AssetID: 54923285
Headline: RAW VIDEO: Twinkle Trunk! Researchers Create Elephant Robot That Can Bowl
Caption: Researchers at EPFL (École Polytechnique Fédérale de Lausanne) have created a robot elephant that can go ten pin bowling. The twinkle-trunked bot does have a vital purpose - showing how 3D-printing can be utilised to create muscles, tendons, ligaments and bones that work in concert to deliver energy, precision and a vast range of motion. Up to now, engineers have largely relied on multi-material 3D printing to create robots that combine hard and soft elements. While this method can emulate the range of biological tissue types, it does not allow for continuous control over properties such as stiffness or load-bearing strength throughout a structure. But researchers at EPFL (École Polytechnique Fédérale de Lausanne) believe they have found a new approach. The team led by Josie Hughes from the Computational Robot Design and Fabrication Lab (CREATE), part of EPFL’s School of Engineering, has developed an innovative programmable lattice made from simple foam material. Each lattice is made up of individual ‘cells’ that can be configured in more than a million ways – and potentially combined into infinite geometric variations. “We used our programmable lattice technique to build a musculoskeletal-inspired elephant robot with a soft trunk that can twist, bend and rotate, as well as more rigid hip, knee, and foot joints,” says postdoctoral researcher Qinghua Guan. “This shows that our method offers a scalable solution for designing unprecedentedly lightweight, adaptable robots.” They tested these advances in various ways - including sending the elephant to a bowling alley. The breakthrough lies in how the researchers can ‘programme’ the geometry of each cell in two dimensions: shape and position. The team uses two main cell types – the body-centred cubic (BCC) cell and the X-cube – each with different mechanical characteristics. BCC cells are typically softer and more elastic, while X-cube cells offer greater rigidity and strength. However, rather than being limited to these two configurations, the CREATE team has developed a method to create hybrid cells with shapes that blend anywhere between BCC and X-cube – allowing for smooth transitions across a robotic limb, much like in real muscle and bone. “This approach enables the continuous spatial blending of stiffness profiles and allows for an infinite range of blended unit cells. It’s particularly suited for replicating the structure of muscular organs like an elephant trunk,” explains PhD student Benhui Dai. The team can also programme the position of each cell within the lattice, allowing them to rotate or shift cells along different axes. In more complex structures, they can even superimpose cells to create entirely new combinations. A cube with just four overlapping cells can yield around four million configurations, while five superimposed cells can produce more than 75 million. This new level of control enabled the EPFL researchers to replicate a wide range of biological joint types in their elephant model, including a sliding joint (akin to bones in the foot), a bending uniaxial joint (like the knee), and a two-way bending biaxial joint (as seen in the toes). The elephant’s trunk – a marvel of muscular engineering in nature – was also modelled using lattice sections specially designed for twisting, bending and rotating, all connected through fluid, seamless transitions. Looking ahead, the research team sees vast potential in their foam-based lattice approach. “Like honeycomb, the strength-to-weight ratio of the lattice can be very high, enabling very lightweight and efficient robots,” says Josie Hughes. “The open foam structure is well-suited for motion in fluids, and even offers potential for including other materials, like sensors, within the structure to provide further intelligence to foams.” With the flexibility to embed sensors, waterproof materials and further functionality, the lattice technique could open the door to a new generation of adaptable, intelligent robots – ones that move and behave less like machines and more like living creatures - and that can bowl like the best of us. Their findings have recently been published in Science Advances.
Keywords: feature,photo,video,elephant,robot,tech,technology
PersonInImage: