Artificial Life Lecture 19

1
Artificial Life Lecture 19

• Summing up!
• Survey of Artificial Life topics covered
• And Alife skills
• Work in Alife academic research ..
• and commercial ..
• Recent robotics at Sussex
• Presentation on EASy research
• Personal prejudiced list of hot research topics
• Time for questions?

2
Alife Topics covered - Evolutionary

• Evolution
• Evolutionary Algorithms
• Co-evolution
• Evolution of communication
• Development, L-Systems, G-gtP mappings
• Fitness Landscapes
• Neutral Networks

3
Alife topics evolutionary, more

• Information and Life and Evolution
• Tierra and Avida
• GP and Classifier Systems

4
Alife topics DS and robotics

• Dynamical Systems approach to cognition
• Braitenberg vehicles
• CTRNNs
• Evolutionary Robotics
• Passive Dynamic Walking

5
Alife topics beyond

• CAs and RBNs
• Models of Genetic Regulatory Systems
• Morphogenesis, L-Systems
• Homeostasis
• Gaia Theory, Daisyworld
• Artificial Chemistry
• Autocatalysis, autopoiesis

6
Alife Skills

• Programming a GA
• Microbial GA
• Programming a CTRNN
• Programming a Lego EASy-mind robot
• (coming next term) ODE physics engine
• How to prepare a project proposal
• (hopefully) how to carry through a project

7
Work in Artificial Life

• Research. In Spring term I shall give a seminar
  on How to apply for a PhD/DPhil
• Main points will be you have to find a research
  topic and a supervisor
• UK/Europe, apply in Spring, N. American system
  different
• jobs outside academia ..

8
Companies interested in Alife
BT a few years ago claimed they took 15 of all
computer science graduates in the UK surely no
longer true! BT RD labs at Martlesham (nr
Ipswich) used to have 4000 in Development and 600
in Research. Now slimmed down, sort of privatised
into cost centres and called BT Exact. EASy
contacts specifically with their Future
Technologies Group (formerly from 1996
Artificial Life Group) (Richard
Tateson) http//more.btexact.com/projects/ftg Evol
utionary computing, biological metaphors, fruit
flies and cell phones. Has funded DPhils
9
Companies interested in Alife

• Hewlett Packard - HP Labs Bristol
• BICAS Biologically-Inspired Complex Adaptive
  Systems http//www.hpl.hp.com/research/bicas/
• More recently cut down on research HP financial
  problems

10
Companies interested in EASy

• Many other big companies with research labs have
  an active interest in EASy issues eg British
  Aerospace, Logica, Nortel
• Smaller labs spinoff businesses in Sussex
  Innovation Centre
• Natural Motion www.naturalmotion.com (CEO Torsten
  Reil is EASy graduate)
• Brighton media companies
• Etc etc

