1 Self -Or ganizat ion - St igmergy - Advant ages Charact erist - PDF document

Last t ime Self -Organizat ion ❒ Cellular aut omat a ❒ Pat t ern ❍ One-dimensional ❍ A par t icular, or ganized ar rangement of obj ect s ❍ Wolf r am’s classif icat ion in space or t ime ❍ Langt on’s lambda par amet er ❒ I nt eract ions ❍ Two-dimensional ❍ Based on local inf or mat ion only - no global • Conway’s Game of Lif e inf ormat ion ❒ Pat t ern f ormat ion in slime molds ❍ P hysical laws ❍ Dict yost elium discoideum ❍ Genet ically cont r olled proper t ies of t he ❍ Modeling of pat t er n component s 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 1 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 4 Out line f or t oday Self -Or ganizat ion - I ngredient s ❒ Self -Organizat ion ❒ Posit ive f eedback ❍ Act ivit y amplif icat ion ❒ Aut onomous Agent s ❒ Negat ive f eedback ❒ Real Ant s ❍ Act ivit y balancing ❒ Virt ual Ter mit es ❒ Amplif icat ion of random f luct uat ions ❒ Virt ual Ant s ❒ Mult iple int eract ions ❒ Ant Algorit hms 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 2 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 5 Self -Organizat ion Self -Or ganizat ion - I nf ormat ion ❒ Signals ❒ ”Self -organizat ion is a process in which ❍ St imuli shaped by nat ural select ion specif ically pat t ern at t he global level of a syst em t o convey inf ormat ion emer ges solely f rom numer ous int eract ions ❒ Cues among t he lower-level component s of t he ❍ St imuli t hat convey inf ormat ion only incident ally syst em. Moreover, t he r ules specif ying ❒ Gat hered f rom one’s neighbors int eract ions among t he syst em’s ❍ St imuli-r esponse, simple behavioral r ules of component s are execut ed using only local t humb inf ormat ion, wit hout ref erence t o t he ❒ Gat hered f r om work in progr ess global pat t ern.” – Camazine et al, p. 8 ❍ St igmer gy ❍ Random f luct uat ion and chance het er ogeneit ies 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 3 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 6 1

Self -Or ganizat ion - St igmergy - Advant ages Charact erist ics ❒ Dynamic syst ems ❒ Permit e simpler agent s ❒ Exhit emer gent propert ies ❒ Decrease direct communicat ion bet ween ❍ At t ract or s agent s ❍ Mult ist abilit y ❒ I ncrement al improvment ❍ Bif ur cat ions ❍ P ar amet er t uning ❒ Flexible, since when environment changes, ❍ Environment al f act ors agent s respond appropriat ely ❒ Adapt ive syst ems ❒ Dif f erent pat t erns may result f rom t he same mechanism ❒ Simple rules, complex pat t erns 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 7 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 10 Aut onomous Agent Self -Or ganizat ion – Alt ernat ives ❒ Cent ral leader ❒ ”a unit t hat int eract s wit h it s environment (which probably consist s of ot her agent s) ❍ Need ef f ect ive communicat ion and cognit ive abilit ies ❒ but act s independent ly f rom all ot her ❒ Blueprint s agent s in t hat it does not t ake commands ❍ Most be st ored f rom some seen or unseen leader , ❒ Recipes ❒ nor does an agent have some idea of a ❍ Hinder s f lexibilit y global plan t hat it should be f ollowing.” ❒ Templat es - Flake, p. 261 ❍ Must be avaiable 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 8 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 11 St igmergy Real Ant s ❒ A recursive cont rol syst em ❒ I magine if art if icial syst ems could do t he t hings ant s can do? ❒ Ef f ect ive f or coor dinat ion in space and t ime ❒ Why ant s? ❍ Amazonas: 30% of biomass is ant s/ t ermit es ❒ A sequence of qualit at ively dif f erent st imulus-response behavior s ❍ Amazonas: dr y weight of social insect s is f our t imes t hat of ot her land animals ❒ Two t ypes: ❍ Eart h: ~10% of t ot al biomass (like humans) ❍ Qualit at ive st igmergy ❍ Quant it at ive st igmer gy 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 9 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 12 2

Army Ant s Harvest er Ant s ❒ 100 000s in colony ❒ Find shor t est pat h t o ❒ Cr eat e t empor ar y f ood ”bivouacs” ❒ Pr iorit ize f ood sour ces ❒ Act like unif ied ent it y based on dist ance and ease of access (Picture from The Texas A&M University System) (Pictures from AntColony.org) 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 13 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 16 Fungus-Gr owing Adapt ive P at h Ant s Opt imizat ion ❒ "A Leaf Cut t er Colony can st r ip t he t allest of t r ees in a single day. Equivalent consumpt ion of a f ull grown cow in t he same t ime!" ❒ ”Cult ivat e” f ungi under ground ❒ Fert ilize wit h compost f rom chewed leaves (Pictures from AntColony.org) 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 14 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 17 Fungus Cult ivat or Nest Virt ual Termit es ❒ The assigment ❒ Why does t he number of piles decrease? ❒ How t o improve t he perf or mance wit h t wo t ype of t ermit es and t wo t ype of chips? ❒ How does dest royers af f ect t he syst em? (Picture from AntColony.org) 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 15 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 18 3

Langt on’s Virt ual Ant s Virt ual Ant s - Conclusion ❒ Grid wit h whit e or black squares ❒ Even simple, reversible local behavior can lead t o complex global behavior ❒ Virt ual ant s can f ace N, S, E, W ❒ Such complex behavior may creat e ❒ Behavioral rule: st r uct ures as well as apparent ly random ❍ Take a st ep f orwar d behavior ❍ if on a whit e square then paint it black and t urn 90º right ❍ if on a black squar e then paint it whit e and t urn 90º lef t 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 19 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 22 Virt ual Ant s - Example Ant Algorit hms ❒ Ant colony opt imizat ion (ACO) ❒ Developed in 1991 by Dorigo (PhD dissert at ion) in collaborat ion wit h Colorni and Maniezzo 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 20 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 23 Virt ual Ant s – Time Reversibilit y Summar y ❒ Virt ual ant s are t ime-rever sible ❒ Self -Organizat ion ❒ But , t ime-reversibilit y does not imply ❒ Aut onomous Agent s global simplicit y ❒ Real Ant s ❒ Even a single virt ual ant int eract s wit h it s ❒ Virt ual Ter mit es own prior hist ory ❒ Virt ual Ant s ❒ Demonst rat ion ❒ Ant Algorit hms 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 21 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 24 4

Next t ime ❒ Flocks, Herds, and Schools ❒ Boids 19/11 - 04 Emergent Systems, Jonny Pettersson, UmU 25 5