Reading Lists

Put yourself in the position of a reviewer for each paper. Using the headings below, write a review for each paper (250-500 words for each paper).

State where and when each paper was published and who is the intended readership for the particular journal or conference (you will need to access the journal or conference home page to find such information).
Describe the central idea underlying the paper, the stated aim of the study and summarise the results presented. Use your own words.
What do the authors claim as the main contribution of each paper?
Who is likely to be interested in the results of these studies, and why?
Write a brief evaluation of the technical soundness and the contributions of the paper.
As a reviewer, what additional information would you give to the authors?

Bring your reviews each week for discussion. In week 2, we will discuss the readings from week 1 and look at your reviews, and so on through the semester.You may wish to revise you reviews before final assessment. The complete set of reviews must be submitted during the exam period. For details of submission and assessment criteria see the course profile.

Week 1: Introduction to Biorobotics

V. Braitenberg (1986) Vehicles: Experiments in Synthetic Psychology, MIT Press

Week 2: Sensing and Moving

R.A. Brooks (1990) Elephants Don’t Play Chess, Robotics and Autonomous Systems, vol 6, pp 3–15.
B. Webb (2002) Robots in invertebrate neuroscience, Nature 417, 16 May 2002, pp 359–363

Week 3: Walking and Swimming

R.D. Beer, H.J. Chiel, L.S. Sterling, (1990) A Biological Perspective on Autonomous Agent Design, Robotics and Autonomous Systems, vol 6, pp 169 - 186.
A.J. Ijspeert, et al. (2007) From Swimming to Walking with a Salamander Robot Driven by a Spinal Cord, Science 315, pp 1416-1420
H. Kimura, Y. Fukuoka and A. H. Cohen (2007) Adaptive Dynamic Walking of a Quadruped Robot on Natural Ground Based on Biological Concepts, The International Journal of Robotics Research 26; pp. 475 - 490

Week 4: Learning

M.A. Lebedev, J.M. Carmena, J.E. O’Doherty, M. Zacksenhouse, C.S. Henriquez, J.C. Principe, M.A.L. Nicolelis (2005), Cortical ensemble adaptation to represent actuators controlled by a brain-machine interface. J. Neurosci. 25: 4681-4693
G. Wyeth, Training a Vision Guided Mobile Robot (1998), Autonomous Robots, v.5 n.3-4, July-August 1998, p.381-394
(Not the world's most significant paper but a nice link to Braitenberg vehicle ideas).
S. Mahadevan , J. Connell, Automatic programming of behavior-based robots using reinforcement learning (1992), Artificial Intelligence, v.55 n.2-3, June 1992, p.311-365
(Mahadevan and Connell also have written a book called Robot Learning which is an essential read in this area, but the above paper is a good start.)

Week 5: Natural evolution

The complexity of multicelled organisms is due to two great forces: (1) development, which during an individual's lifetime transforms a single cell to an adult form; and (2) evolution, which over many generations enables the developmental process to experiment with different ways to grow a body. The study of both together is called evo devo. The key to understanding evo devo is to understand heritability - how genes code for the core properties of development in a species, and how they also allow for the variation of individual members.

S.B. Carroll, (2005). "Endless forms most beautiful: the new science of evo devo", NY: WH Norton. [you are welcome to borrow my copy if you can't find another copy]

Background reading (doesn't need a written review):
Chapters 3 and 4. Use these chapters and any other material you choose to understand the basics of biology, in particular the process of development from DNA to RNA to protein within each cell (Fig 3.2), the role of homeobox genes (Fig 3.5) and the general logic of embryo geography (Fig 4.4).

Assigned reading (for written review):
Chapter 7. Little bangs: wings and other revolutionary inventions.

Week 6: Artificial evolution

The idea of using a genome, mutation and selection has been adopted as a computational paradigm and used extensively in engineering. The basic idea of using generations of designs, with variation and selection at each generation is simple to understand. The challenge in practice concerns the choice of representations, how to introduce variation, and effective fitness functions. The subtle aspects are important to making an evolutionary algorithm effective, but they are beyond the scope of this week's work. The goal is to understand the basics of algorithms and how they can be applied to robotics.

Background reading (doesn't need a written review):

Find definitions for evolutionary algorithm, artificial genome, mutation operator and fitness function. Review a simple evolutionary algorithm (EA) for a mutant-champ paradigm (also called 1+1 EA, or stochastic hill-climber). (If you can't find enough online, for a set of good definitions, see chapter 1 in M. Mitchell, 1999, An introduction to genetic algorithms, Boston: MIT Press)

Assigned reading (for written review):

H. Lipson, (2005). Evolutionary robotics and open-ended design automation. Biomimetics, CRC Press (Bar Cohen, Ed.) pp. 129-155. (Available from http://ccsl.mae.cornell.edu/papers/Biomimetics05_Lipson.pdf)
J. Bongard, V. Zykov, and H. Lipson, (2006). Resilient machines through continuous self-monitoring, Science, 314, 1118-1121.

Week 7: Navigation (robots)

H. Durrant-Whyte and T. Bailey, “Simultaneous localization and mapping: Part I,” IEEE Robotics and Automation Magazine, pp. 99–108, JUNE 2006.
T. Bailey and H. Durrant-Whyte, “Simultaneous localization and mapping: Part II,” IEEE Robotics and Automation Magazine, pp. 109–117, JUNE 2006.

Week 8: Navigation (rodents)

Bruce L. McNaughton, Francesco P. Battaglia, Ole Jensen, Edvard I. Moser and May-Britt Moser, “Path integration and the neural basis of the ‘cognitive map’”, Nature Reviews Neuroscience Vol 7, August 2006, pp. 663-678
Michael Milford and Gordon Wyeth, Mapping a Suburb with a Single Camera using a Biologically Inspired SLAM System, under review.