C.2 Source Annotations

C.2.1) Behavior or Capability

Control in Serpentine Locomotion

Serpentine locomtion, specifically lateral undulations, are the wave like movements along a slender body that allow for forward movement. Slender, and unimpeded by wheels and limbs, these serpentine robots demonstrate excellent mobility and capability to navigate through spaces. While this snake-like movement has been replicated by many researchers, the challenge remains in control of motion such as to allow for unrestricted, fluid motion through any confined space. The fact remains that these robots are slow and have limited motor control. Because snake robots have so many links and joints this also results in a more complex configuration plane of movement. Each joint cannot be operated individually. The collection of joints must move in synchrony in order to achieve purposeful movement, avoiding Brownian motion. Thus far, robots do not demonstrate autonomous movement, important when navigating unstructured environments, where the user cannot always anticipate obstacles and direct every movement, especially when the snake robot is navigating inaccessible locations not visible to the user.

Uses of Serpentine Locomotion

In search and rescue (SAR) missions, the purpose is to access collapsed areas, where it is necessary to not only navigate over, but through rubble. It also has applications in the medical field, where minimally invasive procedures are favored. Clearing a blood clot or plaque buildup in the bloodstream without performing open heart surgery, for instance. Also, due to a lower weight surface area ratio, compared to robot counterparts with wheels and appendages, these snake robots are better at navigating terrains like sand or water, without getting stuck and bogged down by its own weight.

C.2.2) Capabilities of Existing Technology

While searching for sources for C.1., I focused more so on recent publications to get a sense of what was 'new' and developing in the field. I realized that I did not have much information on what technologies already existed. I did have [1] from C.1., which I did not use because it was too 'old'; I would be looking at it again now. I ventured back into the literature and located another source of information in [7]. [7] is quite reliable - author's have h-indices of 8 and 21, the article is cited 30 times, though it appeared in proceedings from IEEE International Conference on Robotics and Automation in 1999. [1] was cited 88 times as reported on Scopus, with authors having h-indices of 1,2, and 15 (the first two authors were only active for a year or so), and it was published in a credible venue: IEEE Control Systems Magazine.

[1] provides a more general overview of the existing mechanisms for serpentine locomotion. [1] reported on how they created a wheel-less robot that achieved serpentine motion, incorporating control feedback to delegate efficiency of movement in a five-link snake robot. Considering directional friction coefficients and locomotion inertia, the authors were able to receive feedback and control for speed: "our control strategy yielded a reasonable velocity control system for the locomotion of the snake robot." However, [1] also recognizes the challenges: “A mathematical framework is established for modeling, analysis, and synthesis of the serpentine gait, but the tremendous adaptability of the snakelike mechanism to rugged environments has not yet been fully exploited. Studies of other gaits remain to be done…” So, the authors have established an useful framework for controlling locomotion when friction coefficients are known, that is, when the environment parameters are well understood, but authors have not yet extrapolated this capability to unstructured and unforeseen environmental elements. Nonetheless, this paper has shown that control of locomotion is possible; more steps need to be taken to decipher environmental elements.

C.2.3) Potential Biological Solutions or Bioinspired Approaches

[3] acquired from C.1, among other sources, pointed out that “an important problem in the control of locomotion of robots with multiple degrees of freedom (e.g., biomimetic robots) is to adapt the locomotor patterns to the properties of the environment.” So while serpentine motion is well documented, the dynamics that allow for active response to environmental changes have not been. I chose to look at [6] from C.1 source acquisitions. This is a high quality paper, cited 102 times, as reported by Scopus. With the authors having h-indexes of 18, 7, 3, and 8, and its journal venue having an ISI reported impact factor of 29.747, this 2007 published paper certainly presents high impact information.

[6] uses the salamander as a model for gait studies. The research [6] presents on the capabilties of the salamander, an amphibian, in roaming on land and in water. The salamander's ability to naturally transition between the two states indicates that a transition from wheeled robots traditionally used to serpentine robots is possible. Investigation of the salamander's mechanisms for controlling its swimming versus walking states would produce additional insights on how environmental changes are 'logged' by a salamander and how they prompt the salamander to change locomotion mode.

While swimming, the salamander's limbs are folded back and fast axial undulations propagate from its head to its tail. Swimming is a manifestation of serpentine, lateral undulatory, movement. The authors study the different mechanisms that are active during swimming versus during walking. In doing so, they ascertain that the oscillatory centers in the body are activated during swimming, and those in the leg activated while walking on land. This again goes back to the idea of Central Pattern Generators, presented in my C.1 assignment [4].

This paper exposes some of the basic underlying mechanism governing wave-like, lateral undulatory movements, as are seen while a salamander swims. It suggests the use of coupled oscillators to achieve purposeful and directed motion. The authors also suggest future directions in incorporating this capability into robots. As the authors so aptly point out: “There is currently no well-established methodology for controlling the locomotion of robots with multiple degrees of freedom, in particular for non–steady-state locomotion in complex environments.” Overcoming the challenge of dealing with multiple degrees of freedom, and hence a wider range movement configuration, to control for locomotion is significant.

