Researchers at MIT have developed a simulation technique that promises to make creating bouncy, stretchy characters for films and video games both more reliable and true to physical laws. By identifying a hidden convexity in the equations that describe material deformation, the team has unlocked a way to model elastic materials—like rubber—with precision and control.
Leticia Mattos Da Silva, an MIT graduate student and the lead researcher on the project, explained, “The way animations look often depends on how accurately we simulate the underlying physics. Our approach not only respects those laws but also offers animation artists greater stability and control.”
For anyone who’s ever wrestled with unstable simulations or unrealistic motion, this breakthrough is a welcome change. Traditional methods often sacrifice physical accuracy for speed, causing energy losses and erratic behavior. In contrast, MIT’s new technique separates the stretch and rotation components of deformation, with the stretch part forming a convex optimization problem that is both more manageable and consistent when solved.
The team’s experiments have demonstrated that their solver reliably simulates a range of elastic behaviours—from bouncing objects to squishy cartoon characters—while keeping vital properties like energy and momentum intact. This stability means that animators can trust the simulation to perform well even under challenging conditions.
While the method may not match the speed of some speed-first tools, its reliability makes it an attractive option for those prioritising accuracy over rapid results. The researchers are already looking to boost computational efficiency, aiming to speed up the process without losing the stability that sets this method apart.
Beyond animation, there’s potential for this technique in engineering and design applications, such as modelling flexible materials for toys, clothing, and other products where precise deformation is essential. As Silva remarked, “We were able to revive an old class of integrators in our work. My guess is there are other problems where uncovering a hidden convexity could bring significant advantages.”
