Tensegrity in Computer Generated Imagery (CGI)

Simulation of a tensegrity body model can be used to create realistic-looking biological movement for video games and animated films. This can be done with unactuated tensegrity structures: satisfactory video can be produced from simulations in which the tensegrity body model is pushed into motion.

Context: Tom Flemons Archive

A double-layered tensegrity model for efficient CGI rendering

(May 25, 2015 repeated from Layered Design for a Full-Body Tensegrity Model) We also talked about the application of tensegrity models to improve CGI rendering in programs such as Maya. I explained to Perry about my theory that the best way to model vertebrate anatomy i.e. quadrupeds and bipeds is to build bodies with two layers of tensegrity. …[full text in Layered Design]… I pointed out that such meshes enveloping a stick model would create a fairly life like rendering of a body in motion without a lot of computation needed – the form would naturally move like a living being does by nature of its tensegral geometry. Perry felt that applying this principle to a modeling engine for CGI would prove to be revolutionary. I have no idea whether it is possible to build a tensegrity rig which would speed up rendering and save studios millions in the cost of making movies and in the gaming world would allow revolutionary life-like creations of avatars who moved with the fluidity of living beings but he must. If he thinks it’s possible then it’s probably worth investigating.

Simulate a tensegrity structure returning to equilibrium from a stressed position

(May 29, 2015)  Perry, Snelson talks about 3D weaving. It’s an important paper for what I’m talking about. Also check out Tachi’s work [Tomohiro Tachi (2012) Interactive Freeform Design of Tensegrity]. This is what I had in mind when I was talking about meshes. I read this paper some time ago and was excited to see how they derived a tensegrity envelope for anything. Like Maya make a mesh but then do the next thing and turn that into a tensegrity grid. If it is interactive like in NTRT then you have a structure that will respond to pushes and return to homeostasis. As it returns to equilibrium from a stressed position it follows a minimal energy path and oscillates like those youtubes I showed you of Gerald de Jong’s. This renders the simulations very believable.

If you have to create the complex form bit by bit like I make my models – there is a lot of trial and error – different lengths of struts and cables, number of struts, types of geometry etc. If an approach that combined a mesh generator with a tensegrity grid transformation along the lines of this paper and Dorothea’s idea to overconnect a tensegrity and let an evolutionary algorithm pare down the structure to the minimum form – you could very rapidly generate a lot of different forms – it would somewhat automate form finding I suspect…

Create illusion of directed movement by reversing the time sequence

(May 29, 2015, continued) Keep in mind that a perfectly balanced tensegrity has just enough prestress to hold its shape – it is a perfect minimal energy machine. We exist best in a low prestress state – a perfectly toned and balanced body is in a low energy state of equilibrium. All tensegrities given an opportunity return to a balanced state after having a force impact them. If you ran the time sequence in reverse you would have the illusion of a controlled (and yet fluid) directed movement. e.g. you could model an arm attached tensegrally to a torso throwing a punch by stretching it out in the direction of the punch and then letting go… run in reverse I’m betting it would look pretty smooth. So this principle could be applied to any moving figure.

Bipedal gait discovery for film animations

The impressive 2013 bipedal gait discovery work by Geijtenbeek et al. uses a genetic algorithm  to discover a fascinating variety of bipedal gaits for a body model consisting of rigid blocks linked by articulations. More compliant and biologically-plausible motions would result if a tensegrity body model were used instead. Tom comments on the feasibility of doing this:

(Feb 22, 2014) I am not sure if there is much reason to expect a tensegrity modelling of bipedal gait from this team of researchers. What they have achieved is an imitation of human gait that seems to predict and resemble a version of bipedal motion that perhaps has already paid off in such robotic mimicry found in Boston Dynamics (now Google) robots Petman and Big Dog. While I think it is necessary and should be possible to build a biotensegrity computer model I don’t know if anyone knows how to go about this.

Tom made progress toward creating a tensegrity model of an entire organism, including this 2015 biped design. A small number of parameters might suffice for pushing a full-body tensegrity model into motion, as evidenced by Tom’s use of a single control rod to push his tensegrity torso through a large movement repertoire.