C.2.4) Value of Quality Sources

I found that a few of the sources I initially encountered for C.1 were not applicable as I delved deeper into a more specific domain of the problem. For instance, [4] presented results on a computational model of the polychaete annelid worm's central pattern generator. Upon further reading, I realized that this was simply a model that had not been applied to a robotic system. It was purely a conceptual rendering, and while it was helpful in producing additional papers on CPGs [5], it did not really contribute anything physical. Nonetheless, the paper was a reliable source.

[8] was another source that came up. Its authors having h-indices of 3 and 21, and having been cited 20 times, and published in IEEE/ASME Transactions on Mechatronics , this paper is a quite credible source. The abstract tells us that the authors developed a joint actuator system that is "highly optimized for use in serpentine robots" and allows for "simultaneous proportional control." Thereby, I thought it might have some applications to the problem of interest. Delving further into their results and conclusions, it became apparent that these integrated joints provide better control in motion and integrity in position as stipulated by the user. So, the position of joints as dictated by the user is more accurate, but this does not convey any information on how to improve automated control of serpentine motion without user control.

A search query of snake motion robot* in PubMed yielded [9], "Automated kinematic generator for surgical robotic systems", which really has nothing to do with serpentine motion. The search term was too broad. The search query also yielded [10] which talks about navigation through the gastrointestinal tract through bioinspired robots: "Understanding motion and perception systems of lower animal forms, such as parasites, worms, insects and snakes, can help to design and fabricate bio-inspired robots able to navigate in tortuous, slippery and difficult-to-access cavities of the human body." While the paper has over-arching applications to general control in locomotion, it is only a peripheral 'perspective' paper delving into issues to overcome and not so much producing new results. It is also too focused on human tissue, capitalizing on its adhesive properties in formulating robotic systems, that it is not generalizable to different and variable environments.

As Mr. McGee recommended to me in C.1. , I should "continue to keep track of keywords that can help refocus your searches. For example there may be specific aspects of your project that can serve as useful search terms such as terrain, materials selection, method of control, power source, weight, other components for your system such as actuators, sensors or the like, etc." I was using broader search queries for C.1, like "serpentine", "undulatory", etc., without narrowing my search to specific aspects of interest, like "CPGs" or "control" or "oscillatory".

C.2.5) Open Problems

The most important open problem is how to delegate automated control. Because there are so many links and joints in a snake robot, synchrony and control are crucial to ensure purposeful, directed movements, especially when navigating through confined spaces. The many parameters cannot be constantly updated by the user. The robot must 'learn' how to adapt and control its many links, joints, actuators, and motors in response to external stimuli. Salamanders[6], eels[7], tadpoles[2], and snakes[1] have been investigated as ideal animal models for serpentine motion. The Central Pattern Generator (CPG) has been implicated in many papers [4][5][6] as the ultimate control over movements. While computation models [4] and mathematical frameworks have been formulated to capture the mechanisms of the CPG, the CPG networks have not been completely elucidated though some simple models have been applied successfully to robotic systems. [13] for instance, successfully implemented CPGs, but each link only had one degree of freedom : "We currently have only one degree of freedom per element. This may be a problem in two cases: when the robot has to get over an obstacle (this would require some vertical flexibility) and if the robot falls on one side."


Existing Technology Paper [1]


[11] is the most heavily cited, 661 times, according to Scopus. It is a book, entitled "Legged robots that balance", published in 1986, and probably the foundation for many developments in the field since. The second most heavily cited, 180 times according to Scopus, is [12], another book called "Machines That Walk: The Adaptive Suspension Vehicle", published in 1988.


[13] and [14] are the two most heavily cited successors, with 62 and 17 citations, respectively. [13] reports on an amphibian robot controlled by central pattern generators. [13] states that "locomotion of the robot is controlled by a central pattern generator (a system of coupled oscillators) that produces travelling waves of oscillations as limit cycle behavior." Scopus reported h-indexes of 7,2,5 and 18, respectively. The paper was published in a journal with an ISI reported impact factor of 1.361. [14] reports on Authors have Scopus h-indexes of 3,11,17, and 9 and was published as part of a proceeding for an IEEE sponsored conference. They closely controlled a snake-robots turning motions by: "[To] compensate offset and orientation errors of the robot, we propose an amplitude modulation method and a phase modulation method based on analysis of the serpenoid curve." The tracking control method collects information from sensors.

Potential Biology Applications Paper [6].


[15] is the most heavily cited, at 265 citations, and this article bears heavily on the biological side, presenting information on "networks of the brainstem and spinal cord that co-ordinate locomotion and body orientation in lamprey are described", along with "the cycle-to-cycle pattern generation of these networks." and the "vestibular control of body orientation during swimming." This likely provided the basis for study of salamander swimming patterns in [6]. With a primary author having a h-index of 31, and the journal venue, Trends in Neuroscience, having an impact factor as reported by ISI of 12.794, [15] is certainly a reliable source.

