제조업 사례로 살펴본 “big data roi 측정 및...
TRANSCRIPT
제조업 사례로 살펴본 “Big Data ROI 측정 및 투자”
김기남 Manager / SK 텔레콤 Data사업본부
Consideration for Big Data in Manufacturing
Areas of Greatest Benefits for Manufacturing
Source: A TCS 2013 Global Trend Study
#1. Big Data 투자, ROI에 대한 고민
ROI Calculator
Source: TDWI
ROI Calculator
Company Profile
Labor Cost
HW, SW Cost
Direct Cost Savings
Productivity
Revenue Increase ROI
ROI 등식
Cost of Investment - Gain from Investment = ROI
Server Perpetual Licenses
Storage ) (
Support Subscription Licenses Network and Security
Customizations Cloud
( )
CapEx OpEx Revenue
Optimization Cost
Optimization + - -
) New Revenue
Existing Revenue + (
Employee Average Cost per Hour x Total Number of Increased Productivity Hours ( )
+ Human Capital Gain
Existing Solution Gain + ( )
Running Costs Efficiency + Value of Minimized Revenue Risk ( )
Revenue Optimization
Big Data ROI에 대한 OTL
#2. Big Data Infra 투자를 통한 현재 가치만 계산
자동제어 / 실행
Big 데이터 분석
데이터 센싱/수집
지능형 의사결정
제조 Intelligence를 첫 단계
Bottom Line Contribute
Revenue Optimization
Cost Optimization
CapEx OpEx + - - ROI =
Net present value 만 고려
Bottom Line Contribute
Big Data Preparation SCM
ERP
P&S
MES
• Planning • Scheduling • Real Time
Dispatching
• Execute Prod • Inventory Mgt.
• Material Mgt. • History Mgt.
YMS
• Yield Mgt. • Quality Analysis • Statistical Analysis
Enterprise
MES
Level
생산시스템, IT부서 Data
DW
Data Lake
①
②
Ingestion
Transformation
Barriers to Entry
Data Lake
DW
#1 Application
#2 Application
#3 Application
#4 Application
#5 Application
#6 Application
Leverage existing SQL query language and existing BI tools against data within Hadoop
Barriers
But.... More Data, Bigger Challenges
Data quality rules
Data catalog Metadata
store Data lineage
Big Data Governance
• Management Scalability • Metadata Tools • Business monitoring
It's not all 100% there
OpEx
Hadoop 운영
인건비 증가
1) What do we know about our information? 2) Where did this data come from and who can use it? 3) Does this data adhere to company policies and rules?
Ingestion
Data Lake
Cleansing
Transformation
#3. Big Data 분석을 활용한 Business 미래가치 증명
자동제어 / 실행
Big 데이터 분석
데이터 센싱/수집
지능형 의사결정
제조 Intelligence를 위한 다음 단계
Business Insight Contribute
Revenue Optimization
Cost Optimization
CapEx OpEx + - - ROI =
? ? ? ? Biz 가치 증명 예산 부족 인력 부족 Biz 가치 증명
?
Investment
Small Start with Small Team
PoC
Dot’ know the Result
Small Team Exploration Test & Valuation
Revenue Optimization
? Biz 가치 증명
Cost Optimization
? Biz 가치 증명
Data Lake
#1 Application
#2 Application
#3 Application
DW
Data sandboxes
• Individual data views in the data lake
• Self-Discovery-BI
• Advanced Analysis
Explore Raw Data Sets
CapEx
예산 부족 Data scientists and other analysts need to be allowed to explore raw data sets in an unfettered way
Data Exploration
Data Agility
Machine Learning
Metatron for Big Data Discovery
데이터 수집
저장 & 처리
분석
시각화
데이터 패턴 탐지 / 실시간 수집
Sub-second 처리 엔진
In-memory Grid 시각화
A A A
Hadoop 에코시스템
Big Data Storage
Industry 특화 고급분석(DNN ML, etc.)
모니
터링
및 관
리/보
안
• Workspace: 개별 유저가 독립적인 분석
공간을 논리적으로 할당 받으며, 해당
공간에서 수집, 처리, 분석, 시각화를 수행
탐색 분석(Discovery/리포팅) SQL 분석 (Ad-hoc Query)
심층 분석(R, ML, Python) ETL/Workflow
Rethinking Business
분석 전문업체 무상 PoC Evaluation Investment AS - IS
TO - BE Proof of concept business
model with a partner
Evaluation Cultivate
New business with a partner
OpEx
기술 인력 부족
Revenue Optimization
Biz 가치 증명
CapEx
예산 부족
Rethinking Business with SKT
Let's put your data to work
Unlock operational efficiency
Capture new growth opportunities
Let's build something together
Solution for Business Excellence with partners
Big Data Analysis Professional Service
On-Premise or Cloud
Data Collection Data
Processing
Data Model Manager
Data Preparation
ML / DNN Algorithm
예지 정비
Appl.
품질 관리
Appl.
운영 최적화 Appl.
....