Soft robots preview videos come online!!

Happy New Year!!
Finally, my robot videos are here, on the last day of the year 2009!

After searching around for a host server, I decided that the good old YouTube was still the best so far. My last post was some preliminary tests on video linking. Now let's check out some preview videos of my soft-bodied caterpillar robots!!

First off, the InchBot I-III are the early versions of the my soft robot implementation, dated back to March 2009. It's a process by which I developed inching gaits and learned about frictional control. My colleague Chris successfully modeled these Early InchBots in a finite element environment as well. He really spiced up the video, too.

Then, here come the InchBot IV-VII which twitch, inch, burrow, and climb with much smaller body size. Chris also implemented the InchBot-V in FEA. These soft robots featured open-loop robust inching/crawling/climbing gaits.

Finally, the newest class of caterpillar soft robots, GoQBot, have an escape ballistic rolling behavioral inspired by the caterpillar of Pleuroptya ruralis (mother-of-pearl moth). This class of robots can initiate a rolling behavior within 300ms and hit top speed over 15cm/s. In addition, the updated versions have include almost all the previous InchBot series capabilities and are radio controlled fully untethered. Simple intelligence is implemented into the body structures and active sensing will be next. To hear more, stay tuned to my publications coming soon.

Interesting videos of caterpillers!!

Everybody in my lab knows me for training caterpillars to perform various sportive activities. Well, inducing caterpillars to crawl underwater was a true story. My motivation was to test the role of gravity and external pressure on a behaving animal. As it turned out, 5th instar caterpillars float in water and thus experienced a force negative to gravity, but none of the kinematics characteristics changed. See my video "Manduca underwater walk" on YouTube.

Looking closely, I found that bubbles could be seemed to come out of the spiracles as the animal compressed itself. This observation illustrated the potential change of body volume due to tracheal compression (see video). If caterpillars can squeeze air out under the influence of water pressure, they must perform quite a lot of gas exchange in the air. In other words, locomotion facilitates gas exchange by compressing and squeezing the air out of the trachea.

Finally, I would like to share a video I shot the other day when one big caterpillar was crawling on top of a smaller one. It's quite an pathetic scene because the smaller caterpillar was actually in the molting process and could not fight back. Nevertheless, as the big caterpillar crawled along, I observed appropriate deformation on the substrate (in this case another caterpillar) as illustrated by my new ground reaction forces paper (to appear in Journal of Experimental Biology).

The missing post released!!

Dear readers,

There has been one missing post about the DARPA meeting which I started back in October but never finished. The reason was quite simple: I wanted to wait for better graphics. In any case, since I was orchestrating the live robotic demos for Tufts, there was no way I could take photographs as I always do. It turned out that it was not allowed anyways. DARPA actually hired prefessional film crew and photographers for the event.

Two months has gone by since the event so it's not news anymore. Nevertheless I thought some of you might be interested in reading my story a few days before the meeting! Now, according to the non-disclosure document I signed, I am not supposed to share anything I saw at the meeting. So in this post, I simply described what happened the very last week before the meeting in the labs. The ME professor that I worked with on this project said: "(It was) the most stressful/intense academic experience that I've gone through". Indeed, for the last month we worked at least 15 hours everyday. To hear more, see October 11 post which is newly released.

Hydrostatic skeleton model for caterpillars

As my regular readers, you might have noticed my dual role as a biologist and a roboticist. Indeed, I currently have 6 projects running in parallel, three in locomotion and other three in robotics. Of course, there are also many more side projects. In any case, the posting about caterpillar prolegs configuration was the motivation of a locomotion project. Here let me present to you another locomotion project of mine with some literature review. This one concerns modeling hydrostatic skeleton in caterpillars. The following photos show how an anesthetized caterpillar can lose turgor and fail to "stay in shape".

