Adam Komoroski: C.2 Source Annotation

C.2 Source Annotation

C.2.1) Behavior or Capability

Snake-like Locomotion

Designing robots to imitate the way snakes move around brings about a number of advantages over wheeled or legged methods of locomotion. In general, snake-like robots are multi-link articulated robots that often resemble snakes or other animals, such as salamanders, that have similar characteristics. [3] explains that the redundancy in the design itself offers the benefit of robustness to mechanical failure. More importantly, by imitating the motion strategies of snakes and other snake-like animals researchers hope to achieve effective locomotion that exceeds the capabilities of wheeled and legged robots in certain situations. Snakes themselves are very good at moving quickly over rough and varied terrain, and are extremely capable when it comes to adapting to different environments. The idea is that by imitating the design and motion of these animals, similar advantages will be seen in the robots as well. A number of sources from both robotics and biology such as [6], [5], and [3] discuss three common gaits that are used by snakes, which are lateral undulation, concertina locomotion, and sidewinding. By observing and modeling locomotion strategies such as these, researchers are attempting to develop control strategies for snake-like robots that will mimic the motion of real snakes and will allow them to move effectively over rough and varied terrain.

Uses

Since these robots are particularly good at moving in and adapting to new and irregular environments, they are very well suited to tasks that make it necessary to traverse such terrain. They could be used effectively to explore unknown environments and map these areas or find objects or people in them. They could transport and deliver items such as food and supplies or they find and retrieve items in unknown environments with terrain that would be difficult for wheeled or legged robots to get through. [1] suggests the following about the applications of snake-like robots:

Robots with these properties open up several critical applications in exploration, reconnaissance, medicine, and inspection.

C.2.2) Capabilities of Existing Technology

Recent advances in our ability to develop control strategies that allow snake-like robots to take advantages of a snake's ability to adapt to varied environments are detailed in [1]. Their approach seeks to reach this goal by attempting to achieve:

…automatic design through genetic programming (GP) of the fastest possible (sidewinding) locomotion of simulated limbless, wheelless snake-like robot (Snakebot).

This research suggests that this approach is an important step toward achieving fast, effective, and adaptable locomotion using snake-like robots. The authors describe the effectiveness of this solution, stating:

Robustness of the sidewinding Snakebot, initially evolved in unconstrained environment (considered as ability to retain its velocity when situated in unanticipated environment) was illustrated by the ease with which Snakebot overcomes various types of obstacles such as piles of and burial under boxes, rugged terrain, and walls. The ability of Snakebot to adapt to partial damage by gradually improving its velocity characteristics was discussed.

Thus, this method of achieving useful locomotion from snake robots is quite promising. The major goal of generating locomotion that is robust and adaptable in order to be effective in challenging environments was the primary focus of this paper. The results suggest that this approach could be a reliable way to achieve this goal. The robot shows its ability to change its motion in order to move between two narrow walls or to move out from under a pile of boxes. However, it is only a step in the right direction and it not a complete solution. Improvements could be made to allow the robot to move more efficiently and to adapt more quickly. Additionally, the experiments detailed in the paper were performed only in simulation. The authors note the realism of the simulations:

The realism of simulation is ensured by employing the Open Dynamics Engine (ODE), which facilitates implementation of all physical forces, resulting from the actuators, joints constrains, frictions, gravity, and collisions.

However, this is not a complete substitute for performing experiments using real robots, and new challenges will likely need to be addressed when the time comes to move from simulation to a real robot.

C.2.3) Potential Biological Solutions or Bioinspired Approaches

Recent efforts to solve these problems from the biology side are explained in [5]. The biological approaches to creating effective snake-like motion focus on observing and modeling the structure and motion of actual snakes, slugs, earthworms and similar creatures. This particular paper looks at a number of aspects of snake locomotion in order to appropriately approach the design of snake-like robots. These areas of focus include aspects such as frictional characteristics, weight distribution, and gait. This paper also mentions the three common gait types (lateral undulation, concertina locomotion, and sidewinding) and chooses to focus on lateral undulation. Focusing on these aspects, [5] then attempts to develop a model to describe the motion.

The authors conclude by explaining that the models they developed captured general trends in the motion of actual snakes but with some error. The offer some potential explanations for these errors, stating:

we have modeled the tail as having the same cross-section and mass per unit length as the body

and

Another aspect that we have not considered is that snakes are likely able to dynamically change their frictional interactions with a surface by adjusting the attitude of their scales, a possibility that bears on our assumption of independent friction coefficients in Eq. 2. As suggested by the theoretical predictions of motion on a ramp in Fig. 2D, which fit the experimental data better with forward friction coefficients of 0.3 rather than 0.1, snakes may be able to alter their frictional properties when moving up and down a slope.

Thus, this approach has been effective in generally describing the way that snakes move, but more work needs to be done before it is reasonably accurate at predicting the motion of actual snakes.

C.2.4) Value of Quality Sources

The search terms I chose such as 'snake robot' and 'snake locomotion' generally provided papers that were relevant to my topic. The papers that I rejected from my original search were mostly due to issues of recency, number of citations, quality of the venue, and quality of the authors. I ultimately chose my papers based on what I found to be the most interesting and most closely related to my topic, even though almost all of the papers I clicked on were acceptably close to my topic of choice.

