Sunday, September 19, 2010

My primary Ph.D projects

It gets pretty confusing about what I do for Ph.D. I am in the Biology Department at Tufts University, getting my Ph.D in biology. However, I've worked with people from biomedical engineering, mechanical engineering, electrical engineering, and computer sciences in the course of my Ph.D training. In fact, most of my colleagues start to think that I'm from "their department" (whatever it is). What exactly am I? That's a very good question which I actually does not have an answer to. Perhaps after I summarize my primary Ph.D research projects below, you may call me whatever you wish with a better understanding of what you are referring to.

My first project at Tufts was to design a force beam array to collect ground reaction forces from all the leg contacts from Manduca caterpillars in two directions simultaneously. This system involves synchronized video tracking and alternative force transducer design. The results suggested that caterpillars may use the substrate to transmit force much like an external skeleton. I'm currently working on the manuscript for the second part of the animal study.

For soft-bodied animals, tissue material properties can contribute greatly to the overall behaviors. My second project at Trimmer Lab was to instrument a lever-arm system to perform uniaxial tests on soft cuticle from the Manduca caterpillar. These soft specimens were often 5mm long and less than 1mm wide. Real-time video extensometry was necessary to control the strain. We then attempted an constitutive model for the material.

To understand how tissue mechanical properties affect the overall animal behaviors, one must go to the structural level. My third project was to investigate the overall structural properties of the Manduca caterpillar over a scaling range. We tracked down all the body tissue that could take mechanical loads and simulate the hydrostatic skeleton in a FEA model. We found a dramatic increase of flexural stiffness associated with body miniaturization. This implies several evolutionary constraint on caterpillar body plans.

Going beyond the Manduca crawling, I wanted to explore some clues about how caterpillars developed inching gaits in the course of evolution. I first simulated different crawling and inching gaits in my first soft-bodied robot lineage. Then I created a second robot lineage to simulate some ballistic behaviors which were supposed to derive from the normal locomotor patterns. Finally, I conducted a field work in Costa Rica to examine all these caterpillar behaviors in nature across different species.

In summary, my practical skills in the lab allowed me to design systems, acquire all the parts/supplies, manufacture/implement devices, and program systems to execute experiments. I would summarize my scientific contribution in my little niche in three statements.

1) During locomotion, soft-bodied animals can use the substrate as their external skeleton, and therefore gain stability and robustness.

2) The mechanical scaling of hydrostatic skeleton may limit the prolegs configurations and locomotor modes in caterpillars.

3) Soft-bodied robots, by definition, cannot rely on exact postural control. The very minute such a robot force a conformation, it becomes rigid.

Wednesday, September 15, 2010

Mechanics of a ballistic roll - Part 2

GoQBot is a soft-bodied robot designed to simulate the ballistic rolling escape behavior in caterpillars. The robot mimics the scale, timing and morphing of the behavior while allowing us to explore the control parameters and perform dynamics tracking. Please refer to earlier posts for more background information.
Kinematics tracking can be very tricky especially with high speed erratic movements. In order to compute the angular momentum of GoQBot, we must know the mass distribution of the robot as it deforms quickly within the 0.2 second window. One way to do it is to break down the body into many segments and track each segment individually. However, that requires installing more than twenty 1-mm size IR emitters on a small soft body no longer than 12cm. The soldering would be simply a torture, not to mention the wiring. Luckily, there is an alternative: kinematics-based model extrapolation.The idea is to constrain a virtual model of the GoQBot using the five IR-marked coordinates as the reference. By applying some deformation characteristics on the model, I can easily extrapolate the positions of every bit of the robot. This is exactly what I have done.

Monday, September 6, 2010

Sharing research tools and links

Hi there,

A new section has been added to my blog called "Huai-Ti's research tools". There you would find some useful links from general referencing to good online stores. There are also some good deals such as a free SEM imaging program (but the image will be public). In addition, you may find some useful software for making presentations or publications. For example, there is a Screen Capture Software that streamlines the print-screen function in most computers. Anyways, stuffs like this could be useful when you need it. I will populate this list gradually.

Over the past four years of my R&D, I have designed over 12 instruments, written more than 5 systems of feedback control programs, placed more than 85 orders for supplies and parts. I guess the number of projects and pet projects in experimental science directly correlates to the spending and instrumentation. Also, people in my research team tend to leave purchasing jobs to me since I've been really good at communicating with vendors and company associates to get the right parts for laboratory research. But I have to say such job is really stressful, although all the spending came from various research grants and not from my pocket.

Friday, September 3, 2010

Mechanics of a ballistic roll - Part 1

First off... I would like to remind you of the behavior that I'm studying. See the ballistic rolls and flips in some leaf roller caterpillars, and check out my biomimetic GoQBot's performance.

To study the the dynamics of such impressive behavior, the robot has been a great tool to test hypotheses and facilitated the collection of mechanical and control data. Back in April this year, I successfully collected the kinematics of GoQBot using my own adaptation of the VICON motion capture system. However, the kinematics data alone only tells half of the story. We could compute the angular momentum as well as the linear momentum from pure motion tracking. However, there is no data on the mechanical power output and loading condition. To extend the analysis, I took one of the force beams I used for caterpillar ground reaction forces measurement and implemented it for robot GRF collection. In fact, VICON system has built-in capabilities to take in force-plate measurements and any other analog signals.With some modification and instrumentation, I obtained GRF in two directions from the head anchor and simultaneous recording of the actuator current draw during the high speed kinematics recording. Here are some very preliminary data.I'm currently processing the data to calculate the center of mass and the development of angular momentum. More results are coming!!