BHGE’s approach to Digital Transformation?
Confidential. Not to be copied, distributed, or reproduced without prior approval.
…and how does it unleash value?
8
Confidential. Not to be copied, distributed, or reproduced without prior approval. 9
Key Challenges in Production
95% of data collected is
never used
35% average recovery
50% workforce retiring
in 5-10 years
13% average
downtime
9
Confidential. Not to be copied, distributed, or reproduced without prior approval. 10
About BHGE Digital125
Outcome Driven with AI at the Core
Software Developed By Oil & Gas Experts For Oil & Gas Experts
in the Oil & Gas industry
Delivering software for outcomes you care about:
Years of experience
Maximizing recovery
Optimizing production
Reducing NPT
Improving safety
Enabling enterprise-wide digital transformation
To really deliver the promise of the Industrial Internet of Things (IoT), you need a software company that knows the “Things” better than anybody.
We harness the power of data from connected devices, develop differentiated algorithms, and allow you to materially improve operational effectiveness and efficiency.
Our core DNA is Digital
We know IIoT
10
Confidential. Not to be copied, distributed, or reproduced without prior approval.
March 21, 2019 11
BHGE IIoT SW Offering
Cloud Native – IaaS & PaaS Services (GCP / AWS / AZURE / Private Cloud)
AI Factory – Enterprise Scale Platform Services
Workbench • Studio• Kernels & Techniques• Catalog
AI Runtime • Deep Learning• Model Lifecycle Manager• Orchestration• Data Fabric
Visualization • Analytics UI• Application Framework
Field Dev & Planning• Subsurface lithology
identification, prediction• Optimizing Drilling ROP
Production Optimization• Field Optimization with
constraints• Well Lifecycle Management
Reliability• Deep learning Anomaly
Detection • Anomaly Forecasting• Event Correlation
• Degradation Analysis• Corrosion Forecasting• Remaining Useful Life
Prediction
Process• System-of-System Prediction• Flow Models• Network Optimization• Global Optimizer
Industrial AI Applications & Use Cases
Enterprise Fit for Purpose Solutions BHGE Product Line EnablementCustomer-facing digital services
Integrity
Confidential. Not to be copied, distributed, or reproduced without prior approval.
Data science is truly a science, with complex concepts, mathematical models, physics-based techniques, etc.,requiring a data scientist with Ph.D. credentials.
BHGE Applied AI takes all of those techniques, AI and Deep Learning capabilities, and distills it down so it issimple for all users. You don’t have to be a data scientist to take advantage of Applied AI, since it is able toconnect from nearly any disparate data silo, perform the data wrangling, ingestion, curation, etc., and take in thecustomers’ data, domain expertise, and their existing analytical models during the learning process.
By applying Deep Learning techniques and capturing the learnings into the Semantic Knowledge Capture engine,Applied AI is capable of analyzing the data and pulling from its deep repertoire of techniques to automaticallyrecommend the best techniques to convey what the data is indicating.
Applied Artificial Intelligence (AI)
12
Confidential. Not to be copied, distributed, or reproduced without prior approval.
BHGE Applied AI – Data Science Made Simple
Data Science techniques for AI and Deep Learning are powerful, but applying those concepts to real-life
O&G problems can be intimidating…
Learning and Automatic Recommendations for Best AI
Techniques, Best Models, and Why
Your Existing
Analytical Models
Your Data
Your Domain
Expertise
BHGE Applied AI
13
Confidential. Not to be copied, distributed, or reproduced without prior approval.
BHGE Digital Applied AI
Access & Gather Data
Ensure Data Quality
Ensure Data Consistency
Access ALL the Data
Create Models/ Apply Artificial Intel. Techniques
Supercharge the Computation at Scale
Visualize the Outcomes
• Connectors• Automated Field (Tag) Mapping across Data Stores
• Cleansing• Missing Data Estimation
• Ready to use Catalog of Oil & Gas models• “Machine Learn” with Semantic Knowledge
Capture• Orchestration of System-Level Analytical
Models
• Nvidia partnership• Google partnership
Foundation: Security, User Management, Metering
• Alerting, Cases• Customer-
Configurable User Experience
• Feature Rich• Fast Deployment• Efficient• Scalable• Open Ecosystem
Scale Customer & BHGE Engineering Expertise
Deploy Analytics at Scale
14
Confidential. Not to be copied, distributed, or reproduced without prior approval.