Modeling mechanics of biological soft structures has been a long time endeavor for functional morphologists as well as theoretical biologists. In general, this field focuses on morphologies without any rigid skeleton (internal or external). They are usually soft tissues supported by some fluid which allows very large deformation. The notion of “hydrostatic skeleton” became well-known by the 50’s largely due to research on worms (cnidarians, annelids, and nematodes). Clark and Cowey established how soft-bodied animals achieved extreme extension with helical reinforcing fibers in the body wall (Clark and Cowey, 1958). The oblique fibers winding around the body allows very large longitudinal stretching. Soon this type of fiber reinforcement was found in many other cylindrical biological structures including those of plants. In 1980’s, a new wave of theoretical investigation of soft-bodied animal locomotion began. Keller and Falkovitz attempted a model of worm crawling using finite difference method which calculated the transverse traveling wave along the body and its associated contact friction (Keller and Falkovitz, 1983). A few years later, Dobrolyubov generalized this line of reasoning to traveling deformation (both transverse and longitudinal). He claimed that the transverse traveling wave can represent caterpillar locomotion while the longitudinal traveling wave resembles crawling worms. Then he gave an example on how this model could describe snake’s locomotion (Dobrolyubov, 1986). These models proposed credible mechanisms for locomotion, but did not explain how animals achieved those body deformations. In another Journal of Theoretical Biology paper, Wadepuhl presented probably the first comprehensive finite element hydrostatic skeleton model based on medical leech which had been well studied by then (Muller et al., 1981; Sawyer, 1986; Stern-Tomlinson et al., 1986; Wadepuhl and Beyn, 1989). This model included geometry, elastic properties of the body wall, internal volume, and body pressure. It revealed some principles of antagonism in worm-like structures as well as the pressure-volume interactions (265 Wadepuhl, M. 1989). At about the same time, Wainwright nicely summarized the mechanics of cylindrical biological structures in his famous little book “Axis and Circumference” (Wainwright, 1988).
Before the turn of the 21st century, Journal of Theoretical biology continued to host models of hydrostatic skeleton. However, experimental data gradually dominated the modeling efforts. Skierczynski et al constructed an updated leech model empirically based on dimensions of animals in limiting cases, passive properties of the tissues, muscle responses to activation, and the transform from motor-neurons to muscles. It assumes elliptical shapes for cross-sections, constant volume, and that the shape tends to minimize the potential energy. It simulates the vermiform elongation and predicts the pressure changes (Skierczynski et al., 1996). Similarly Alscher and Beyn simulated the motion of leech using Lagrangian mechanics and a large system of differential-algebraic equations (Alscher and Beyn, 1998).
While leech models seemed to develop with fast pace, earthworm studies were thriving as well. Dobrolyubov refined his mass transfer wave model and published another paper in JTB with Douchy on peristaltic transport. This general model attempted to explain the digestive transport as well as locomotion by caterpillars, earthworms, snake and snails (Dobrolyubov and Douchy, 2002). Accoto et al added to JTB another earthworm kinematics model, again based on constant volume and simple friction (Accoto et al., 2004). With these numerous hydrostatic skeleton models, it was thought that soft-bodied animal locomotion is more or less realized and what we learned from worms can be applied to others such as caterpillars. Unfortunately, caterpillars are simply not worms in all biomechanical respects.
Caterpillar’s body differs from that of a worm in several essential features: 1) Extension in the longitudinal direction is accounted by numerous inter-segmental folds instead of body wall stretching. 2) Body pressure is highly variable and less predictable. 3) It contains more compressible volume in the body. 4) There is no segmental septum that compartmentalizes the animals. 5) Caterpillars are legged systems with discrete and on-off attachments. As the results, the helical fiber-reinforced cylinder model does not apply. The constant volume assumption does not hold, and real-time pressure recording lacks correlation to body movements. Frictional model based on mass transfer is useless in this system. What’s more, caterpillars don’t move with one single gait and/or body configurations. In this study, we seek an alternative approach to model this worm-like structure that is so much unlike worms.

References

Accoto, D., Castrataro, P. and Dario, P. (2004). Biomechanical Analysis of Oligochaeta Crawling. J. Theor. Biol. 230, 49-55.
Alscher, C. and Beyn, W. J. (1998). Simulating the Motion of the Leech: A Biomechanical Application of DAEs. Numerical Algorithms 19, 1-12.
Clark, R. B. and Cowey, J. B. (1958). Factors Controlling the Change of Shape of Certain Nemertean and Turbellarian Worms. J. Exp. Biol. 35, 731.
Dobrolyubov, A. I. (1986). The Mechanism of Locomotion of some Terrestrial Animals by Travelling Waves of Deformation. J. Theor. Biol. 119, 457-466.
Dobrolyubov, A. I. and Douchy, G. (2002). Peristaltic Transport as the Travelling Deformation Waves. J. Theor. Biol. 219, 55-61.
Keller, J. B. and Falkovitz, M. S. (1983). Crawling of Worms. J. Theor. Biol. 104, 417-442.
Muller, K. J., Nicholls, J. G. and Stent, G. S. (1981). Neurobiology of the Leech: Cold Spring Harbor Laboratory Pr.
Sawyer, R. T. (1986). Leech Biology and Behaviour: Clarendon Press Oxford.
Skierczynski, B. A., Wilson, R. J. A., Kristan Jr, W. B. and Skalak, R. (1996). A Model of the Hydrostatic Skeleton of the Leech. J. Theor. Biol. 181, 329-342.
Stern-Tomlinson, W., Nusbaum, M. P., Perez, L. E. and Kristan, W. B. (1986). A Kinematic Study of Crawling Behavior in the Leech, Hirudo Medicinalis. Journal of Comparative Physiology A: Neuroethology, Sensory, Neural, and Behavioral Physiology 158, 593-603.
Wadepuhl, M. and Beyn, W. J. (1989). Computer Simulation of the Hydrostatic Skeleton. the Physical Equivalent, Mathematics and Application to Worm-Like Forms. J. Theor. Biol. 136, 379-402.
Wainwright, S. A. (1988). Axis and Circumference: The Cylindrical Shape of Plants and Animals. Cambridge: Harvard University Press.

