DWS15 - Smart City Forum - Boosting Digital Transformation - François Stephan - SystemX

SystemXBoosting Digital TransformationAre predictive technologies the next stage in the smart city development ?

Smart City Forum 19 nov 2015

CONFIDENTIEL

SystemX is part of Paris-Saclay World class Innovation Cluster

Technology Research Institute on Complex Systems Digital Engineering

2

60.000 students

11.000 researchers in 10 departments out of which 5.000 in:

- Mathematics

- Information Science & Technology

- Humanities & Social Science

two Nobel Prize winners,

six Fields medalists.

Source: MIT Technology Review

> Strong development 2010-2020 backed by

French government funding

> Quality of « French » mathematical

engineers

> Paris moving to a Start up friendly region

> Presence of Worldwide actors: Transport,

Energy, Defense & Security

Largest French

competitive cluster: Focus on digital technologies

for systems

160 world-class large

companies

480 SMEs & start ups

20 VCs & Bus. Angels

Paris-Saclay

top eight World Innovation Clusters

CONFIDENTIEL

Scientific & technological roadmap 2016-2020

3

Industrial, economical & societal benefits

Industrial, economical & societal benefits

SystemX added value applied on 4 strategic domains

SystemX added valueSystemX added value

AgilityProductivity

Competitiveness

Collaborative engineering methods, processes and software tools for complex systems

Security Availability

User Experience

New safe and secure architectures, integrating new usages, critical embedded systems and

infrastructures evolution

SustainabilitySecurity

Attractiveness

Decision-aid tools for optimal & operational planning of Smart Territories evolution

FlexibilityPerformance

Security

Digital Infrastructures architectures, protocols, algorithms and control & management software

Systems Engineering

Systems Engineering

Autonomous Transport

Autonomous Transport

Smart Territories

Smart Territories

Digital Infrastructure

Digital Infrastructure

CONFIDENTIEL

Prediction for Human Mobilty: Goals

Study and model traffic flow in multimodal road and public transportation networks

Understand how people move (spatial and temporal distributions, most visited locations…)

Predict and estimate the flow among these networks and possibly to predict the future position of a moving object (individual or a vehicle)

CONFIDENTIEL

Data Analytics for Human Mobility

Where and when people move

How far, how fast, and by what means they travel

Individual and aggregate scales

Correlations with infrastructure, demographics, and social behavior

Use of Telecom Mobile Networks, City Sensors, Connected objects, social networks data for inference

CONFIDENTIEL

Smart City Energy analytics Use Case : Smart Community

6

Data Storage (50 to 100 millions meters).

Electric Consumption data Access.

Temporal & Spatial Data Agregation (seconds/minutes/days).

Consumption prediction at short (seconds) and very short terms(minutes).

Renewable Energy Production Prediction.

Multi-sources, multi-energy planning.

Smart Community

CONFIDENTIEL

Smart City Energy analytics Data Generic Workflow

7

Data CollectionData Collection

Actions (Users, Systems,

components, etc.)

Data StorageData Storage

Data Cartography

Data Cartography

Data Management

Data QualityData Quality

Data Interface Acquisition

Data Interface Acquisition

Data CorrectionData Correction Data SecurityData Security

Data Pre-ProcessingData Analytics

Visualization Visualization Knowledge ExtractionKnowledge Extraction

Interactive Analysis

Interactive Analysis

Simulation & Modelling

Expert ModelExpert Model Statistical ModelStatistical Model

SimulationSimulation

Behaviour Improvement Suggestion

Behaviour Improvement Suggestion

Dashboard ObservationDashboard Observation

11

22

33

444’4’

55

CONFIDENTIEL

Example of Energy Consumption Prediction

8

horizon de prévision : 1 jour 19/02/2013

Prediction from observation on the past (business building)

Similarity based algorithm on a large set of past consumption data

Target: real time prediction

Prediction

Observed

Thank you for your [email protected]

Tel +33 1 69 08 05 62 – +33 6 78 68 65 70 - @FrancoisStephan

International & Development Director