Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Standard view
Full view
of .
Save to My Library
Look up keyword
Like this
0 of .
Results for:
No results containing your search query
P. 1
Artificial Intelligence

Artificial Intelligence

Ratings: (0)|Views: 237 |Likes:
Published by ARVIND

More info:

Published by: ARVIND on Mar 30, 2010
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as DOC, PDF, TXT or read online from Scribd
See more
See less





 Technical Paper Presentation Contest
ARTIFICIAL INTELLIGENCE(Biologically Inspired Intelligent Robots Using Artificial Muscles)
 Artificial Intelligence is a branchof Science which deals with helping machines finds solutions to complex  problems in a more human-like fashion.This generally involves borrowincharacteristics from human intelligence, and applying them as algorithms in a computer  friendly way. A more or less flexible or efficient approach can be taken depending on the requirements established, whichinfluences how artificial the intelligent behavior appears. Humans throughout history have always sought to mimic the appearance, mobility, functionality, intelligent operation, anthinking process of biological creatures. This field of biologically inspired technology,having the moniker biometrics, has evolved  from making static copies of human ananimals in the form of statues to theemergence of robots that operate withrealistic appearance and behavior. This paper covers the current state-of-the-art and challenges to making biomimetic robotsusing artificial muscles.
: EAP, artificial muscles, artificialintelligence, biometrics
AI is generally associated with
Computer Science
, but it has many importantlinks with other fields such as
, among many others. Our abilityto combine knowledge from all these fieldswill ultimately benefit our progress in thequest of creating an intelligent artificial being.
Why Artificial Intelligence?Motivation...
Computers are fundamentally wellsuited to performing mechanicalcomputations, using fixed programmed rules.This allows artificial machines to performsimple monotonous tasks efficiently andreliably, which humans are ill-suited to. For more complex problems, things get moredifficult... Unlike humans, computers havetrouble understanding specific situations, andadapting to new situations. ArtificialIntelligence aims to improve machine behavior in tackling such complex tasks.
When will Computers become trulyintelligent?Limitations...
To date, all the traits of humanintelligence have not been captured andapplied together to spawn an intelligentartificial creature. Currently, ArtificialIntelligence rather seems to focus on lucrativedomain specific applications, which do notnecessarily require the full extent of AIcapabilities. Researchers know this limit of machine intelligence as narrow intelligence.There is little doubt among the communitythat artificial machines will be capable of intelligent thought in the near future. It's just aquestion of what and when. The machinesmay be pure silicon, quantum computers or hybrid combinations of manufacturedcomponents and neural tissue. As for the date,expect great things to happen within thiscentury!
How does AI work ?
There are many differentapproaches to Artificial Intelligence, none of which are either completely right or wrong.Some are obviously more suited than others insome cases, but any working alternative can be defended. Over the years, trends haveemerged based on the state of mind of influential researchers, funding opportunitiesas well as available computer hardware.Over the past five decades, AIresearch has mostly been focusing on solvingspecific problems. Numerous solutions have been devised and improved to do so efficientlyand reliably. This explains why the field of Artificial Intelligence is split into many branches, ranging from
 Pattern Recognition
 Artificial Life
, including
Artificial life through robotics:
Laws of Robotics:1. A robot may not injure a human being or,through inaction, allow a human being tocome to harm.2. A robot must obey the orders given it byhuman beings except where such orders wouldconflict with the first law.3. A robot must protect its own existence aslong as such protection does not conflict withthe first or second law.Robotics has been anevolution of the field of automation wherethere was a desire to emulate biologicallyinspired characteristics of manipulation andmobility. In recent years, significant advanceshave been made in robotics, artificialintelligence and others fields allowing to makesophisticate biologically inspired robots [Bar-Cohen and Brea zeal]. Using these advances,scientists and engineers are increasinglyreverse engineering many animals' performance characteristics. Biologicallyinspired robotics is a subset of theinterdisciplinary field of biomimetics.Technology progress resulted in machines thatcan recognize facial expressions, understandspeech, and perform mobility very similar toliving creatures including walking, hopping,and swimming. Further, advances in polymer sciences led to the emergence of artificialmuscles using Electro active Polymer (EAP)materials that show functional characteristicsremarkably similar to biological muscles.Making creatures that behave like the biological model is a standard procedure for the animatronics industry that is quite wellgraphically animates the appearance and behavior of such creatures. However,engineering such biomimetic intelligentcreatures as realistic robots is still challengedue to the need to physical and technologicalconstraints.
