C.1 Source Acquisition

Instructor's Example Solution for C.1 Source Acquisition

I chose to explore legged locomotion since it is a part of the literature that my colleagues and I have contributed to extensively over the years. I felt like I could play the best leadership role in class using this literature - even though our premise is that students are most likley coming to the assignment with no previous experience at all in the area of bioinspired capability they pick to explore.

The overarching question I posed was: “How can animals improve their locomotion patterns to succeed stably over very high speeds and very complex and varied terrain? What would it take to get a robot with such a capability?”

Various Search Results

As we've noted in class, much of the material you will need is in the "Deep Web" so you will need to go through the Penn Portal to gain access - this happens automatically from on-campus connections and you are prompted for your Penn key from home. Just to get a quick start in such situations, I maintain a permanent link on my browser toolbar to the Penn Engineering Database & Article Indexes where the principal electronic search resources are linked.

To get my robotics and bioinspired robotics references, I started with general citation indices: Scopus, ISI, and, of course, the ubiquitous Google Scholar. Since the nature of my search led largely through the engineering literature, and because the assignment requires some evidence from the biological literature, I ended up needing to work with data bases more oriented toward the life sciences as well: PubMed, and Biosis. Although the assignment requires only four backup sources in addition to the main target, I found it advisable to keep track of a larger number of possible sources (at least three) in all of these database searches, just to make sure I would have enough suitable choices to justify my claims in the final presentation of the area of interest.

Scopus Search Results

I decided to start with Scopus and entered the terms “bioinspired robot” and “locomotion” in Article, Title, Abstract, Keywords mode. I then sorted the returned page on "Cited By" with these results.

The most highly cited reference was [1] with 35 citations (on Jan. 17, 2011 when last run), but this reference is now more than five years old, so it will not work for the present assignment: we should make sure to record it (in our bibliographic database [1]) for later consideration, and move ahead with the assignment. The next most highly cited publications are [2] with 31 citations (on Jan. 17, 2011 when last run)
and [3] with 30 citations (on Jan. 17, 2011 when last run). The next most cited [4] is also too old for this assignment, and the citation volume begins to drop off precipitously after this reference, so let's just move on and see what turns up in other databases.

ISI Search Results

I also used ISI Web of Science and again used the topic terms “bioinspired robot” and "locomotion", sorting the response by "times cited". This again yielded the references [1] and [4] at the top of the list of citations, but which I had to record for future use since they are both insufficiently recent for this assignment. Instead, [3] (19 citations on Jan. 17, 2011), and two new references - [5] (17 citations on Jan. 17, 2011) and [6] (8 citations on Jan. 17, 2011) - emerged as the three top cited publications under this combination of topics at ISI.

Google Scholar Search Results

Finally, I went to Google Scholar and used the (long, so as to cut down the number of hits) search phrase
"bioinspired robot locomotion fast stable running gait", and then produced a different outcome using the (GoogleScholar-suggested) variant, "bio inspired robot locomotion fast stable running gait. There were so many hits, and the results were so very mixed by topic and year and citation rate, that I tried to immediately cut back by refining the search to "since 2005". There were still a great number of hits, all mixed in a way that made little sense to me (notwithstanding Google's claim to be ranking "the way researchers do, weighing the full text of each document, where it was published, who it was written by, as well as how often and how recently it has been cited in other scholarly literature").

Going way down into the second page of search results, I found a seemingly relevant general article [8] with a large citation rate (116 on Jan. 17, 2011) in a very highly regarded journal. In more than one case, the first page of results had articles that turned out to be well more than five years old. At the end of my various checks and cross-checks, it appeared that on the first page, the two top cited recent references were [3] (with 38 citations on Jan. 17, 2011) and [9] (with 29 citations on Jan. 17, 2011)).

PubMed Search Results

To look for biology sources, I first went to Pubmed. I used the keywords “bioinspired robot locomotion” yielding: the reference [6], with these data; the reference [11], with these data; the reference [12] with these data; and the reference [13] , with these data.

Biosis Search Results

Finally, I went to Biosis and used the search keys for all fields “robot” AND “locomotion” AND “run$” to get
seven hits (on Jan. 17, 2011). Discarding those more than five years old and those not apparently concerned with unstructured locomotion, left four references:
[6], [14], [15] , and [16] .