[16] is the second most heavily cited, at 252 citations. Both authors have Scopus h-indexes of 27 and the journal, Science, has an ISI reported impact factor of 29.747. [16] applies mathematical and theoretical concepts to observe dynamic pattern generation in complex systems, having applications to the control of undulatory wave motions of the salamander while swimming.


[5] has been cited the most, 69 times, as reported by Scopus. Its author holds an Scopus h-index of 18, and the journal has an ISI reported impact factor of 1.879. [5] is a review article, talking about locomotion control in animal models through CPGs.

[17] has been cited 68 times. Its authors have h-indexes of 11, 10 and 8, the journal is Science, with an impact factor of 29.747. [17] appears to be a theoretical review paper, exploring the achievements, gaps, and possibilities in the bioinspired robots field. They propose that "Exploiting the dynamics provided by materials and morphological properties as well as the interaction between physical and information processes promises to extend the capabilities of established control-based robot design methodologies", for the development of "adaptive autonomous systems." While this idea is great, this is only a proposal. The authors did not execute any experiment.

C.2.8) Bibliographic Data

1. Saito, M., Fukaya, M., & Iwasaki, T. (2002). "Serpentine locomotion with robotic snakes". IEEE Control Systems Magazine, 22(1), 64-81.
2. Kim, B., Kim, D. -., Jung, J., & Park, J. (2005). "A biomimetic undulatory tadpole robot using ionic polymer-metal composite actuators". Smart Materials and Structures, 14(6), 1579-1585.
3. Crespi, A., & Ijspeert, A. J. (2008). "Online optimization of swimming and crawling in an amphibious snake robot". IEEE Transactions on Robotics, 24(1), 75-87.
4. Sfakiotakis, M., & Tsakiris, D. P. (2007). "Neuromuscular control of reactive behaviors for undulatory robots". Neurocomputing, 70(10-12), 1907-1913.
5. Ijspeert, A. J. (2008). "Central pattern generators for locomotion control in animals and robots: A review". Neural Networks, 21(4), 642-653.
6. Ijspeert, A. J., Crespi, A., Ryczko, D., & Cabelguen, J. (2007). "From swimming to walking with a salamander robot driven by a spinal cord model." Science, 315(5817), 1416-1420.
7. McIsaac, K. A., & Ostrowski, J. P. (1999). "Geometric approach to anguilliform locomotion: Modelling of an underwater eel robot." Proceedings - IEEE International Conference on Robotics and Automation, 4, 2843-2848.
8. Granosik, G., & Borenstein, J. (2005). "Integrated joint actuator for serpentine robots." IEEE/ASME Transactions on Mechatronics, 10(5), 473-481.
9. Jung DL, Dixon WE, Pin FG. (2004). "Automated kinematic generator for surgical robotic systems." Stud Health Technol Inform. 2004;98:144-6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/15544260
10. Menciassi A, Dario P. "Bio-inspired solutions for locomotion in the gastrointestinal tract: background and perspectives." Philos Transact A Math Phys Eng Sci. 2003 Oct 15;361(1811):2287-98. PubMed PMID: 14599320.
11. Raibert, MH. "Legged robots that balance." Massachusetts Institute of Technology Cambridge, MA, USA, 1986. ISBN:0-262-18117-7
12. Song, SM, Waldron, KJ. "Machines That Walk: The Adaptive Suspension Vehicle." MIT Press Cambridge, MA, USA, 1988. ISBN:0262192748
13. Crespi, A., Badertscher, A., Guignard, A., & Ijspeert, A. J. (2005). "AmphiBot I: An amphibious snake-like robot." Robotics and Autonomous Systems, 50(4), 163-175.
14. Ye, C., Ma, S., Li, B., & Wang, Y. (2004). "Turning and side motion of snake-like robot." Paper presented at the proceedings of the 2004 IEEE
International Conference on Robotics 8 Automation , 2004(5) 5075-5080.
15. Grillner, S., Deliagina, T., Ekeberg, Ö, El Manira, A., Hill, R. H., Lansner, A., et al. (1995). "Neural networks that co-ordinate locomotion and body orientation in lamprey." Trends in Neurosciences, 18(6), 270-279.
16. Schoner, G., & Kelso, J. A. S. (1988). "Dynamic pattern generation in behavioral and neural systems." Science, 239(4847), 1513-1520.
17. Pfeifer, R., Lungarella, M., & Iida, F. (2007). "Self-organization, embodiment, and biologically inspired robotics." Science, 318(5853), 1088-1093.

C.2.9) Brief Annotation of Content, Contribution and Relevance

All sources are mentioned in above sections, with brief explanations of content and relevance