Research odds and ends...

Sorry for the delayed posting again! I've been occupied with the end of the year academic madness: grants writing. One can tell how bad the economics is by looking at how people panic about funding. That's very much the case in academia. In any case, I didn't find these experiences extremely appealing to talk about so I blanked out last Sunday when I was supposed to update my blog.

Regarding the robot videos many people inquired about, I've done my best to push the school administration for video hosting. Unfortunately Tufts is still an university and all administrations eat time. Actually, that's the main reason why I started this blog. Our labs websites update never caught up with our research pace, because we researchers cannot access the webpages ourselves.

By the way, I do have a facebook account, but I don't use it very often. My colleague asked me about it a couple of weeks ago here is how I responded:

Caterpillar prolegs diversity

While we are waiting for the caterpillar robot video links, let's come back to some morphological discussion of lepidoptera larvae. All my biomechanics studies so far are based on a well-known model system tobacco hornworm (Manduca sexta). It is a fair size macro-lepidopera species commonly found in the America. It has 4 pairs of abdominal prolegs plus 1 pair of anal prolegs (or terminal prolegs). This is thought to be the ancestral form.
However, we are perfectly aware of the diversity of lepidoptera species. Caterpillars really vary in numbers and arrangement of prolegs. They also adopt different gait patterns accordingly. So how do we generalize what we learned from Manduca? Or can we?

To find out why, I am planning a field study to compare body overall scaling across different species of caterpillar. Although much work has been done on comparing morphological changes in the evolutionary context of species interaction, little is known about the physical constraints during evolution. I believe that there is a link between locomotor biomechanics and the evolution of prolegs configuration.

From InchBot to GoQBot - video links coming soon!

Dear readers,

If you see this message, then I must thank you for your loyalty to my blog. I apologized for missing out almost a month of blogging, but really I can't afford to get into trouble. In any case, welcome back to my rapid pace of locomotion research and biomimetics. Currently I have two robotic lineages: InchBot[7 generations] and GoQBot[4 generations]

Each robot generation has a very specific research aim and target performance. I think it's time for me to present them to you in a video format. While all the videos have been edited and annotated, I still need to link them from the Tufts Media Server. Please be patient! They are coming soon.

DARPA Review

It's peculiar how things evolve. I never meant to get into this DARPA project until I realized how far I've been sucked in. So here I am working on my robots in the second room of Biomimetic Devices Laboratory every minutes of my awareness.

System engineering requires a lot of collaboration but creation really has to be solo. A roboticist has to be able to work with people as well as alone. I was fully aware of the importance of this final review, because a lot of people depend on this funding at Tufts, including many friends and colleagues. For me, that's enough at stack even if I had alternative funding source for my primary research. I simply had to do everything in my power to make this robotic demonstration rock solid and fail-proof. I was really up for the job!

In September, I produced a plan for preparing this robot demonstration. Following is a rough list of the issues we had to consider:

1. Transportation protection
2. Demonstration platforms
3. Displays cases and proper Logos
4. Portable power supply and charging units
5. The robots (obviously) and their doubles
6. The control system and a secondary backup
7. Primary robot operator, backup operator
8. Demonstration program and rehearsal
9. Robot specification data
10. Backup video shots in case some live demos fail
11. Tool boxes

This list just went on... My job was to coordinate the designs, search for the parts, arrange the purchases, find appropriate help, and do as much as possible to complete my robots and their accessories. In a typical day, I would spend 5+ hrs doing micro-soldering and bonding, 2+ hrs programming, 2+ hrs literature research, 3+ hrs robot testing, 2+ hrs casting/molding. On top of that, I had to purchased about $500 worth of components everyday on average. Actually, I found shopping the most stressful task among all. In order to get the right thing on time, I had to check the mechanical/electric compatibility, availability, pricing, shipping, and make payments. That's probably why R&D companies always have a person or a group of people specifically do purchasing. It's such a tough job.

Anyways, it's Sunday night now! We are up for the show in two days. I better get some sleep and hope for the best.

Intelligence embeded in morphology?

Recent years, the bio-inspired robotic community became very interested in the idea of embedding control strategies into the structures or morphology of the robots. In a fast ballistic locomotion, many animals have some structural designs which allow them to tolerate certain degree of perturbation. This control method has been named "pre-flex" as oppose to neural "re-flex". The phenomenon is well represented by the example of cockroach recovering from lateral perturbation. Sprawl biomimetic hexapod robot nicely demonstrated this damping effect in polypedal running locomotion. In a way, the morphology effectively reduces the computation required to control balance and recovery. Thus some people like to call this effect: "morphological computation".