11
Slides from a presentation at the Dec-2005
Informatics Research Awayday
7 slides to cover EASy research in
Informatics. Omits further CCNR research that is
more biologically oriented. Exactly a year old,
so some people and topics are out of date, but
still gives an overall picture. Expansion
happening right now New faculty/RFs
Anil Seth Luc Berthouze Thomas Nowotny
Paul Graham Andy Phillipides
12
EASy (Evolutionary and Adaptive Systems)and
Vision (in italics)
Faculty (51V) Phil Husbands Inman Harvey
Ezequiel Di Paolo Adrian Thompson Emmet Spier
(V) Paul Brown (2) David Young
Hilary Buxton
RFs (93V) Andy Philippides Lionel Barnett
Patricia Vargas Xiaobing Hu Linc Smith
Kyran Dale Jon Bird
Peter Passaro Dustin Stokes (V)Hiroyuki
Iizuka (V)Andy Wuensche (V)Peter de Bourcier
Doctoral Students (28)
Matt Bardeen Bill Bigge Thomas
Buhrmann Daniel Bush James Dyke Alice Eldridge
Chrisantha Fernando Tom Frose Ian
MacInnes Simon McGregor Alex Penn Marieke Rohde
Linc Smith Sampsa Sojakka Eric Vaughan Rob
Vickerstaff Nathaniel Virgo Rachel Wood Caroline
Ong Gary McHale Matt Quinn Andy Balaam
Helen Riabinina Julian Worby Mike
OddHayward James Mandelis Fernando Almeida e
Costa Eduardo Izquierdo-Torres (4)
Neil Robinson Kingsley Sage Aisha Thorn John
Rowston
(V) Visiting Prof/RF
Research-oriented EASy MSc- 5-10 have PhDs
before MSc, 50 go on to doctoral research,
100 ex-EASy PhDs around world
13
EASy and VisionTypical Journals and Conferences
Artificial Life, Adaptive Behavior, Evolutionary
Comp, IEEE Trans Evol. Comp, Robotics Aut Sys,
Genetic Prog Evolvable Hardware, Phil Trans Roy
Soc A, J Exp Biology, Biosystems, J Theor
Biology, J Neuroscience. Int Conf Artificial Life
(Alife), Eur Conf Artificial Life (ECAL)
Simulation of Adaptive Behavior (SAB), Int Conf
Climbing and Walking Robots (CLAWAR), IEEE
Congress Evol Comp (CEC), Genetic Evol Comp
Conf (GECCO), Parallel Prob Solving from Nature
(PPSN), Int Conf Evolvable Systems IEEE Pattern
Analysis Mach Intell, Int J Computer Vision,
Image Vision Computing. Eur Conf on Computer
Vision, Int Conf. on CV, British CV conf.
14
EASy and VisionGeneral Research Areas
Interfaces between Biology and IT/AI Evolutionary
and Adaptive Robotics, Artificial Life,
Neurocontrol, Computational modelling of Neuronal
Systems, Studies in Minimal Cognition,
Homeostasis, Dynamical Systems approach to
Cognition, Evolutionary Electronics, Insect and
Robot Navigation, Theory of Natural and
Artificial Evolution, Ecological modelling,
Hybrid wetware-silicon devices, Social
Interaction Dynamics, Creativity in robots,
Applications of adaptive technology in art and
music
Biological and computational vision
15
EASy and VisionSome core research questions
What are minimal useful models of how animals (eg
insects) with their nervous systems can perceive
the world, navigate around it, communicate? How
can we get robots to achieve comparable
capabilities? How can we use artificial evolution
to design robots, electronic circuits, drug
molecules ? How do natural/artificial
organisms/agents communicate in society/interact
in ecologies?
etcetc What low-level mechanisms are needed
to support visual control of action?
16
EASy and VisionTypical methodologies used
Collaborate with biologists cf CCNR Computer
modelling in minimal simulations Use of physics
engines, ODE Building and programming
robots Computational models of
vision Mechanical camera mount for saccades
17
EASy Current funded Research Projects
Rapid insect-like visual learning algorithms,
3y EPSRC 530K Spatially
Embedded Complex Systems Engineering 3.5y
EPSRC 620K Non-linear Media based computers
3.5y EPSRC
496K Computational Intelligence, Creativity and
Cognition 3y AHRC 335K Artificial
evolution to design robust circuits exploiting
Physics from
Microelectronics to Nanotechnology 5y
EPSRC 209K Bionics and Space Systems Design
1.4y Eur Space Agency 150K
18
EASy and VisionNew areas -- potential
collaborations
Humanoid Bipedal walking robots exploiting
passive dynamics Neutral Networks in Fitness
Landscapes for Artificial Evolution Maximum
Entropy Production principle and Life Adaptive
Text Entry methods Development of adaptive
cochlear implants Development of adaptive music
generation software/hardware ????
Development of cheap, fast, intelligent active
camera mounts.
19
From the CCNR web-page
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
Other CCNR research students not featured
Bill Bigge programmable springs (talk at
beginning of term) Simon McGregor Variadic
neural networks Fernando Almeida e Costa -- The
Morphological Turn in Cognitive Science new
routes towards a theory of meaning (Alergic