Examples of O&G Specific Applied AI Models
15
Confidential. Not to be copied, distributed, or reproduced without prior approval. March 21, 2019 16
Our Growing Library of O&G Specific Applied AI Models
• ESP Remaining Useful Life• Multi-well ESP Optimization• AI-Based Rod Lift Fault Detection• Digital Twin• Virtual Flow Meter• Automated Anomaly Detection
16
Confidential. Not to be copied, distributed, or reproduced without prior approval. Improving P d i
17
Remaining Useful Life• Minimize unplanned downtime due
to unplanned workovers• Asset specific, unique to each ESP
based on real-time operational data (not averages based on history only)
• Manufacturer agnostic • Prediction with probability
(confidence)• Model pre-trained with clients data• Automated anomaly detection with
unsupervised learning for detection…supervised learning for identification
Predicting failure allowing for advanced notice to improve planning and decrease NPT, 30 – 100+ day lead time
17
Confidential. Not to be copied, distributed, or reproduced without prior approval. Improving P d i
18
Multi-Well Optimization• Increased production due to ultra-fast
optimization techniques • Simultaneously solve for well-level and
cluster-level constraints• 1000x faster than current solutions
delivered through parallelized cloud computing, accelerated well models and techniques
• Industry first simultaneous multi-well optimizer
Production field optimization, with cluster constraints (common topside infrastructure such as water disposal, separators, power supply) to highlight and deliver production upside.
18
Confidential. Not to be copied, distributed, or reproduced without prior approval.
AI Rod Lift Fault Detection & Identification
No manual rule-based training >50% improvement vs. best alternative Automatic fault identification in real time AI learns from human expert for unknown
faults…and packages, distributes “expertise” to all sites instantly.
March 21, 2019 12
©2018 Baker Hughes, a GE company, LLC (“BHGE”). All rights reserved.
Confidential. Not to be copied, distributed, or reproduced without prior approval.
March 21, 2019
Digital Twins…our Reliable Crystal Ball
Casing (Pipe) ModelReservoir Model Pump Model Tubing Model Pressure Gradient
@ Surface
Flow Rate, Q
Head
Efficiency
Horse Power
Best Efficiency Point
A live up-to-date digital representation of an asset, system, or process Unique ever-green model per well…that scales across a field Predicts total production with uncertainty estimates 10x improvement in accuracy vs. leading on-premise solution
Probabilistic learning
AI (Deep learning)
Deep domain models
20
Confidential. Not to be copied, distributed, or reproduced without prior approval.
Model Based Flow Metering
Estimate production (multi-phase) on a per-well basis without requiring new flow meters.
Probabilistic network models with uncertainty quantification
• Identify drivers for production variability
• Reduce uncertainty in production estimates (update Digital Twin)
21
Confidential. Not to be copied, distributed, or reproduced without prior approval.
Automated Anomaly Detection Unsupervised learning for detection…supervised learning for identification No explicit rule or signal specific threshold/condition necessary Scales well across very large number & combinations of tags Can work with linear, non-linear and discontinuous signals Equipment Agnostic Learn Locally…distribute globally >50% false-positive reductions vs. rule based systems
Example: ESP, RLS, Compressors
• Intrusion Detection• Slow Degradation• Excitations
22
©2018 Baker Hughes, a GE company, LLC (“BHGE”). All rights reserved.
All this is real Right Here…Right Now
23
Demystifying Digital Transformation for O&G�By leveraging AI at the heart of decision making �Slide Number 2Slide Number 3Slide Number 4Slide Number 5Slide Number 6Slide Number 7Slide Number 8Key Challenges in ProductionAbout BHGE DigitalBHGE IIoT SW OfferingSlide Number 12Slide Number 13BHGE Digital Applied AIExamples of O&G Specific Applied AI ModelsOur Growing Library of O&G Specific Applied AI ModelsRemaining Useful LifeMulti-Well OptimizationSlide Number 19Slide Number 20Slide Number 21Slide Number 22Slide Number 23Slide Number 24