The power of "morphological computation" relies on the structure's ability to response "differently" to varying conditions. A linear transform is hardly any good because its result only depends on one order of the input. For example, an Hookean spring produces certain tension at certain stretch regardless of how fast it is stretched. Therefore most interesting logic components are non-linear. As the results, the input can be processed conditionally, taking more things into account.

Now, what exactly can be computed in morphology and what parameters are critical? That's a fundamental question for the field of functional morphology and biomechanics. I think I might have found my version of the answer in my robots. Let me get back to this topic in a few weeks after consulting with some caterpillars.

Unleash the robots!!

A complete robotic agent has to be independent. Going untethered is a very important step in any robotic development. This week, my robot controller pulled free!!

Taking the 3D wiring approach, I successfully customized a micro-RC system for my robots. Traditional electronics are limited by the 2D wiring of PCB boards. Making thin electronics is easy as long as the integrated circuits utilize the surface on the PCB efficiently. That's why cellphones and iPods can be so thin. However, if bulk size is an issue (my robots need to fit through a small hole), traditional circuit layout cannot accommodate that down to a critical size. The only solution to improve packing is to do away with any sort of PCB and stack integrated circuits one on top of each other. By fitting these electronics components carefully according to their geometry, one can minimize the overall dimensions. This is exactly what I did.

Well, piecing 3D puzzles is only the first step. Since overall dimension has to be as small as possible, I worked with the smallest of everything from diodes to wires. I started with the World's smallest RC receiver and built on top of that. Typical soldering dot size is less than 500 micron and most soldering joint spacing is 800 micron. Due to the complex 3D structure, wiring and insulation became very complicated. All the wiring and soldering has to be done manually under a microscope. My micro-dissection skills really came to the rescue!! This is perhaps a good example of how human beats machines. [the scale in the following images is mm]

The result was quite satisfactory. I could radio control 7 actuators on my robot with all the electronics packed into a ~6mm cube volume (not including external wires) over a 300 meters range. The customized transmitter can receive commands from a computer interface via a USB cable. This allows the robot operator to gain various computer support including an library of CPG gaits. Image below showed a test setup where actuators were distributed across several different worm-like bodies.

Huai-Ti is up for the DARPA challenge -- 21 days countdown!

With all these robotics attempts, Huai-Ti has taken up the DARPA challenge together with several fellow Tufts roboticists. According to the ChemBot Phase-I challenge, a soft robot has to be produced for covert access.

Primary challenges:
1. Cover 5m in 20min (average speed at 25cm/min)
2. Reduce the largest dimension via morphing (10 fold)
3. Traverse an arbitrary 1cm opening
4. Reform original functions and capabilities

This effort will be evaluated in a live robot demonstration in exactly three weeks or 21 days. Other ChemBot teams include Harvard, MIT, and U. Chicago. We must not look bad in front of them. Start the count down, and wish me good luck. Go Jumbo!!

Intelligent Design versus Evolution in Robotics

Ok, between posts of robotics, I'm going to risk my neck and touch on the famous American debate on Intelligent Design vs Evolution. Since I'm not only a physicist but also a biologist, my position is hard to be neural. Thus I decide to speak in the context of robotics only.

Robotics is a multi-disciplinary engineering by itself. A roboticist not only need to know mechanical design and to be familiar with materials, but also need to understand control strategies and to learn electronic interfacing. It is an integrated system design process. All robotic systems must be designed through our superior intelligence then. Is there any question?

More recently, the Genetic Algorithm (GA) approach to engineering optimization became very popular. The idea is that by introducing variations similar to mutation in a model, we could come up with very unintuitive solutions quickly. This approach can be very powerful for complex systems, which we don't fully understand. Some people start put much hope on this especially in tough engineering problems.

Being a philosophical person, I ask: is there a fundamental difference between human intelligent design and an optimization process such as GA. The working behind GA is try-and-error via simulation. Now let's analyze the process of engineering design to find out how it could be different.

First of all, I would like to quote a saying in the community of mechanical engineering: "you design by intuition, model for conscience, or don't model at all." In many well-designed human artifacts, there was no modeling involved during the designing process. Often times, the designer "simply knew" what would be a better configuration without knowing exactly why. And this knowledge came from experiences. These experiences include personal observations, hands-on works, conceptual reflections, and communication with others. Our engineering knowledge is not legislative. We often go with rules of thumb.

Secondly, the key to good engineering is to build a good sense of physics. In fact, personal observations and hands-on experiences provide physical episodes of how things work. Then conceptual reflection and communication setup virtual simulations of physical systems in our brain. Every good engineer has a physics engine inside hisz/her brain working out solutions by simulating ideas. And the predictive power of this simulator is directly proportional to experiences and our understanding of physics.