Artificial muscles:
Muscles are the key to the mobilityand manipulation capability of biologicalcreatures and when creating biomimetic it isessential to create actuators that emulatemuscles. The potential to make such actuatorsis increasingly becoming feasible with theemergence of the electro active polymers(EAP), which are also known as artificialmuscles [Bar-Cohen, 2001]. These materialshave functional similarities to biologicalmuscles, including resilience, damagetolerance, and large actuation strains.Moreover, these materials can be used tomake mechanical devices with no traditionalcomponents like gears, and bearings, whichare responsible to their high costs, weight and premature failures. The large displacementthat can be obtained with EAP using lowmass, low power and, in some of thesematerials also low voltage, makes themattractive actuators. The capability of EAPs toemulate muscles offers robotic capabilitiesthat have been in the realm of science fictionwhen relying on existing actuators.
FIGURE 1: A graphic illustration of the grandchallenge for the development of EAPactuated robotics an arm wrestling matchagainst human.Unfortunately, the EAP materials thathave been developed so far are still exhibitinglow conversion efficiency, are not robust, andthere are no standard commercial materialsavailable for consideration in practicalapplications. In order to be able to take thesematerials from the development phase toapplication as effective actuators, there is aneed for an established infrastructure. For this purpose, it is necessary to developcomprehensive understanding of EAPmaterials' behavior, as well as effective processing, shaping and characterizationtechniques. The technology of artificialmuscles is still in its emerging stages but theincreased resources, the growing number of investigators conducting research related toEAP, and the improved collaboration amongdevelopers, users, and sponsors are leading toa rapid progress. The Responsibility of AIResearch andDevelopment The question that concernsmany in regards to the development of machines capable of intelligence comparableto that of a human being, is that oconsequences of creating servants capable of  becoming masters. This is a question in itself containing many more: How much power willthese machines - (robots, as many like to think of them) have? Who is going to control them?What are they going to be used for? Whatresponsibilities will they carry? These are justa few of the questions, which are on the mindsof many who ever stopped to think what AI isabout.Robots, which could build other robots, tried to protect humans fromeverything until people could not do anything by themselves. They became totally dependenton the machines. The creator of these robotscould not destroy them because they wouldkeep him away from doing so. After all howcould they protect people after they would bedestroyed? How much responsibility andauthority should these machines have? Thisquestion becomes even more important whenone considers a fact that one of the biggestsupporters of AI research is military. One of the reasons is that in "...a nuclear age, ... a newgeneration of very intelligent computersincorporating AI could actually defend thecountry better, faster, and more rationally thanhumans."4 Even if this is true we do not think that a machine should handle theresponsibility for the fate of humancivilization, no matter how intelligent it is.
Biometric robots using EAP :
Mimicking nature wouldsignificantly expand the functionality of robots allowing performance of tasks that arecurrently impossible. As technology evolves,great number of biologically inspired robotsactuated by EAP materials emulating biological creatures is expected to emerge.The challenges to making such robots can beseen graphically in Figure 2 where humanlikeand dog-like robots are shown to hop andexpress joy. Both tasks are easy for humansand dogs to do but are extremely complex to perform by existing robots.FIGURE 2: Making a joyfully hoppinghuman-like and dog-like robots actuated byEAP materials are great challenges fo biomimetic robots

Activity (34)

You've already reviewed this. Edit your review.
1 hundred reads
1 thousand reads
1 thousand reads
Asif Ali Khan liked this
zucamlo liked this
trickers123 liked this
Purna Chandra liked this
divinp liked this
patilamrita liked this
Diptam Dutta liked this

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->