An Aside: The Challenge of Assessing Quality of Technical Information

In the class lecture we stressed the importance of checking metrics as the ISI Journal Citation Reports and other public citation indexing services. The careful student will also want to use more qualitative (and harder-to-gauge) measures such as the mission statement and the quality of the editorial board to judge the value of various scientific venues. These will be the mechanisms stressed in our class since they are readily accessible to novices and experts alike, and they are surely better than taking no measures at all to weigh the relative value of your potential sources. But the fact remains, one of the most daunting barriers to moving beyond a novice acquaintance with a technical field is the problem of determing which of the many available sources of expertise carry the greatest value and thus deserve the greatest focus of the learner's attention.

In truth, despite the allure of numerical impacts as measured by outfits such as ISI, no single assessment really gives an adequate picture. For example the, PLoS Medicine Editorial Board has written an incisive critique of the journal impact factor. Taking a different look at this issue, the CRA Best Practices Memo (typically intended for tenure review committees and department chairs and deans rather than researchers :) presents a compelling defense of the generic value of refereed conferences (and other less traditional vechicles for technical exchange) whereas such venues are not catalogued by ISI at all. Similarly, the IMU Citation Statistics Report offers a convincing case for the superiority of qualitative assessment by the peer community over any numerical measures of quality. Of course, the premise of this class is that a learner may likely not have access to reliable participants in the technical peer community, particularly in the early stages of some inquiry. Perhaps this feature of the research university - its offering access to such individuals in dense, concentrated form - will remain the enduring value of such institutions long into the age of acclerating information access now upon us.

Addressing the Evaluation Rubrics

Assuming I know ''nothing'' about this field a priori, I must start with a blind assessment "by the numbers", and then follow with a somewhat more qualitative inquiry according to the way the numbers steer me, as follows.

C.1.0) Source

To address the requirement "Identify one recent paper …. Cite it appropriately, using the References format found in the Wiki.", I chose to use reference [3]. I tentatively decided to focus on this paper since it had gotten a hit in the greatest number (three out of five - the only comparably frequently encountered references in my search was the "older" paper [1]) of the varied search engines I consulted in my search process.

This paper is concerned with vertical climbing, in particular, quoting,

…an integrated, systems-level view of several novel design and control features associated with the biologically inspired, hexapedal, RiSE (Robots in Scansorial Environments) robot.

hence it became clear that my claims- and fact-checking sources would need to be concerned with topics such as locomotion over rough, non-level or generally unstable terrain.

C.1.1) Venue

I now offer evidence that the venue is appropriate for the topic under considerations and of sufficient quality to be worth reading more carefully.

Mission and Scope of Journal

The article under consideration was published in a journal whose stated aims are "to promote scholarly publications dealing with the fundamentals of robotics in unstructured and dynamic environments.'' It remains to make sure that the journal and the authors enjoy reasonable credentials.

Relative Impact Factor

The ISI Journal Citation Report for this paper listed an impact factor of 1.99 which places it fourth on
the list of 16 journals in robotics tracked by this citation index.

Similarly, this journal is ranked 0.091 by the Scopus Journal Analyzer placing it also among the top four of the 16 robotics journals tracked by that index as well.

C.1.2) Authors' Qualifications

Most of the authors of this paper are university professors at reputable institutions. To make sure they are also reasonably well considered within their fields it is simplest (although always simplistic, as we have discussed in the aside) to run some quick citation analyses. The first author registers Scopus-computed h-index of 10; the last two authors (often, the most senior are listed at the end) show Scopus-computed h-indices of 27 and 20 respectively. In other words, the lead author has written more than ten different papers that have been cited at least that number of times by other experts publishing in the scientific publications tracked by Scopus. And other, more senior co-authors have even deeper citation records. Knowing that the Scopus-computed citation rate (focusing as it does on specific archival journal venues) tends to be conservative relative to fields such as robotics that stress conferences, these numbers seem convincing.

C.1.3) Source Identification Methods

I will "pull rank" and not consult our SEAS librarian as all students in this class are required to do, relying instead on my careful account of source identification methods above.

C.1.4) High Quality Bioinspired Robotics Contribution

Having established the quality and relevance of [3], I can now quote directly from the conclusion of that paper to restate the problem area and advance the claim that animals still exceed and remain appropriate targets for improving locomotory efficacy in robotic science:

The speed and versatility of the platform can be improved so that future climbing robots begin to approach the performance seen in animals.

C.1.5) General Robotics Literature

Judging from titles alone, the three references, [13] and [15], [9] appear to be the robotics papers I turned up that best address the question of how well robots fare over rough terrain: specficially, whether there remain substantial research challenges and whether bioinspiration is a likely source of solutions.

Quality of Sources

Reference [9] has a lower impact factor than many of the other
16 robotics venues Scopus can analyze but a closer inspection reveals it is an IEEE conference, hence, as a refereed venue sponsored by a major scientific organization, it should be quite acceptable, particularly since it has impact superior to a number of archival journals.

