A vision on collaborative computation of things for personalized analyses

52
A Vision on Collaborative Computation of Things for Personalized Analyses Dr. Eng. Sc. Justyna Zander SIMULATEDWAY, Harvard University Copyrights reserved. © 2012

description

Presentation delivered at the 3rd IEEE Track on Collaborative Modeling & Simulation - CoMetS'12. Please see http://www.sel.uniroma2.it/comets12/ for further details.

Transcript of A vision on collaborative computation of things for personalized analyses

Page 1: A vision on collaborative computation of things for personalized analyses

A Vision on Collaborative Computation of Things

for Personalized AnalysesDr. Eng. Sc. Justyna Zander

SIMULATEDWAY, Harvard University

Copyrights reserved. © 2012

Page 2: A vision on collaborative computation of things for personalized analyses

Agenda

� Science and Interdisciplinarity

� Democratizing Computation, Modeling and Simulation

� System Analysis Case Study

� Collaborative Technical Engine

� Simulation Engine Semantics

Page 3: A vision on collaborative computation of things for personalized analyses

Science Interdisciplinarity

Page 4: A vision on collaborative computation of things for personalized analyses

Dr. Justyna Zander - MBD for CPS4

One Scientific Discipline

… still Stable and Quiet

Page 5: A vision on collaborative computation of things for personalized analyses

Dr. Justyna Zander - MBD for CPS5

What if Science Disciplines start interacting…

… to create Big Data

Page 6: A vision on collaborative computation of things for personalized analyses

Dr. Justyna Zander - MBD for CPS6

Science Disciplines Interact to create Emerging Behavior …

Page 7: A vision on collaborative computation of things for personalized analyses

Movie… and Dynamic Behavior of the System

Page 8: A vision on collaborative computation of things for personalized analyses

The Personalized Mirror

of Human Life

Health

Record

Lifestyle

Interests

Travels Social

Network

Emotional

Intelligence

Genetics

DNA

Body

Building

Nutrition

Habits

Education

Knowledge

Wisdom

Information for the

Benefit of an

Individual

Brain

Capacity

Geolocation

Pollution

Collaborative Analysis

Page 9: A vision on collaborative computation of things for personalized analyses

Computation, Modeling and Simulation to the Rescue

9

Page 10: A vision on collaborative computation of things for personalized analyses

Agenda

� Science and Interdisciplinarity

� Democratizing Computation, Modeling and Simulation

� System Analysis Case Study

� Collaborative Technical Engine

� Simulation Engine Semantics

Page 11: A vision on collaborative computation of things for personalized analyses

Science Democratization

Page 12: A vision on collaborative computation of things for personalized analyses

12

Page 13: A vision on collaborative computation of things for personalized analyses

Computation of Things

� Mission: increase sustainable wellbeing and happiness

� Vision: increase personal awareness in any possible aspect

of life based on:

Page 14: A vision on collaborative computation of things for personalized analyses

A Future Personalized Virtual Advisor

User

Interface

CoTh

SYNTHESIS: Forecast and Prediction

ANALYSIS

Participatory Sensing

Physical Systems, Infrastructure, etc

Geolocation, Patterns,

Assessment criteria,

!

Intelligent Modeling

and Simulation

for Sustainability

?

Page 15: A vision on collaborative computation of things for personalized analyses

Agenda

� Science and Interdisciplinarity

� Democratizing Computation, Modeling and Simulation

� System Analysis Case Study

� Collaborative Technical Engine

� Simulation Engine Semantics

Page 16: A vision on collaborative computation of things for personalized analyses

System Analysis

Page 17: A vision on collaborative computation of things for personalized analyses

What should I do to attain

in 2 years from now

a cyclist performance

of Armstrong’s performance from 2004?

Page 18: A vision on collaborative computation of things for personalized analyses

Individual

YOU!

Page 19: A vision on collaborative computation of things for personalized analyses

Group Dynamics

Page 20: A vision on collaborative computation of things for personalized analyses
Page 21: A vision on collaborative computation of things for personalized analyses
Page 22: A vision on collaborative computation of things for personalized analyses
Page 23: A vision on collaborative computation of things for personalized analyses
Page 24: A vision on collaborative computation of things for personalized analyses
Page 25: A vision on collaborative computation of things for personalized analyses
Page 26: A vision on collaborative computation of things for personalized analyses

Simulation findings

� The cyclist who finished second in 2004 was reported to

be 5 cm taller than Lance Armstrong.

� If the body height of the virtual cyclist is increased

from 179 cm to 184 cm, the model simulation

predicts that the time needed for the time trial becomes

about 3 s longer.