Finally, we call our ability to create and design “intelligence”! This ability is how we can determine what could work and fail without actually building the entity. In the ultimate sense, it is our ability to simulate different situations in our brain quickly and come to a good but non-perfect decision. We can avoid physical try-and-error because we did that quickly in our mental simulator. We can often skip many unnecessary simulations because we remember the results from similar episodes before.

Now, let me come back to my question: is there a fundamental difference between human intelligent design and an optimization process such as GA? The difference can only be in the processing units. While we simulate with biological neural networks, GA relies on computers and programming. For a robot, designing a locomotor gait is the same as evolving one. The only difference is brain vs. computer algorithm. So if you can build a computer that interprets our physical understanding well for the system of interest, then robot evolution by GA saves your brain labor. Otherwise, human design and try-and-error is more efficient. We teach computers how to “think” remember?

This is how I solve the question of Intelligent Design vs. Evolution in robotics by arguing their equivalency, or how I create a debate with my radical views.

A Brief Review of Huai-Ti's Caterpillar Robot

It's time for a review of all my robotic attempts with soft materials. I think I am starting to grasp the core principle of making a piece of rubber move. Of course, it's the "rubber soul"!


Coming up... it's GoQBot-II(Roll-n-Jump) and GoQBot-III(Shake-n-Glide). All future robots will be tether-less because I just finished the micro-RC system with dual control [PC+Mannual]. More will be coming up in the near future. Stay tuned!!

Special post on rolling caterpillar robot!!

If any of you remember in April when I featured the fastest locomotion by caterpillar (15" per second), then you would recall this impressive ballistic roll performed by "mother-of-pearl moth" larvae. BBC had a nice little video introduction about it. I posted the YouTube link on my blog under "Interesting Links". In any case, here is that link if you want to refresh your impression on what caterpillars can do.

Well, this August when I came back from the UK, I was compelled to consider replicating this biological feat with my robot. It was the week right after I developed a gait for my caterpillar robot to crawl through a 1cm hole all by itself. The Tufts robotic team was concerned with meeting the DARPA speed metric. Crawling gaits are robust, but rather slow in general. To defended my love of biological inspirations, I introduced this more dramatic locomotor mode. A few days later I had lunch with two of my best colleagues again at WholeFood, it came across our minds that it might not be too crazy to replicate what the rolling caterpillar has done. So the idea of "GoQBot" came into being, but it took us another month of hard work to translate what's on a napkin blueprints to a physical entity (yes, the original ideas were outlined on two WholeFood napkins).

Why is it called "GoQBot"? That's another interesting story. My buddy Tim is really good at drawing comics around engineering schematics. He was also one of the two colleagues in that brainstorming lunch break, the other being Dr. Gary Leisk. Tim envisioned this robot being so quick that it will go on and on as it rolls. So he drew a robot rider who shouted "GoGoGo". Gary later interpreted that into Go to the cubic power like "G^3". I thought the name could be a little more sexy and turned it into "Go-Q". The letter "Q" pictorially resembles the configuration when the caterpillar rolls into shape, and phonetically retained the cubic power of Go. Interestingly, the first generation of GoQ-Bot literally rolled three rotations (Go^3) in the Q configuration. See the brief introduction below!

"GoQBot-I" retains the same body plan from the InchBot series (InchBot IV~VII) which could performed three kinds of inching gaits, two variations of crawling gaits and a spacial climbing gait. This time, the robot has two additional flexible tail appendages that provided stability and guided curling trajectory. How does it perform a ballistic roll? See the following snapshots for yourself! It's quite obvious why it has to "Go-Q"I apologize for the bad contrast on my robot. I simply forgot to mix in rubber dye and sparkles! Next GoQBot will definitely dress up in a flamboyant color with pink sparkles (maybe some fluorescent markers for kinematics analysis as well). It might be Go-Cute Bot instead!

My discipline does not fracture

This past fortnight was a tough one, for I managed to fracture my right foot two weeks ago. I was immobilized for 4 days before I couldn't stand for anymore home arrest and went back to the laboratory anyway. Indeed, it was a hard time for everything from financial impact to simply getting somewhere. I deeply thank all my friends who helped me survive. I shall get back on my feet very soon.

Well, I am not about to continue this unfortunate story although that was part of what I had to deal with last two week. Since I spent more time at home due to the injury, I started to work on that pile of papers that I "was" supposed to read... and guess what? These were papers about the kind of skeleton that doesn't fracture: hydrostatic skeleton!!

