Your data. Your choice.

If you select «Essential cookies only», we’ll use cookies and similar technologies to collect information about your device and how you use our website. We need this information to allow you to log in securely and use basic functions such as the shopping cart.

By accepting all cookies, you’re allowing us to use this data to show you personalised offers, improve our website, and display targeted adverts on our website and on other websites or apps. Some data may also be shared with third parties and advertising partners as part of this process.

@harvardengineering / YouTube
News + Trends

New 3D printing technology improves the control of soft robots

Kim Muntinga
10/2/2026
Translation: machine translated

A new 3D printing process from Harvard anchors the movement of soft robots directly in the material. Instead of time-consuming readjustment, the desired deformation is created during printing.

Imagine you print out a robot, pump air into it and it folds exactly as you planned: it bends to the left, grips, unfolds like a flower.

This is exactly what a team from the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) has managed to do. The researchers have developed a method with which soft robots can be produced in a single printing process and their movement behaviour is already anchored in the material. The study was published on 6 February 2026 in the journal Advanced Materials.

The problem with soft robots

Soft robots are made of flexible, often biocompatible materials and are seen as a promising alternative to rigid machines. They can interact with sensitive objects, adapt to the human body and operate in confined or irregular environments, for example in minimally invasive operations. However, the very feature that makes them so useful is also their biggest challenge: their softness.

In the past, anyone wanting to build a soft robot for a specific task had to produce complex moulds, pour silicone layer by layer, apply pneumatic channels to surfaces and connect several components together. This is time-consuming, not very flexible and hardly scalable. Every adjustment to the design requires new moulds, new casting runs and new tests.

The result: even experienced teams need many iterations before a soft robot behaves as it should. And even then, the movement behaviour often remains difficult to predict.

Rotation is the key

The Harvard team solves this problem with a process called «Rotational Multimaterial 3D Printing» (RM 3DP). The idea behind it is elegant: a single nozzle dispenses two materials simultaneously and rotates during the printing process. This rotation controls where which material ends up within the printed filament.

The image series documents the path from computational path planning to printing and controlled pneumatic moulding.
The image series documents the path from computational path planning to printing and controlled pneumatic moulding.
Source: SEAS

The researchers print filaments with a flexible outer shell made of polyurethane and an inner channel made of poloxamer, a polymer that is also found in hair gel. As soon as the outer shell hardens, they wash the gel out of the inside. What remains are hollow tubular structures, i.e. precisely positioned channels inside the material.

The researchers determine the position, shape and size of each individual inner channel by precisely controlling the rotational speed, material flow and nozzle geometry. When air is pumped into these channels, the structure deforms in exactly the direction that was programmed during printing.

We use two materials from a single opening that can be rotated to programme the direction in which the robot bends when inflated.
Forscher Jackson Wilt

From theory to demonstration model

To test the process, the team printed two demonstration objects: both in a single, uninterrupted print path, without separate assembly.

The first is a spiral actuator in a flower pattern: when air is introduced, it unfurls like an opening flower. The second is a five-fingered hand gripper with defined knuckle joints that wraps around objects when inflated. Both were created in a continuous 3D-printed path, without any separate assembly steps.

In a way, the geometry replaces the programme code. Changing the shape of the channel changes the movement behaviour of the robot. This makes the process exceptionally flexible:

We don't have a mould. We print the structures, programme them quickly and can adjust the actuation quickly.
Forscher Jackson Wilt

What this means in practice

The application possibilities extend far beyond laboratory demonstrators. As the structures are made of flexible and potentially biocompatible materials, the technology could be useful in surgical robotics, assistance devices and in the human-machine interface.

In medicine, it is possible to imagine instruments that unfold inside the body in a targeted manner, without any rigid components that could damage tissue. In production, grippers could handle fragile objects without damaging them. And in assistive technology, soft exoskeletons or orthoses could be created that dynamically adapt to the wearer's body.

A process with scaling potential

What sets the process apart from many other approaches is its scalability. The printing parameters - rotation speed, flow rate, nozzle geometry - can be adapted on the software side without having to modify the hardware. A new design does not require a new mould, only new printing parameters. This significantly shortens the development cycle.

Header image: @harvardengineering / YouTube

6 people like this article


User Avatar
User Avatar

My interests are varied, I just like to enjoy life. Always on the lookout for news about darts, gaming, films and series.


News + Trends

From the latest iPhone to the return of 80s fashion. The editorial team will help you make sense of it all.

Show all

These articles might also interest you

  • News + Trends

    Roller, mop and mop plate: Roborock launches three new top models at once

    by Lorenz Keller

  • News + Trends

    Spielwarenmesse Nürnberg: These are the 6 winners of the Toy Awards 2026

    by Stephan Lamprecht

  • Background information

    My first time with resin: how SLA works

    by Kevin Hofer

3 comments

Avatar
later