Robotic Soft Material

Invented by Manoj Sahi

Imagine a future where a material itself is a robot that can sense, actuate, compute and communicate!


Mattrbot is an early stage robotic soft material that is developed through machine learning for humanoids, exosuits, autonomous vehicles, cobots and energy harvesting robots.


Many of the emerging robots are heavily reliant on Silicone. This limitation of material choice leads to limited forms, behaviour, intelligence and applications of robots. The first version of Mattrbot is a robotic material that is better than Silicone. Silicone is considered as a pollutant and an unwelcome material for the other forms of life. A future where robots are based on silicone is not desirable. Mattrbot has nine important characteristics with respect to environmental, user and technical perspective which silicone doesn’t.


Mattrbot comes with MattrBox and MattrBin.

MattrBox is a system where Mattrbot is created. The starting point is through machine learning. Millions of materials are organized in material libraries. From these libraries, machine learning picks the right ingredients based on selection criteria to formulate Mattrbot which is then simulated with accurately predicted properties.

In the next step, selected ingredients are combined and 3D printed if required. This process enables the combined material to acquire functions such as shape memory, self-healing, conductive, light emitting and piezoelectric. The custom made Mattrbot with intriguing properties can go to various robotic applications.

Upon the end of application lifecycle, retired Mattrbots are moved to MattrBin. Here, Mattrbots are subjected to triggered degradation which is initiated with a light source. Recovered materials are sent back to the material library. Thus, the full circle of Mattrbot completes.