how a global manufacturing company built a data science capability from scratch

Post on 21-Mar-2017

435 Views

Category:

Data & Analytics

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

How a global manufacturing company built a data science

capability from scratch@carlotorniai

Head of Data Science and Analytics Pirelli

Outline§ Why Data Science and Analytics in Pirelli § What did we do differently § Lessons learned

Pirelli§ The 5th world’s largest tyre

manufacturer

§ Leader in the Premium and Prestige market

§ Only supplier of Formula 1 tyres

§ The Calendar

Why Data Science and Analytics in Pirelli?§ Capitalize on the amount of

data available

§ Build services around data

§ Drive a cultural change

Main clusters of activities

Smart Manufacturing Integrated value Chain Demand forecasting

Services built on top of Cyber Technologies

What didn’t work before?

§ Tech-centered approach within IT

§ Old approach: client - supplier relationship

§ Core competence outsourced

What did we do differently?

§ People - Org structure and team

composition

§ Process - Agile to break silos

§ Technology - Right tools for the task

People: outside the company grid§ Start up

§ Outside ICT

§ Reporting directly to the CTO

People: insource the right talents§ Diversity of backgrounds

§ Small and flat organization

§ Be as much “independent” as you can across the full DS spectrum

Process: agile to break silos§ Transparency and trust

§ Break the contract game

§ Dealing with uncertainty

§ Cross team and cross hierarch interaction

Process: how to stick around§ Business driven

§ Have clear KPIs

§ Identify actionable items

§ Redefine the “idea” be data driven

Technology: right tools for the task§ It’s never about the tools

(first)

§ Democratising data and enable smart data interaction at every level of the organization

§ Choose the right tools at the right time

Tech stack and architecture evolution

MES Local repo

HadoopCluster

ETLpipelines

Pirelli VPC AWS Factory

Tech stack and architecture evolutionPirelli VPC AWS Factory

MES Local repo

HadoopCluster

ETLpipelines

AnalyticsInfrastructure

Tech stack and architecture evolutionPirelli VPC AWS Factory

MES Local repo

HadoopCluster

ETLpipelines

AnalyticsInfrastructure

Local analytics Infrastructure

Data ProductsDev & Deploy

Tech stack and architecture evolutionPirelli VPC AWS Factory

MES Local repo

HadoopCluster

ETLpipelines

AnalyticsInfrastructure

Local analytics Infrastructure

Data ProductsDev & Deploy

Issue trackingNotification

SmartAlerts

Tech stack and architecture evolutionPirelli VPC AWS Factory

MES Local repo

HadoopCluster

ETLpipelines

AnalyticsInfrastructure

Local analytics Infrastructure

Data ProductsDev & Deploy

Issue trackingNotification

SmartAlerts

Requirements§ Top management

commitment

§ Integration with the business

§ Relations with IT

Challenges§ Expectations and portfolio

management

§ Recruit and maintain talents

§ Resistance to change

top related