Top Banner
Self-assembly Strategies - an ascens case study - Andrea Vandin 10th-16th June 2012 Edinburgh ascens autonomic service component ensembles
33
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Self-assembly Strategies

Self-assembly Strategies

- an ascens case study -

Andrea Vandin

10th-16th June 2012 Edinburgh

ascensautonomic service component ensembles

Page 2: Self-assembly Strategies

10th-16th June 2012 Edinburgh

1. ASCENS and its case studies

2. Robotics self-assembly strategies

3. Proposals for possible projects

Self-assembly Strategies

Page 3: Self-assembly Strategies

ascensautonomic service component ensembles

Page 4: Self-assembly Strategies

“...open-ended, highly parallel, massively distributed systems...

...made of self-aware, self-adaptive, self-expressive, self-*, autonomic components.”

- ascens-ist.eu

ensembles

Page 5: Self-assembly Strategies

“...build ensembles in a way that combines

software engineering, formal methods and autonomic, adaptive, self-aware systems.”

- ascens-ist.eu

the goal of ascens

Page 6: Self-assembly Strategies

Robot Swarms

E-vehicles

Page 7: Self-assembly Strategies

Swarm robotics & s-botsscenario

z

limited battery life

limited tolerance to radiations

Page 8: Self-assembly Strategies

Swarm robotics & s-botsscenario

Page 9: Self-assembly Strategies

Swarm robotics & s-botsscenario

Page 10: Self-assembly Strategies

Simple informationlocal perceptionunknown environment

Simple individual behaviour

Simple communication means

Complex global behaviour

Swarm robotics & s-botsmain characteristics

Page 11: Self-assembly Strategies

Swarm robotics & s-botschallenges

What is a good strategy forexploration?building?assembly/transportation?

What do those strategies need in terms of...knowledge/awareness?adaptation mechanisms?

How can we analyse those strategies?

Page 12: Self-assembly Strategies

10th-16th June 2012 Edinburgh

1. ASCENS and its case studies

2. Robotics self-assembly strategies

3. Proposals for possible projects

Self-assembly Strategies

Page 13: Self-assembly Strategies

Self-assembly strategies

Auton. Robots 28(4): 439-455 2010

Video: 3 s-bots crossing a holehttp://sysma.lab.imtlucca.it/tools/ensembles/

Courtesy of Rehan O'Grady (IRIDIA@ULB)

Page 14: Self-assembly Strategies

Basic self-assembly strategy

Page 15: Self-assembly Strategies

IP

AP AGG

SA AS

W CP

Move towards light

Move away from light

Grab an admissible LED

Move in any direction

Move in any directionidle

Move towards light

Basic behaviours

Page 16: Self-assembly Strategies

IP AP

Transitions

?

? ?

? ?

?

?

IP AP?

? ?

? ?

?

?

IP AP

IP

AP

see hole orsee green orsee red

Page 17: Self-assembly Strategies

10th-16th June 2012 Edinburgh

1. ASCENS and its case studies

2. Robotics self-assembly strategies− Analysis

3. Proposals for possible projects

Self-assembly Strategies

Page 18: Self-assembly Strategies

Discrete event simulations

Video: simulating 3 s-bots crossing a holehttp://sysma.lab.imtlucca.it/tools/ensembles/

Page 19: Self-assembly Strategies

CP

CP

Bizarre behaviours

CP

Page 20: Self-assembly Strategies

Prob(become seed) andclose to green anddon't see red

IP

AP AGG

SA AS

W CP

see hole orsee green orsee red

Assembled

See red

Timeout

See hole

Don't see green

Timeo

ut

Bizarre behaviours

Close t

o red

Page 21: Self-assembly Strategies

Prob(become seed) andclose to green anddon't see red anddon't see hole

IP

AP AGG

SA AS

W CP

see hole orsee green orsee red

Assembled

See red ortimeout andnot gripped

Timeout

See hole andnot see red

Don't see green

Timeo

ut an

d

gripped

Bizarre behaviours

Close t

o red

Page 22: Self-assembly Strategies

Execute enough simulations of probabilistic systems, statistically estimating properties, given a confidence interval”

✔ No state-space explosion

✔ Quantitative properties

✗ No absolute confidence on results as model checking

✗ Statistical errors

CALCO 2011: 386-392

Statistical Model Checking

Page 23: Self-assembly Strategies

Quantitative analysis“What is the expected percentage of robots reaching the goal?”

vs

~ 33%

~ 50%

WHY?

Page 24: Self-assembly Strategies

Quantitative analysis“What is the expected percentage of robots reaching the goal?”

Page 25: Self-assembly Strategies

Quantitative analysis“What is the expected percentage of robots reaching the goal?”

Page 26: Self-assembly Strategies

Quantitative analysis“What is the expected percentage of robots reaching the goal?”

Page 27: Self-assembly Strategies

10th-16th June 2012 Edinburgh

1. ASCENS and its case studies

2. Robotics self-assembly strategies

3. Proposals for possible projects

Self-assembly Strategies

Page 28: Self-assembly Strategies

Proposals for possible projects

Literature review

High level (visual/graphical) strategies definition

Strategies implementation and analysis

Page 29: Self-assembly Strategies

Literature review

High level (visual/graphical) strategies definition

Strategies implementation and analysis

Proposals for possible projectsClaytronics@CMU,Intel

CKBots@ModLab

Morphogenesis@Iridia

@MIT

Page 30: Self-assembly Strategies

Literature review

High level (visual/graphical) strategies definition

Strategies implementation and analysis

“impose” particular morphologies … line? find&solve “bizzarre” behaviours? …

“impose” particular morphologies … line? find&solve “bizzarre” behaviours? …

Proposals for possible projects

Page 31: Self-assembly Strategies

We will discuss this tomorrow!

Page 32: Self-assembly Strategies

Thanks!

Page 33: Self-assembly Strategies

[email protected]/andrea.vandin