Want to speak at AMA: Energy 2025 or AMA: Automotive & Mobility 2025? Submit your application now!
Researchers at Ecole Polytechnique Federale de Lausanne (EPFL) have developed a 3D printed robot that mimics the mechanical complexity of muscles and bones using just one material.
Inspired by elephants, the robot combines soft, flexible components with rigid, load-bearing structures without switching materials. Led by Josie Hughes in the Computational Robot Design and Fabrication Lab (CREATE) in EPFL’s School of Engineering, the novel development lies in controlling internal lattice geometries to produce a wide range of mechanical behaviors within a single elastic resin.
Published in Science Advances, the study introduces programmable lattice design as a way to replicate musculoskeletal systems with tunable stiffness and directionality. Using two methods, topology regulation and superposition programming, the team achieved Young’s moduli from 25 to 300 kPa and shear moduli from 1.38 to 40 kPa, covering the stiffness spectrum from soft tissue to cartilage-like rigidity.
Lattice-based stiffness programming enables complex motion
To bring the design to life, the researchers developed a computational pipeline using custom MATLAB scripts to map motion and stiffness requirements into programmable lattice geometries. Designs were exported as STL files via OpenSCAD and Hob3l, then printed on a Halot-Mage Pro 3D printer using F80 elastic resin from Godsaid Technology. Tendon-driven actuation was implemented with Bowden cables and Dynamixel servo motors, also controlled through MATLAB.
With this setup, topology regulation was used to construct the robot’s trunk. This method continuously adjusts stiffness by blending two lattice types, bcc and XCube, allowing the trunk to be divided into three sections for bending, twisting, and helical motion, all powered by four motors.
A parameter called the topology index controlled transitions between soft and rigid zones, enabling the tip to use fine, thin cells for delicate gripping and the base to provide structural support. Weighing just 150 g, the trunk could lift up to 500 g and handle objects ranging from 0.1 mm to 100 mm in diameter.
Superposition programming created the rigid joint structures in the robot’s legs. This method combines unit cells with varied orientations and translations to produce discrete, directional stiffness. The legs feature active joints at the hip and knee, and a passive ankle that adjusts to ground contact.
Controlled by two motors and four tendons, the hip enables flexion, extension, abduction, and adduction, while the knee uses a single motor. Capable of supporting up to 4 kg, more than the robot’s 3.89 kg body weight, the legs enable walking with step lengths of about 150 mm at speeds of 7.5 mm/s.
The robot demonstrated both forward and lateral gaits and maintained balance while standing on three legs. The feet were designed with stiffer lattice regions at the front for weight-bearing and softer regions near the heel for ground conformity. The open lattice design reduced overall weight and allowed the robot to function in water without modifications.
Mechanical testing confirmed how changes in beam thickness, cell type, and arrangement influence stiffness and anisotropy. The trunk’s twisting segment achieved rotation angles up to 78.1°, while the bending module showed a 30% increase in range compared to a uniform structure. Over one million unique lattice configurations were generated using the design methods, and the number could exceed 75 million by expanding the underlying geometric variations.
This work offers a scalable way to embed mechanical intelligence directly into a robot’s structure. Future versions may integrate sensors, fluids, or other components to expand into soft robotics, prosthetics, and lightweight systems.
Bio-inspired efforts in 3D printed robotics
Research into the additive manufacturing of bio-inspired robotics has taken on a variety of forms in recent years.
In February, researchers from the University of Twente (UT) and the University of Southern Denmark (SDU) developed a low-cost method to strengthen the bond between soft and rigid materials in hybrid robots using standard FDM 3D printers.
By intentionally inducing under-extrusion, they create a porous interface that mimics biological connective tissue, improving stress distribution and adhesion. Their approach outperformed traditional adhesives by up to 200% in lap shear and peel tests, and withstood three times more pressure in pneumatic tests. This bio-inspired solution could significantly improve the durability and accessibility of soft-rigid robotic systems.
On another note, Cornell University researchers developed a bio-inspired 3D printed soft robotic muscle that regulated its internal temperature through synthetic sweating. Using hydrogel-based composite resins and stereolithography (SLA), they fabricated fluidic elastomer actuators with pores that opened and closed in response to heat.
As temperatures rose, the pores released water, enabling cooling over 600% faster than non-sweating equivalents. The thermal regulation was entirely material-driven, requiring no sensors. While the sweating prevented overheating during grip tests, it reduced surface grip, prompting future plans to adjust hydrogel texture.
What 3D printing trends should you watch out for in 2025?
How is the future of 3D printing shaping up?
To stay up to date with the latest 3D printing news, don’t forget to subscribe to the 3D Printing Industry newsletter or follow us on Twitter, or like our page on Facebook.
While you’re here, why not subscribe to our Youtube channel? Featuring discussion, debriefs, video shorts, and webinar replays.
Featured image shows optical image showcasing the physical appearance of the elephant robot. Photo via EPFL.