I can't locate the journal in which [13] appears in the ISI catalogue at all, but Scopus Journal Analyzer shows the citation rate beginning to take off for this apparently very new venue (which only began publishing in 2006 according to the report), and this establishes its viability as a supporting reference. Although the author list for this paper overlaps substantially with that of my target [3], their appearance in independently refereed archival journals justifies in part the attribution of the authors' motives and goals to that of the larger scientific community.

Unfortunately, I cannot locate any information at all about the publisher or nature of the meeting on which the book containing [15] is presumably based. Thus, I will drop it from further consideration.

Checking Claims via Two Additional Quality Sources

A glance at the Scopus published abstract for [9] yields immediate evidence, e.g., this quote,

We conduct extensive experiments to verify that the controller is able to robustly cross a wide variety of challenging terrains, climbing over obstacles nearly as tall as the robot's legs,

that rough terrain presents a difficult challenge in robotics. I grabbed a copy of the article via the Penn Text button on this abstract page (note, it is not easy to negotiate the publisher's site, so copying the DOI is a particularly efficient means to get the article).

Similarly, a quote from the Scopus published abstract of [13],

Terrestrial arthropods negotiate demanding terrain more effectively than any search-and-rescue robot.

bolsters the case for bioinspired research on robotics in tough terrain and convinced me I needed to get my hands on the paper. Because of some problem with the publisher's link I was not able to download it immediately via Penn Text. As it turns out, one of the authors has privately posted the entire paper as revealed by this Google Scholar Search. Moreover, somewhat annoyingly, a subsequent exercise of Penn links succeeded in opening the Journal Home page after all.

C.1.6) Biology Literature

As it turns out, few of the articles discovered in my original search turn out to have appeared in the mainstream biology literature, as the engineering-oriented names of the venues make clear.

Running Through the Reference List

The abstract of [12] states that the article concerns:

… a bioinspired polymer material with the ability of multiple glue-free bonding and debonding …successfully applied … to the 120 g wall-climbing robot Mini-Whegs.

and hence it does seem to fit the prescription in assignment C.1.6) to find evidence of animal capabilities related to locomotion on difficult terrain. One might argue that the announced scope of the journal Bioinspiration and Biomimesis in which appeared the references [13] and [12] has a claim on biologists' attention, but we have already used [13] as evidence from the robotics literature, thus it seems a little suspect to turn around and use [12] as support from the biology literature.

There is only one more clearly relevant reference from the previous search: reference [16], reveals itself to represent a mainstream biological setting from the very name of the journal, and it supports the utility of seeking new, fast locomotory capabilities in animals according to the summary which states:

Rapid, vertically climbing cockroaches produced climbing dynamics similar to geckos, despite differences in attachment mechanism …The general force patterns used by cockroaches and geckos have provided biological inspiration for the design of a climbing robot named RiSE (Robots in Scansorial Environments).

Generating More Evidence from the Biological Literature

To finish the assignment, I will need to find one additional, arguably biological, source that similarly supports the value of looking for new approaches to robotic rough terrain locomotion inspired by animals - since there are no further relevant entries in the original research. In such a situation, it seems to make most sense to review those articles citing reference [3] for evidence that bioinspiration may be of help .

For example, we find [17] an article in the Journal of Compative Physiology whose abstract states:

Discharges of tarsal campaniform sensilla could effectively signal active substrate engagement when the pretarsal claws and arolium are used to grip the substrate in climbing, traversing irregular terrains or walking on inverted surfaces.

Although this article is too recent to expect many citations, its Scopus Journal Analyzer Impact is 0.237 (far higher than the robotics journals we examined) and its ISI JCR score is near 2, which places it in the 45th place out of all 75 physiology journals compared in the ISI Journal Summary List - an entirely reasonable basis for accepting it.

We now have our two biological papers to complement the two robotics references, and the assignment is now complete.

