Elephant Robot demonstrates bioinspired 3D printing technology

A powerful sprint of a cheetah, a supple slither of a snake, or a clever grasp of a human: each is made possible by a seamless interaction between soft and hard tissues. Muscles, tendons, ligaments, and bones work together to provide the energy, accuracy, and range of motion needed to carry out the complex movements seen throughout the animal kingdom.

Replication of this musculoskeletal diversity in robotics is extremely difficult. Until now, 3D printing using multiple materials is one way to create soft, rigid robots, and while this approach may mimic the diversity of biological tissues, it means that critical properties such as stiffness and load-bearing strength cannot be controlled continuously throughout the robot structure.

The team, currently led by Josie Hughes of the EPFL School of Engineering’s Computational Robot Design and Manufacturing Lab (Create), has developed an innovative lattice structure that combines biological tissue diversity with robotic control and accuracy. A lattice made of simple foam material consists of individual units (cells) that can be programmed to have different shapes and positions. These cells can take on over a million different configurations and even combine them to produce endless geometric variations.

“We used programmable lattice technology to build soft trunks that twist, bend and rotate musculoskeletal-inspired elephant robots, as well as stiffer hips, knees and foot joints.” “This shows that our method offers a scalable solution for designing unprecedented lightweight, adaptable robots.”

This study is published in Science Advances.

Two programming dimensions. Infinite geometric variations

Team programmable grids can be printed using two main cell types of different shapes: body-centered cubic (BCC) cells and X-cube. When using each cell type to 3D print a robotic “tissue” the resulting lattice has different stiffness, deformation, and load bearing properties. However, the Create Lab method also allows printing of lattices made of hybrid cells with shapes anywhere on the spectrum between the BCC and the X-Cube.

“This approach allows for continuous spatial blending of stiffness profiles, allowing for an infinite range of mixed unit cells. It is especially suitable for replicating the structure of muscle organs, such as elephant stems,” says Ph.D. Student Ben Huidai.

In addition to modulating the shape of each cell, scientists can also program their positions within the grid. This second programming dimension allows you to rotate and shift (translate) each cell along an axis. Cells can even be superimposed on one another to create an entirely new cell combination, and the resulting lattice gives even broader mechanical properties. To grasp the vast scale of potential variations, a lattice cube with four superimposed cells could result in around 4 million possible configurations, with over 75 million configurations of five cells.

Waterproof and sensor compatible

In the elephant model, this dual programming feature allowed the manufacture of several different tissue types with unique ranges of movement, such as sliding plane joints (found in small bones in the foot), bent uniaxial joints (located in the knees), and two-way bent biaxial joints (found in the TOES).

The team was even able to replicate the complex movements of the elephant muscle trunks, engineering individual lattice sections specialized for torsion, bending and rotating movements, by maintaining smooth, continuous transitions between them.

In addition to altering foam material and incorporating new cell shapes, Hughes says the proprietary foam lattice technology structure offers many exciting possibilities for future robotic research.

“Lattice strength-to-weight ratios are very high, allowing for very light and efficient robots, like honeycombs. The open-form structure is suitable for liquid movement, offering the possibility of including other materials such as sensors in the structure, and even providing additional intelligence in the foam.”