According to my ground reaction forces analysis of caterpillar crawling, hydrostatic skeleton was not active during normal locomotion (manuscript still under review at J Expt Biol). However, we know that Manduca caterpillar can pressurize itself substantially and cast about with precise control. Then where is that switch that "turns on" the hydrostatic skeleton?
Well, I thought I would look into discussions on body pressure of all the major soft-bodied animal systems. In annelids [leeches, earthworms, sandworms...etcc], the cylindrical body is covered by very tough layers of circumferential and longitudinal muscles. Segments are separated by septum muscles which constrict the flow of incompressible coelomic fluid. This type of hydrostat has been modeled and studied for a long time. Deformation can be mapped given that the body volume and pressure are known. Nematodes, on the other hand, does not have circumferential musculature. But they tend to have much higher internal pressure and some evidence suggests that their body wall has very high residual stress. In this model, longitudinal muscles are working against the highly stressed elastic cuticle through the incompressible body fluid. Now, what is the case in caterpillars?
Unlike worms, caterpillar cannot breath through the skin; it has to ventilate via an extensive trachea system. Morphologically, these are significant air cavities open to the exterior. The body cannot be "incompressible" even if the body fluid is. In fact, air bubbles are expelled when a caterpillar crawls underwater. In addition, caterpillars don't have circumferential muscles. To pressurize the body, a caterpillar can do two things: close the spiracles, and compress the body fluid. Indeed, when we try to tear an caterpillar off its substrate, the animal gets shorter and stiffer. However, what can we say about other caterpillar behaviors where the body seems to extend beyond resting length and still maintain body pressure. I suggest a flattening action via the oblique muscles that cause the animals to compress body fluid. Some EMG data might support it, but there is no sufficient proof at this moment. How much compression is require to pressurize the hydrostatic skeleton is totally unclear?

There will be more robotics coming up in the next posting in two weeks I promise! I've been working my leg off (... almost) on the robot radio control system and other robot behavioral integration.

InchBot-VII starts to climb!!

After two weeks of endeavor (sleeping in the lab etc...), I finally convinced the new InchBot to climb up a steep incline. According to the animal locomotion literature, climbing is defined as moving up a incline over 45 degree. Well, currently this InchBot-VII can handle just over 45 degree... so it's a climbing robot for sure!!
This 124mm long robot has two batteries in the head capsule and the rear capsule respectively. So it does not need to be tethered really. It weighed less than 4g if we include the R/C control circuits. It climbs with three sticky pads and several morphological features. The climbing gait was adopted directly from the principles of motion in Manduca caterpillar. Stay tune to my upcoming JEB paper titled: Substrate as skeleton: ground reaction forces from a soft-bodied legged animal for details.
Oh...this robot also got some pink sparkles in the body which made it kind of cute!
Go InchBot!!

Sticky Pads... for climbing robots

In order to improve crawling efficiency, we started to develop controllable grippers for the caterpillar robot. There are three common mechanisms for gripping: Hooking, Adhesion, and Suction. For the scale of our robot, micro-hooks array and adhesion pads seemed most probable. Manduca caterpillars use the former(crochets) while most insects employ both. However, there are great challenges in both systems.

To effectively dig into substrate of non-uniform stiffness, the micro-hooks array has to vary in hook size and compliance. Up to date, we have very little knowledge about these hooks arrangement and properties. Let alone the task of manufacturing such an micro-array for load bearing.
On the other hand, sticky pads are easier to produce. There are hundreds if not thousands of different adhesive materials to choose from. However, release of sticky pads may be quite a challenge for small robots.

After evaluating the situation, I decided to design a membrane with sticky substance on it. Once I master how to control such a membrane, I could control the sticky pads. Then, micro-hooks array can be incorporated into the system.

Akimidis caterpillars

People often ask me what I do with caterpillars and poke their head toward my various apparatus. Well, this one is obvious! "Why are you dipping caterpillars into water?"

Together with a hard working undergraduate research student, we perform an simple experiment what Akimidis would have appreciated very much. By dipping the caterpillar progressively, we could map the volume contribution from head to tail by quantifying the buoyancy. This simple method easily beats the fancy 3D laser scanning, MRI imaging, and X-ray data our predecessors attempted before us in this lab. The only special skill is to tam the caterpillars to stay still during the measurment. Well, this is what caterpillars do in nature -- "I'm not here!!" Nevertheless, cold water can be irritating, thus controling the temperature is critical. The total emersion directly gives the overal body density of any given caterpillar.

Another collateral experiment was something Sir Issac Newton would have been quite interested. We balance a lightweight beam with the caterpillar on it by providing a pivot support below the beam. By measuring the moment force at one end, and shift the pivot across the animal, the mass distribution can be easiliy calculated. The first immediate result is the center of mass for the animal. Combining this data with the above volume map, we can reconstruct a density map across the animal body length.

Density map and mass distribution are two crucial parameters for any biomehcnaical model of a soft body. In addition, the density change across different animal sizes indirectly reflect the trachael volume and respiratory capacity, since most other tissues in the body are similar to water density. Open gas cavities also affect the use of hydrostatic skeleton for a soft-bodied animal. If some caterpillars don't use their hydrostatic skeleton for locomotion, maybe their bodies are too leaky for economic pressurization.

