Neurorobotics

 

Contact us 

or check out our Neurorobotics Forum
 
 

Join us

in the EBRAINS Community.

Neurorobotics in HBP and EBRAINS

Neurorobotics is an emerging transdisciplinary field of research at the interface between neuroscience, artificial intelligence and robotics. The Neurorobotic Platform is the solution for embodied simulation developed in the context of the Human Brain Project to be at the core of the EBRAINS Closed-Loop Neuroscience service offered by the EBRAINS Research Infrastructure.

 

The Neurorobotics Platform - NRP

The NRP is an integrative simulation framework that enables in silico experimentation and embodiment of brain models inside virtual agents interacting with realistic simulated environments.

  • It is entirely open source

  • It offers a browser-based graphical user interface for online access

  • It can be installed locally (Docker or source install)

  • It can be interfaced with multiple neuromorphic chips (SpiNNaker, Intel Loihi)

Our community of users includes roboticists, computational neuroscientists, and engineers working on applications of neuromorphic computing. However, anyone with a working knowledge of Python can easily learn how to use our platform.


You can download and install the NRP locally for maximum experimental convenience, or access it online in order to leverage the EBRAINS HPC infrastructure for large-scale experiments.

ACCESS THE NRP

REQUEST ACCESS HERE

How the NRP can support your research

The NRP is an integrative ecosystem, with special consideration given to creating software interfaces and APIs that enable integration of not only simulation engines, but also of tools for design, analytics, databases, and more.

The NRP helps you improve and refine novel models of brain functions and architecture.


The functional connection of neural models to embodied agents allows neuroscientists to explore how the brain performs a number of tasks in closed loop, from lower-level sensorimotor tasks, to higher cognitive functions (e.g. contextual awareness, decision making, etc.).


Physical sensorimotor mouse experiment and simulation on the Neurorobotics Platform

The NRP helps you design robots with enhanced decision-making capabilities and contextual awareness for use in autonomous exploration (deep sea, space), search and rescue in disaster areas, and more.


Environment exploration in the NRP

The NRP supports investigation and testing of interaction with AI-enabled autonomous systems in VR settings for development of novel, cost effective robotic control technologies.

A related publication can be found at the bottom of this page.



Ongoing work by TUM researcher Sandro Weber on VR Re-Embodiment.

The NRP allows you full control of experimental parameters and let you define adequate testing setups for functional evaluation. For instance, simulations of the interactions between an agent and its environment are fundamental in order to implement data-driven model-less approaches to synthesising control policies (e.g. reinforcement learning). These are seen as key to producing robotic systems capable of complex online decision-making or adaptation (e.g. to previously unseen situations, changes in operational parameters, etc.) but they usually require large amounts of data for training purposes. This is not an issue for the NRP, as multiple instances of the latter can indeed be deployed and synchronised on the High-Performance Computing infrastructure of EBRAINS.


Accelerate learning through parallelisation

Important publications
 

Ubi-Interact: A modular approach to connecting systems

Sandro Weber, Marian Ludwig, Gudrun Klinker

ICST Transactions on Mobile Communications and Applications, Vol. 6, No. 19 2021-07-14