Research and Advances
Artificial Intelligence and Machine Learning

Physically Based Computer Animation: Introduction

How can images be made to create the visual illusion of physical force?
Posted
  1. Article
  2. Author

Welcome to an artificial world that’s beginning to behave as naturally as the real one.

By incorporating the basic laws of physics—friction, gravity, momentum, drag, pressure—physically based computer animations are beginning to run like dinosaurs, jump like acrobats, flow like water, crack like glass, swirl like mist, even grow and evolve like a tree. These animations are finding a commercial role in industrial simulations and remote sensing and climate modeling, as well as in their most recognizable and popular application—entertainment, from feature films to special effects, to video games, to television commercials and programming, to museum exhibits. In some cases, animated characters are even beginning to plan their own physical and sensing actions, escaping trouble and avoiding physical barriers. Their survival strategies can be based on knowledge acquired from their experience in their virtual domains.

Traditional special effects, even many game sequences, startle the viewer by producing explosions of color and light. Now, effects based on physical simulation might mean a realistic-looking blast and a virtual shock wave hitting nearby structures and passersby. They might also mean a collision between a race car and a wall, propelling shattered parts tumbling down the track. The authors here have not only generated some of the most striking of these images; they are themselves the sources of some of the most innovative simulation and animation tools and approaches available today.

For the non-animator, these approaches might be harnessed to help design better physical products, explain the mysteries of turbulent fluids, from blood to air, refine physical processes, and predict the weather, as well as entertain. One thing about animations unlikely to change, however, is how incredibly time-consuming and labor-intensive they are, as the production credits of movies like Antz, Dinosaur, Stuart Little, The Prince of Egypt, and Toy Story 2 attest.

Looking at how this technology is altering the computer game experience, Hecker describes the often-critical role of physics in today’s multibillion-dollar interactive electronic entertainment industry. Just to keep pace with game players’ expectations, game designers have refined what Hecker calls their games’ built-in “physics simulators” to deliver an increasingly realistic—and fun—experience. As an example of a consistent simulator, Hecker views the classic game Tetris as flawless.

Exploring cognitive modeling for games and animation, Funge doesn’t just add physical realism but gives his autonomous graphical characters the cognitive and sensory capacity to make decisions in (presumably) the best interests of achieving whatever their assigned role happens to be, survival included. The result is increasingly lively and reactive characters and an entertaining and compelling experience, though perhaps less control for the designer.

A challenging question Hecker asks about the physical mobility of such characters concerns how they might walk, run, and jump realistically. Popovi cacute_l.gif developed and now shares his solution by adapting existing motion, through what he calls “motion transformation,” allowing animators to transform input motion sequences to reflect characters’ muscles and bones in motion. The result might be an animated human character that limps, jumps, or walks on the moon. I’m prompted to wonder when synthetic characters might be indistinguishable from real actors.

Covering such environmental staples as water, fracture, and smoke and fire, Foster and Metaxas, O’Brien and Hodgins, and Stam next offer their methods and algorithms for producing convincing simulations. Imagine being able to send ripples through a pool of virtual water, shatter a virtual china teapot, and interact with virtual smoke and fire in real time. The resulting torrent of calculations generates images that look like the real thing to our highly skeptical visual perception.

Acknowledging the limitations of such computer graphics techniques for engineering disciplines, Foster and Metaxas describe a way to simulate flowing water. As demonstrated in the 1998 feature film Antz, a spectacular final sequence, which Foster created, sends a flood roaring and crashing through an ant colony. Such complex pictures are as dazzling as the real-world phenomena they are intended to represent—and a great tool for creating special effects.

Also focusing on realistic animation rather than rigorous physical correctness, O’Brien and Hodgins seek to predict and measure realism by comparing their simulated fractures with real fractures of comparable but real objects, including drinking glasses and china bowls, as shown alongside their simulated real counterparts.

Returning to the challenge of simulating fluids, Stam describes his way of interacting with smoke and fire, even clouds and air, in real time. The solution’s speed, stability, and visual quality suggest a role in such engineering applications as manufacturing design, evaluating, say, the fluid flows past automobile and airplane bodies during shape design. In architecture, they could be used for estimating the circulation of air throughout a building. Finally, Prusinkiewicz explores what may seem among the most specialized of simulations: modeling plants and plant ecosystems. But his inspiring introduction to the field is a revelation, exploring the synergy of biology, computer science, and art. Besides covering how to depict botanically accurate plants, he shows how to let them grow themselves into self-sustaining ecosystems.

The most popular and lucrative application to date of all these technologies is in entertainment media, especially, it seems, for generating photo-realistic 3D digital dinosaurs on a grand scale in intricate detail. The much deeper result, as Prusinkiewicz writes, is not only the visual beauty of nature but insights into the way nature works. Still, the field needs to verify by experiment and technical analysis the accuracy of the simulations’ physical behavior and visual representation. Verification might allow the technology to transfer to such engineering fields as mechanics, hydraulics, aerospace, turbulence, and fluid flow. Other beneficiaries are likely to include the mass market of PC users as well, not just game players. Despite being unskilled animators, these people might then be able to adapt ready-made images and techniques to animate any story they can imagine.

Back to Top

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More