� This illustrates that small differences in body size can

have significant impact on athletic performance.

Page 27: A vision on collaborative computation of things for personalized analyses

Agenda

� Science and Interdisciplinarity

� Democratizing Computation, Modeling and Simulation

� System Analysis Case Study

� Collaborative Technical Engine

� Simulation Engine Semantics

Page 28: A vision on collaborative computation of things for personalized analyses

Collaborative Technical Engine

Page 29: A vision on collaborative computation of things for personalized analyses

Agenda

� https://www.brainshark.com/innocentive/vu?pi=zGtzus4Gsz4IX

8z0&dm=5&tb=0&bg=707070

� http://www.youtube.com/watch?v=G8nlFN17D8E&feature=rel

ated

Page 30: A vision on collaborative computation of things for personalized analyses

Collecting

Analyses

Crowd-sourcing M&S

Engine Infrastructure

and Architecture

Engine

Prediction Query

Predictions

Transformations

Technical Engine Vision

Page 31: A vision on collaborative computation of things for personalized analyses

Domain Expert

Simulation Tool Expert

Infrastructure Expert

Simulation Analyst Expert

Business Analyst Expert

Process Analyst Expert Mass-Scale User

Citizen Analyst

Page 32: A vision on collaborative computation of things for personalized analyses

Human in the Loop

Modeling and Simulation

Ubiquitous Computing

Internet of Things

Wisdom of Crowds

Participatory Sensing

Computational Thinking

Computation Engineering

Engineering Sustainable Development and Human Awareness

Reciprocatory Sensing (AI)

A Merge of Different Approaches

Page 33: A vision on collaborative computation of things for personalized analyses

Modeling and Simulation as a Collaboration and Technology Core

Page 34: A vision on collaborative computation of things for personalized analyses

Multi-disciplines

Page 35: A vision on collaborative computation of things for personalized analyses

Cloud

Technology 1

Technology 2

Technology 3

Users Users

Collection

Models Models

Collection

Big Data Big Data

Collection

Prediction

Collection

Prediction

Queries

Collection

Problems Problems

Collection

Transfor

Collection

Transfor-

mations

Collection

Virtual instance on a

User’s Device 1

Virtual instance on a

User’s Device 2

Virtual instance on a

User’s Device 3

NetworkNetwork

Engine Architecture and Usage Process

Page 36: A vision on collaborative computation of things for personalized analyses

Engine Architecture and Usage Process

Page 37: A vision on collaborative computation of things for personalized analyses

Views

Platform Design

Tech

nolo

gy M

anag

em

ent

Pro

cess

IT Resources

Cloud-based Tools

Guidelines for Users

Technology Management System

Technical Execution

Page 38: A vision on collaborative computation of things for personalized analyses

Cross-sections

Guidelines for Users

Views

Platform Design

Tech

nolo

gy M

anag

em

ent

Pro

cess

Collaborative Platform Objectives

Problem

Man

agem

ent

Pro

cess

M&

S Pro

ject

Man

agem

ent

Pro

cess

Definition

Big Data

Models

People

Solution

Existing Big Data

Available Models

People

Predictions

IT Resources

Cloud-based Tools

Guidelines for Users

Transformations

Simulation

Expertise

Collaborative Knowledge Management SystemTechnology Management System

Technical Execution

User’s Prediction Query

Page 39: A vision on collaborative computation of things for personalized analyses

Cross-sections

Guidelines for Users

Views

Platform Design

Tech

nolo

gy M

anag

em

ent

Pro

cess

Collaborative Platform Objectives

Problem

Man

agem

ent

Pro

cess

M&

S Pro

ject

Man

agem

ent

Pro

cess

Definition

Big Data

Models

People

Solution

Existing Big Data

Available Models

People

Predictions

IT Resources

Cloud-based Tools

Guidelines for Users

Transformations

Simulation

Expertise

Collaborative Knowledge Management SystemTechnology Management System

Technical Execution

User’s Prediction Query

Page 40: A vision on collaborative computation of things for personalized analyses

Cross-sections

Guidelines for Users

Views

Platform Design

Tech

nolo

gy M

anag

em

ent

Pro

cess

Collaborative Platform Objectives

Problem

Man

agem

ent

Pro

cess

M&

S Pro

ject

Man

agem

ent

Pro

cess

Definition

Big Data

Models

People

Solution

Existing Big Data

Available Models

People

Predictions

IT Resources

Cloud-based Tools

Guidelines for Users

Transformations

Simulation

Expertise

Collaborative Knowledge Management SystemTechnology Management System

Technical Execution

User’s Prediction Query

What should I do to attain

in 2 years from now

a cyclist performance

of Armstrong’s performance from 2004?