PS: The caterpillar in the lower photo was groomming after a water bath!!

My UK trip

Having been back to Boston just last night, my first surprise this morning was seeing my chaotic room from my bed. I guess the hectic preparation for this UK trip had sped up the entropy evolution since the second week of June. Nevertheless, I was not discouraged by the mess a bit. What I gained the most from this trip was "motivation". Perhaps meeting more people who appreciate my research made my doing more meaningful. After transfering the dirty clothes from the suitcase to the laundry machine, I started cleaning my room. In the afternoon, I headed straight to the lab to pick up my projects. It's so good to feel motivated.

By the way, today I observed a caterpillar that curves around a rod, trying to crawl onto its own back. As it encountered its dorsal horn, it took a bite. I have seen dogs chasing their own tails but seeing a hornworm trying to eat its own horn was really something more attractive than a gossip. Perhaps this particular caterpillar was somewhat unusual, but it certainly made me doubt caterpillars' self-awareness.

A new behavioral experiment --- walk the substrate

From my caterpillar GRF data, I discovered that in almost any instance the substrate is stretching the animal axially. I was intrigued by the idea that such a soft-bodied animal might be actually "depending" on the substrate stiffness to achieve neccessary deformation during locomotion. So, I thought maybe the stiffness of the substrate would affect the crawling performance in a very predictive way. The first test that came to my mind was to suspend the animal from the head capsule and the rear horn (see photos). By feeding the animal different substrates, they could "walk the substrate backward" by doing their normal forward locomotion in my suspension setup. Interestingly, they crawl just fine with a balsa wood. As I reduced the stiffness of the substrate by changing from wood to rubber and eventually to a soft wire, the caterpillar started to buckle the substrate and failed to produce the stereotypic locomotion pattern. This is the first behaviroal support for the "environmental skeleton hypothesis".

The roles of thoracic legs in caterpillar locomotion

We all know that the six thoracic legs are the true legs that will be retained through metamorphosis in lepidoptera. However, amputation of these limbs does not seem to prevent a Manduca caterpillar from crawling around. Nevertheless, caterpillars appear to use them constantly during each crawl cycle. Originally we considered them as probing devices for sensing the substrate ahead, until I analized the GRF data from thoracic legs. Besides that fact that each leg pair could take up as much body weight as any pair of prolegs, they also exert quite significant amount of forward pull during each crawl. What happened to the amputated caterpillars I cannot figure, but thoracic legs definitely have a important role in normal locomotor performance.

Soft-bodied robotics

For the past few weeks, robotics has taken over my life once more. This time, I challenged the scale and stability with my new designs of soft-bodied inchworm robot. The previous crawling green foam known as InchBot-III (classic): FoamBot. It has been successfully modeled in a FEA implementation with high accuracy.

In order to scale down and maintain the same mechanics, I needed to change the material, or body bending mechanism. For that, I developed the next generation robot InchBot-IV (mini): LeechBot that reversed the working principles of Mckibben artificial muscle. By weaving the SMA into the tubular braid, the structure could create suction durint contraction at two ends, very much like a leech. The suction adhesion morphology is currently under design.


By changing the material, I was able get another inching robot. This little guy measured no longer than 80mm and no wider than 5mm. A newly featured membranous wing allows better adhesion and lateral stability. This winged inchworm was the InchBot-V (nano): InchFly. It does not fly at this moment. But one could easily imaging such a gliding potential from the wing area to mass ratio.

Predatory caterpillars

Recently, I came across some literature about predatory caterpillars. I guess lepidopteran larvae are not strictly herbivorous as I assumed. Several ambush predator caterpillars have been found in Hawaii including the green grappler Eupithecia. A YouTube video demonstrates how this kind of caterpillar modifies the strike reflex for ambushing a pasing termite. They have modified thoracic legs and A6 prolegs for grabing the prey and manuveuring the mostly airborn body.


More recently, a paper in Science reported a case-bearing caterpillar named "Hyposmocoma molluscivora" which feeds exclusivly on snails. This kind of caterpillar spins silk over a resting snail in a spiderlike fashion in order to anchor the prey. Then it would start digging into the shell opening all the way to consume the snail, even if that means leaving its silk case behind.


Another class of caterpillar carnivorousity is consipecific cannibalisim when the population density is too high. Semlitsch and West described the relationship between body size and caterpilalr cannibalism in the journal Oecologia. It was found that smaller caterpillars are more likely to become the victims simply because they are not capable of killing the bigger conspecifics.

Reconsidering "caterpillar locomotion"

Want to guess the highest speed by which a caterpillar can move?
Answer: ~15" per second!!!
No way... what species and how?