Wed) Organiser of ECAL2007 in Lisbon
29
More
James Dyke Homeostasis, Daisyworld, Rein
Control and MEP (workshop on this at ECAL2007 in
Lisbon) Nathaniel Virgo Understanding the
constraints on Ecosystems in terms of MEP, the
Maximum Entropy Production principle.
30
More
Started Oct 2006 Greg Studer understanding how
simple agents can self-assemble in structured
groups, so as to produce behaviours at different
scales Allister Furey using Evolutionary
Robotics techniques to control trajectories of
large kites, so as to extract energy from the
wind (an alternative, potentially more efficient,
form of windpower)
31
Hot research topics
Personal prejudiced list Versions of these might
be suitable for summer projects
32
Passive Dynamic Walking
This is a very embodied, dynamical systems
approach. So far PDWs have just gone down gentle
slopes under gravity. There is scope for adding
small amounts of power input for more general
walking. Matt Williamsons work, associated with
Brooks COG project, using coupled oscillators,
seems a very promising lead that could be applied
here. Eric Vaughan now doing this.
33
Neutral Networks
Wide open for research. Barnetts work shows that
in a formally defined class of (binary) fitness
landscapes full of NNs in a particular fashion,
best strategy is a population of size 11, with a
fixed number of mutations based on getting
expected proportion that are neutral as close as
possible to 1/e 37. Adrian Thompsons hardware
evolution supports this Extensions noisy fitness
evaluations ? Real valued genotypes?
34
Gaia/Maximum entropy?
Relationship between Daisyworld models of
homeostasis, and Thermodynamic ideas that
- Systems try to organise themselves to
produce entropy as fast as possible. Cf. Kay and
Schneider, 4th Law of Thermodynamics. Organised
systems can dissipate junk entropy faster, and
( speculation) Life does it better than
anything else and hence should naturally occur (
given the right circumstances)
35
Homeostasis
Ezequiels work on looking at issues of
homeostasis in minimally cognitive models seems
very important and interesting. As also does Jim
Stones work on ( roughly speaking) what
perceptual systems (eg ANNs) have to do in order
to sift out higher-level invariants from all the
noise. Roughly- (1) output of a system should
not fluctuate wildy (just echoing the noise is
pointless) But (2) Should not stay still it
should fluctuate over the long term if it is to
reflect real things happening in the world.
36
Economy
So, Jim Stone points out that outputs of of
perceptual systems should minimise (Short Term
Variance) divided by (Long Term
Variance) Minimise STV/LTV where short/long refer
to appropriate time scales. This is one way of
ensuring that an output neuron is earning its
keep. Related to Ezequiels interpretation of
homeostasis keepng neurons activations in
the middle of their sigmoids
37
A Speculation
If you evolve CTRNNs for an agent to do a task
(eg phototaxis) with an extra fitness factor
related to making sure each neuron earns its
keep e.g. Stones criteria, minimising
STV/LTV then you can hope to see homeostasis
??? (variant on Ezequiels upside-down glasses
experiments)
38
Adaptive Text Entry methods
The need for easier methods for text entry on
mobile phones, on iPods, PDAs etc. EASy style
approach, adaptive interface.
39
Autonomous Glider
Adding minimal sensors and a relatively simple
Braitenberg-like control architecture to a model
glider. Aim to get autonomous flight gaining
lift/power through ridge-soaring. Minimal
optical-horizon sensor, 3 DoF accelerometer, use
of optic flow.
40
Maintain position relative to a ridge how?
Compare the optical horizon with the gravity
horizon -- should keep you somewhere on the red
line.
But you also need to keep within appropriate
distances from ridge, on that red line.
Use optic flow? Unstructured texture of
ground interesting issues.
41
Skills needed
Skills needed within a group of 3 to 4
people Ability to control a radio-control glider
(also under power) Technical ability to assemble
controller within body Programming ability to
test control methods in simulation And above all,
initiative and drive to push the project forward,
needing commitment, time and effort..
42
Summer research projects
Please remember the range of possibilities
covered here (..and suggest more) when it comes
to planning your main summer projects. Please
also remember that the robots in the ASL are in
general available for your own independent
projects. ODE lab classes in Spring term, from
Greg Studer There will be a talk in Spring Term
on How to apply for a PhD/DPhil place
43
The End
time for more discussion ??