There were a few exceptions to this, however. There are a number of snake-like robots that are used for surgery such as Highly articulated robotic probe for minimally invasive surgery. These showed up in many of my searches, but have little to do with the type of locomotion that my topic is focused on. Many of the articles I looked at seemed to be either too specific or too general for the current task. Papers such as AmphiBot I: An amphibious snake-like robot and GMD-SNAKE2: a snake-like robot driven by wheels and a method for motion control were too narrowly focused for C1.

C.2.5) Open Problems

The most important open problems in this area relate to quickly and effectively adapting to new situations and new environments. Since one of the most impressive aspects of the locomotion of snakes and similar animals is their adaptability, this remains a major goal in the creation of snake-like robots. Papers such as [2] look to use objects in the environment to aid the robot's propulsion. Other papers such as [1] provide more general instructions to the robot that allow it learn about how to move through its surroundings. These approaches have met with some success, but comparison of the movement of these robots to that of actual snakes suggests that there is tremendous room for improvement. With the obstacle aided approach, the motion of the snake in an unstructured environment was clearly suboptimal. While the proposed solutions are promising, the problem of adapting a snake robot's motion to new environments is still very much an open problem.

Annotations

Robotics Paper

As mentioned above, the purpose of [1] is to use genetic programming to automatically design the controller for the robot. They develop a method for achieving this and then test scenarios such as a snake under a pile of boxes and a snake between narrow walls in simulation.

Precursors

Two important precursors of [1] are [4] and [7]. The authors talk about [4] and several other works in their discussion of handcrafted locomotion control. They compare their methods against the results from these precursors. [4] has been cited 110 times according to Scopus. The authors reference [7] in their discussion of accounting for damage of the robot when developing the controller. This precursor has been cited 6 times according to Scopus but it is from IEEE International Conference on Robotics and Automation which I have previously shown to be a reliable source.

Successors

This paper is listed as having 19 successors according to Scopus. An important successor of [1] is [8]. This successor has only been cited 5 times itself, but it is also from IEEE International Conference on Robotics and Automation which has been shown to be reliable.

Biology Paper

The purpose of [5] is to model the motion of snakes, particularly looking at aspects such a frictional characteristics, weight distribution, and gait type. The model is then compared to the behavior of actual snakes.

Precursors

Two important precursors of [5] are [10] and [11]. The authors mention [10] in their comparison of the efficiency of snake-like locomotion to that of legged locomotion. This precursor has been cited 60 times according to Google Scholar. [11] was mentioned by the authors in their discussion of other theoretical analyses. This precursor has been cited 28 times according to Scopus.

Successors

This paper is listed as having 6 successors according to Google Scholar. An important successor of [5] is [9]. This successor has not been cited, but it is a very recent paper. The h-indices of the authors are 2, 12, 4, and 7 respectively, which suggests that they are fairly reliable.

1. Tanev, I., Ray, T., & Buller, A., "Automated evolutionary design, robustness, and adaptation of sidewinding locomotion of a simulated snake-like robot". IEEE Transactions on Robotics, 21(4), 632-645. 2005
2. Liljeback, P.; Pettersen, K.Y.; Stavdahl, O.; Gravdahl, J.T. , "Hybrid Modelling and Control of Obstacle-Aided Snake Robot Locomotion", Robotics, IEEE Transactions on , vol.26, no.5, pp.781-799, Oct. 2010
3. Saito, M., Fukaya, M., Iwasaki, T. , "Serpentine locomotion with robotic snakes", IEEE Control Systems Magazine, 22(1), 64-81, 2002
4. Chirikjian, G.S.; Burdick, J.W., "The kinematics of hyper-redundant robot locomotion", Robotics and Automation, IEEE Transactions on , vol.11, no.6, pp.781-793, Dec 1995
5. Hu, David L. ;Nirody, Jasmine; Scott, Terri ;Shelley, Michael J., "The mechanics of slithering locomotion", Proceedings of the National Academy of Sciences of the United States of America. 106(25). JUN 23 2009.
6. Bruce C. Jayne, "Kinematics of Terrestrial Snake Locomotion", Copeia. Vol. 1986, No. 4 (Dec. 23, 1986), pp. 915-927
7. Bongard, J. C., & Lipson, H. "Automated damage diagnosis and recovery for remote robotics.". Paper presented at the , 2004(4) 3545-3550. 2004.
8. Liljebäck, P., Pettersen, K. Y., & Stavdahl, Ø. "Modelling and control of obstacle-aided snake robot locomotion based on jam resolution." Paper presented at the 2009 IEEE International Conference on Robotics and Automation 3807-3814. 2009.
9. Liljeback, P.; Pettersen, K.Y.; Stavdahl, O.; Gravdahl, J.T.; "A simplified model of planar snake robot locomotion" Intelligent Robots and Systems (IROS), 2010 IEEE/RSJ International Conference on , vol., no., pp.2868-2875, 18-22 Oct. 2010
10. Michael Walton, Bruce C. Jayne, and Albert F. Bennet. "The Energetic Cost of Limbless Locomotion". Science 3 August 1990: Vol. 249 no. 4968 pp. 524-527
11. Burdick, J. W., Radford, J., & Chirikjian, G. S. "'Sidewinding' locomotion gait for hyper-redundant robots.". Advanced Robotics, 9(3), 195-216. 1995