I recently reviewed a few papers about rolling locomotion in nature, and Mother-of-Pearl moth caterpillar (Pleurotya ruralis) is one of the two active rollers ever discovered. Check out a YouTube clip by BBC Aniamls and you will believe what I claimed. In this movie, the caterpillar was rolling downhill, but the rolling action is actually initiated by rapid muscle contraction. According to Dr. John Brackenbury reported in his paper "Fast locomotion in caterpillars", this active rolling could hit a top speed of 39cm/s.



Here are a couple of references for those who want to read into this subject.
J. Brackenbury. Fast locomotion in caterpillars. Journal of Insect Physiology. 45:525-533 (1999)
J. Brackenbury. Caterpillar Kinematics. Nature. 390:453 (1997)

Three forms of caterpillar and their locomotion

Manduca caterpillar only represents a class of Lepidoptera with four abdominal prolegs. Many other species have reduced the size or number of prolegs. Geometrids (loopers, inchworms, and spanworms), for example, retain only the A6 abdominal prolegs which work in conjunction with the terminal (anal) prolegs to facilitate the rear end attachment during a looping action. Lepidoptera larvae in the limacodid-group have very short legs at the belly that their modes of locomotion resemble slug creeping in apparence. For these so called "slug caterpillars", there can be abdominal prolegs from A2 to A7 with a combination of crochets and suckers. After starring at the "typical" 4-Abd prolegs caterpillars for three years, I start to wonder how the differences among these locomotor modes are associated with the evolution of these three forms of caterpillars. First thing that came to my mind was the mass scaling issue. According to my experiences with live caterpillars in the wild, 4 Abd body plan is common in species that can get really big and juicy. Inchworms are rarely bigger than 5cm. Slug caterpillars seems even smaller.

Larger caterpillars --- 4 Abd prolegs (use as many body anchors as the locomotion allows)
Loopers/Inchworm --- 1~3 Abd prolegs (reduce unneccessary attachements for efficiency)
Slug caterpillars --- short prolegs close to the ventrum (increase contact surface for fluid/silk adhesion mechanism)

This line of reasoning is very consistent with my current biomechanical finding: prolegs=attachments. (results from the caterpillar substrate reaction force analysis) Since soft legs cannot provide effective leverage, they are just acting as anchors for caterpillars. And it follows that the evolutionary modifications of prolegs morphologies should be determined by the maximum body size and mechanisms of attachement.

Body pressure controled by the smart integuement...

Integument of soft-bodied animals is often a very interesting biomaterial. Besides unfolding and stretching to accommodate growth, it may function as a body pressure monitor in some instances. More specifically, the viscoelastic effect of the Manduca body wall prevent stress from building up as the body volume increases. When the caterpillar lost body volume, the body wall will automatically shrink to maintain the body pressure. The above effect might sound like a typical barometric reflex in mammals, except that this is done via the passive properties of the cuticular integument. Indeed, a simple reflex may be substituted by a smart material! The following graph shows the "adaptation" response found in the body wall, muscles, and the stretch receptor organ lining along the body axis. More detailed experimentation on caterpillar body pressure is ongoing.

Manduca goes aquatic!!

Caterpillar locomotion is slow and quasi-static. However, it is very robust over a wide range of environmental conditions. In addition to maneuvering in highly branched vegetation, caterpillars can burrow underground and even crawl underwater. In my experiment, water submersion converts gravity into buoyancy and increases the external pressure by a factor of two. Nevertheless, a Manduca 5th instar caterpillar can continue to function up to ~3min before the hypoxia kicks in.

InchBot-III comes alive!

InchBot is a biomimetic soft-bodied inchworm robot actuated by shape memory alloy springs, controlled by a pair of coupled neural oscillators, and powered by a Li-Po battery. It is completely autonomous with the ability to conform to the substrate as a soft body. The exploitation of this compliance allows the robot to operate without feedback over a range of terrain variations, much like the real animal. InchBot-III can crawl up to 5mm/s on a tabletop and carry a payload backpack up to 15g (bottom photo). The whole robot can be rolled up and compressed into one palm (top picture).

Welcome to Huai-Ti's research blog!!

Hi there!

In the field of animal locomotion, fast ballistic motions are always eye-catching. However, on the other end of the spectrum, slow soft body movements can be equally amazing. My research focuses on force transmission in soft-bodied animal locomotion. In particular, we use Manduca sexta caterpillar as the model system.

In response to the feedbacks and requests from recent encounters at professional conferences, I decided to start this blog as a window for people to check on my research progress. I will do my best to update new cool graphics and info about my research. Please note that this site is not intended for formal publication. I cannot post any unpublished data or engineering details. I apologize if the materials cannot satisfy your curiosity, but I promise to publish my research as fast as I can. Thanks for the interest.

HTL