Page 41: A vision on collaborative computation of things for personalized analyses

Cross-sections

Guidelines for Users

Views

Platform Design

Tech

nolo

gy M

anag

em

ent

Pro

cess

Collaborative Platform Objectives

Problem

Man

agem

ent

Pro

cess

M&

S Pro

ject

Man

agem

ent

Pro

cess

Definition

Big Data

Models

People

Solution

Existing Big Data

Available Models

People

Predictions

IT Resources

Cloud-based Tools

Guidelines for Users

Transformations

Simulation

Expertise

Technical Execution

User’s Prediction Query

What should I do to attain

in 2 years from now

a cyclist performance

of Armstrong’s performance from 2004?

Cyclist performance

L. Armstrong

Biomechanics,BiochemistryBody and musclePerformanceGroup dynamicsGeography of the

raceDisease track

Brain capacity

Simulation

community

J. Ullrich

Myself – statistics

Statistics in the

Web

Models in the Web

Page 42: A vision on collaborative computation of things for personalized analyses

Cross-sections

Guidelines for Users

Views

Platform Design

Tech

nolo

gy M

anag

em

ent

Pro

cess

Collaborative Platform Objectives

Problem

Man

agem

ent

Pro

cess

M&

S Pro

ject

Man

agem

ent

Pro

cess

Definition

Big Data

Models

People

Solution

Existing Big Data

Available Models

People

Predictions

IT Resources

Cloud-based Tools

Guidelines for Users

Transformations

Simulation

Expertise

Technical Execution

User’s Prediction Query

What should I do to attain

in 2 years from now

a cyclist performance

of Armstrong’s performance from 2004?

Cyclist performance

L. Armstrong

Biomechanics,BiochemistryBody and musclePerformanceGroup dynamicsGeography of the

raceDisease trackBrain capacity

Simulation

community

J. UllrichMyself – statisticsStatistics in the

Web

Models in the Web

Page 43: A vision on collaborative computation of things for personalized analyses

Agenda

� Science and Interdisciplinarity

� Democratizing Computation, Modeling and Simulation

� System Analysis Case Study

� Collaborative Technical Engine

� Simulation Engine Semantics

Page 44: A vision on collaborative computation of things for personalized analyses

Simulation Engine Semantics

Page 45: A vision on collaborative computation of things for personalized analyses

System

Execution

Engine

Implemen-

tation

System

Model

Implemen-

tation

State of the Past

Computational Framework in the Past

System Model Execution

Engine

Page 46: A vision on collaborative computation of things for personalized analyses

State of the Art

Computational Framework Nowadays

Model

Specification

Model

Implemen-

tation

System Model

Solver

Implemen-

tation

Solver

Simulation

Page 47: A vision on collaborative computation of things for personalized analyses

State of the FutureFuture

Computational Framework

Model

Specification

Solver

Implemen-

tation

Model

Implemen-

tation

Solver

Model

Specification

Simulation Runtime Interface

Solver

Implemen-

tation

Solver

Implemen-

tation

Model

Implemen-

tation

Model

Implemen-

tation

Simulation

Ve

rific

atio

n a

nd

Va

lida

tion

Ve

rific

atio

n a

nd

Va

lida

tion

Page 48: A vision on collaborative computation of things for personalized analyses

A Computational Framework

Analysis &

Synth

esis

TECHNOLOGY

IMPLEMENTATION

Computational Framework User

OPERATIONAL DEFINITION

DECLARATIVE DEFINITION

MODELING

Analysis, Syn

thesis, &

Execu

tion

SimRI

Legend:

SimCI – Simulation Control Interface

SimRI – Simulation Runtime Interface

ExPM– Execution Platform Mapping

SimCI

SPEC

IFIC

AT

ION

MODELINGMODELING

ExPM

Platform

Page 49: A vision on collaborative computation of things for personalized analyses

Declarative Specification of the Solver

Page 50: A vision on collaborative computation of things for personalized analyses

Non-Monotonous Time Notion in Solver

computational evaluation index

time

step size

accepted time step

rejected time step

Page 51: A vision on collaborative computation of things for personalized analyses

Agenda

� Science and Interdisciplinarity

� Democratizing Computation, Modeling and Simulation

� System Analysis Case Study

� Collaborative Technical Engine

� Simulation Engine Semantics

Page 52: A vision on collaborative computation of things for personalized analyses

Continuous Awareness