References

1. D. E. Koditschek, R. J. Full, and M. Buehler, "Mechanical Aspects of Legged Control", Arthropod Structure and Development, vol. 33, no. 3, pp. 251-272, 2004.
2. K. Autumn, A. A. Rizzi, U. Saranli, M. Buehler, A. Saunders, M. Cutkosky, W. Provancher, R. Fearing, R. J. Full, D. I. Goldman, R. Groff, and D. E. Koditschek, "Robotics in Scansorial Environments" , Proceedings of SPIE, vol. 5804, pp.291-302, 2005.
3. Spenko, M. J. and Saunders, J. A. and Haynes, G. C. and Cutkosky, M. R. and Rizzi, A. A. and Full, R. J. and Koditschek, D. E., "Biologically inspired climbing with a hexapedal robot", Journal of Field Robotics, vol. 25, no. 4-5, pp. 223-242,2008.
4. P. Arena, L. Fortuna, M. Frasca, and G. Sicurella, "An Adaptive, Self-Organizing Dynamical System for Hierarchical Control of Bio-Inspired Locomotion", IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, vol. 34(4), pp. 1823-1837, 2004.
5. P. Dario, M. C. Carrozza, E. Guglielmelli, C. Laschi, A. Menciassi, S. Micera, F. Vecchi, S. S. Sant'Anna and I. Pisa, "Robotics as a Future and Emerging Technology: Biomimetics, Cybernetics, and Neuro-robotics in European Projects", IEEE Robotics & Automation Magazine, vol. 12(2), pp. 29-45, 2005.
6. Li, Chen and Umbanhowar, Paul B. and Komsuoglu, Haldun and Koditschek, Daniel E. and Goldman, Daniel I., "Sensitive dependence of the motion of a legged robot on granular media", Proceedings of the National Academy of Sciences, Vol. 106, No. 9, pp. 3029-3034, 2009.
7. S. Kim, J. E. Clark and M. R. Cutkosky, "iSprawl: Autonomy, and the Effects of Power Transmission", Proc. 7th Int. Conf. on Climbing and Walking Robots (CLAWAR), pp. 859-867, 2004.
8. Pfeifer, R. and Lungarella, M. and Iida, F., "Self-Organization, Embodiment, and Biologically Inspired Robotics", Science, Vol. 318, No. 5853, pp. 1088-1093,2007.
9. Kolter, J. Z and Rodgers, M. P and Ng, A. Y, "A control architecture for quadruped locomotion over rough terrain", Robotics and Automation, 2008. ICRA 2008. IEEE International Conference on, pp. 811–818, 2008.
10. S. Kim, J. E. Clark and M. R. Cutkosky, "iSprawl: Design and Tuning for High-speed Autonomous Open-loop Running", The International Journal of Robotics Research, vol.25(9), pp. 903-912, 2006.
11. R. W. Blake, H. Ng, K. H. S. Chan and J. Li, "Fish and Chips: Implementation Of A Neural Network Model Into Computer Chips To Maximize Swimming Efficiency In Autonomous Underwater Vehicles", Bioinspiration & Biomimetics, vol. 3, no.3, 2008.
12. S. N. Gorb, M. Sinha, A. Peressadko, K. A. Daltorio and R. D. Quinn, "Insects Did It First: A Micropatterned Adhesive Tape For Robotic Applications", Bioinspiration & Biomimetics, vol. 2(4), pp. 117-125, 2007.
13. J. C. Spagna, D. I. Goldman, P. C. Lin, D. E. Koditschek and R. J Full, "Distributed Mechanical Feedback In Arthropods And Robots Simplifies Control Of Rapid Running On Challenging Terrain", Bioinspiration and Biomimetics, vol. 2(1), pp. 9-18, 2007.
14. C. Maufroy, H. Kimura and K. Takase, "Towards A General Neural Controller For Quadrupedal Locomotion", Neural Networks, vol. 21(4), pp. 667-681, 2008.
15. S. R. Gajjar, "Biomimetic Robots For Robust Operation In Unstructured Environments", Design And Nature III: Comparing Design in Nature With Science And Engineering, WIT Transactions on Ecology and the Environment, pp. 255-263, 2006.
16. D. I. Goldman, T. S. Chen, D. M. Dudek and R. J. Full, "Dynamics Of Rapid Vertical Climbing In Cockroaches Reveals A Template", Journal of Experimental Biology, vol. 209(15), pp. 2990-3000, 2006.
17. Zill, Sasha N. and Keller, Bridget R. and Chaudhry, Sumaiya and Duke, Elizabeth R. and Neff, David and Quinn, Roger and Flannigan, Clay, "Detecting substrate engagement: responses of tarsal campaniform sensilla in cockroaches", Journal of Comparative Physiology A, Vol. 196, No. 6, pp. 407-420, 2010.
18. G.V. Lauder and P.G.A. Madden, "Advances In Comparative Physiology From High-Speed Imaging Of Animal And Fluid Motion", Annual Review of Physiology, vol. 70, pp. 143-163, 2008.
19. M.A. Daley, G. Felix and A.A. Biewener, "Running Stability Is Enhanced By A Proximo-Distal Gradient In Joint Neuromechanical Control", Journal of Experimental Biology, vol.210(3), pp. 383-394, 2007.