Hvordan legge til rette for dataanalyse i stor skala i industrien?

Transcription

1 Hvordan legge til rette for dataanalyse i stor skala i industrien? Anna Olsson Director Partners & Alliances, Cognite

2 Cognite Our mission: To present a digital representation of industrial reality to make it accessible and meaningful for humans and machines. One industrial data platform. All your data. Collected. Cleaned. Contextualized. Young, fast-growing tech company. More than 100 strong. Industrial expertise and technological expertise. Backed by Aker, the largest industrial group in Norway. Long-term financial foundation and legacy of experience. 2

3 Digitalization is the strategic transformation of business models and activities through digital technologies and liberation of data flows.

4 Beyond B2B, systematic use of vast pools of data combined with machine learning can unlock new value across industries 4

5 Pervasive, low cost Sensor Collaborative Technology Platforms Unmanned Aerial Vehicle Virtual Reality 3D Scanning Real-time Communication and Asset Tracking Algorithms & Models Mobile Connectivity & AR Additive Manufacturing Cloud Computing / Storage Robotics & automation Big Data & Analytics Cyber Security 5

6 What prevents wide adoption of digitalization? What we hear when asking about digitalization: Case Study Project Pilot Test Run Proof of Concept Where are the large scale operationalization projects? SOURCE: Customer dialogues 6

7 A world of data silos Oil Services provider End user 100 operational data systems 4,800 active integrations Applications Dashboards & Reporting Analytics, Optimization & Machine Learning 3D/AR/VR 500 connections rebuilt every year Large O&G Operator Less than 60% of 4M sensor IDs can be mapped to equipment Backfilling of historical data impossible at scale Data sources 7

8 Liberate your data Time spent on Data Science projects Building models 80% to 90% is spent gathering and cleaning the data not building models. Google estimates the time lost to 75% across industries. Gathering data Scalability across assets or equipment is low Time to deployment is lengthy and costly 8

9 Practical implications Increased risk of errors Costly resources involved Unnecessarily lengthy process 9

10 Cognite holds a complete digital representation of the industrial reality both immediate and past

11 1 Break the silos An open, uniform way to access all industrial data Applications Data layer Standard access regardless of assets HORIZONTAL DATA PLATFORM Data sources 11

12 2 Contextualize the data to be able to extract insights Other context: weather, satellite images, maps++ Maintenance logs, ERP data 100,000 sensor tags Operational data in 50+ applications 3D models P&IDs Equipment hierarchy Sensor metadata Type Compressor Manufacturer WILCO S/N Bar, C, F, Hz Sensor values 12

13 3 Enable a thriving, operational application ecosystem Applications Predictive Maintenance Production Optimization Remote monitoring Simulations and visualization Alliance Partners Data layer Standard access regardless of assets HORIZONTAL DATA PLATFORM Data sources 13

14 Real-world use cases

15 "Aker BP is collaborating with Cognite to make our data a strategic resource for accelerated performance, innovation and decision making. This partnership is a key enabler for our quest to digitize the E&P value chain. Karl Johnny Hersvik, CEO AkerBP 15

16 Customer reference Data liberation for all Aker BP operational data All operational data from all Aker BP operations available in the system instantly (close to 250,000 sensors, 4 Tn data points historical and live). Backfill performed in weeks from historians. All live data available in < 700ms Skarv Combined with contextual data: P&ID process files Maintenance modules Alvheim Ivar Aasen ERP D&W logs Valhall/Hod Ula/Tambar Johan Sverdrup 3D/CAD assets Tag Mapping Etc. 16

17 The Cognite Operational Intelligence application provides real time access to operational data in context Democratize access to 3D models and operational data 17

18 Accelerated insights through data and API openness 3rd party applications Advanced analytics Arundo analytics downloaded all required data for their models in 6 hours Implementation design meeting concluded after 15 min, when model was approved GeologIQ got access to relevant data through Cognite APIs Delivered improved application with real time data the very same day Integrity monitoring of risers 3D Models Aker Solutions used less than 1 day to upload and embed 3D models of their equipment in their application QVI delivered to Aker BP Solution Seeker slug prediction No need for costly 3rd party static reports Prolonged maintenance intervals Reduced risk of failure on critical equipment Better slug handling and prediction capabilities Easy to use decision support applications aimed at operators Algorithms for early warnings of imminent slugging 18

19 The Cognite platform in action

20 Thank you! An industrial world operated by algorithms, freeing human creativity for greater pursuits. 20