Research
the new
FRANKA RESEARCH 3
The platform of choice for cutting edge AI & Robotics research
Franka Research 3 is the reference world-class, force sensitive robot system that empowers researchers with easy-to-use robot features as well as with low-level access to robot´s control and learning capabilities.
You don't have the new FR3 yet?
Get a free technical presentation with our experts!
The AI & Robotics tool set
Find out what´s included in the offer.
Robot System
Franka Research 3´s robot system includes the Arm and its Control. The force sensitive and agile Arm features 7 DOF with torque sensors at each joint, industrial-grade pose repeatability of +/- 0.1 mm and negligible path deviation even at high velocities. It comes with a payload of 3 kg, a reach of 855 mm and a workspace coverage of 94.5 %.
Download DatasheetFCI (Franka Control Interface)
FCI is the ideal interface to explore low-level programming and control schemes, providing the current status of the robot and enabling its direct torque control, at 1 kHz. On top of the C++ interface libfranka, integration with the most popular ecosystems ROS, ROS2 and MATLAB & Simulink is available!
DocumentationDESK
Desk is the browser-based user interface that offers quick robot control options, and where Apps can be dragged and dropped into a sequence to create entire tasks in no time. Ideal for rapid prototyping of robot behaviors, setup of experiments, simple human-robot interaction studies and demos.
Watchman
Easy to use and fast to implement safety. Thanks to browser-based user interface Watchman, typically complex safety setups are greatly simplified to ensure that your lab and lab workers are protected.
Excited to take the next step?
Get in touch to request a quote.
„The Franka Emika robot system is ideal for our research in reactive, robust manipulation in open-ended environments. It is also well suited for our work toward collaborative manipulation alongside people.“
Dieter Fox, Senior Director - Robotics Research, NVIDIA
„Humans and robots are differently intelligent; by interacting, we can combine skills to do more than what we could achieve separately.“
Leila Takayama, Associate Professor - Human-Robot Interaction, University of California
„A new era of robotic applications calls for a new robot generation. Franka Emika is laying down a solid foundation for the new challenges in real-world robotics applications.“
Oussama Khatib, Director - Robotics Lab, Stanford University
Resources and Community
An open and global research ecosystem enabled by a powerful robotics platform for quicker time to results and publishing. Franka Research 3 is the reference platform to integrate existing research, share breakthroughs and collaborate on projects, replicate studies and promote papers with the community.
Customize your Franka Research 3
Further enhance the platform with end effectors and additional software.
Franka Hand
The Hand is Franka Emika´s 2-finger gripper with exchangeable fingertips, fully integrated with the software of Franka Research 3, therefore plug-and-use. The fingertips can easily be changed and adapted to the objects to be grasped, e.g. by using 3D-printed fingertips.
App Package for FR3
Apps are modular building blocks that can be combined into App Workflows to prototype robot behaviors rapidly. Each App contains a context menu where the user is guided interactively to enter parameters like speed and force, as well as to set robot poses by demonstration.
RIDE
RIDE is the development interface for writing custom Apps and connecting third-party hardware and external resources. It´s the ideal tool for customizing and extending the system’s capabilities.
Interested in the offer?
Contact us to find out more.
The right interface for each use case
Three access levels to the robot address different needs and skills, for the whole spectrum of robotics research.
DESK
The ease of use and minimal programming time makes Desk the most suitable interface for rapid prototyping, simple human robot interaction studies and demos.
RIDE
It enables researchers to fully integrate the Franka Emika system into experimental setups, and exploit its integrated high performance controllers. It is also a great tool for teaching introductory robotics.
FCI
FCI bypasses the robot’s Control to let researchers run their own control algorithms in external real-time capable PCs at 1 kHz. It is the ideal interface to explore low-level planning and control schemes.
| DESK | RIDE | FCI | |
|---|---|---|---|
| Workflow-based programming using Apps | • | ||
| Quick prototyping of experiments and demos | • | ||
| Execution of robot tasks | • | • | |
| Development of customized Apps | • | ||
| Connection of third-party hardware | • | • | |
| Connection of external resources | • | • | |
| 1 kHz torque, position and velocity control | • | ||
| 1 kHz measurement of sensor signals and robot status | • | ||
| Access to kinematic and dynamic robot model | • | ||
| Integration with ROS and MATLAB & Simulink | • |
The reference robotic platform for research, globally
Start collaborations across a multitude of fields and compare results with a continuously growing network of Franka Emika academic users.
Check the academic community`s publications
From AI, ML, Robot Control and Motion Planning, to Manipulation and HRI.
For researchers at the cutting edge of AI & Robotics, FRANKA RESEARCH 3 provides a reference force-sensitive robotic platform and powerful control interfaces, for quick time to results and publishing. The platform also offers a low barrier to entry for researchers in search of a robot arm to automate their experimental setup, as well as a support for teaching robot control and automation courses.
Guided Uncertainty Aware Policy Optimization
Motion Reasoning for Goal-Based Imitation Learning
RLBench: The Robot Learning Benchmark
Constrained Probabilistic Movement Primitives for Robot Trajectory Adaptation
Reinforcement Learning for Robotic Rock Grasp Learning in Off-Earth Space Environments
Learning Generalizable Coupling Terms for Obstacle Avoidance via Low-Dimensional Geometric Descriptors
6-DOF Grasping for Target-driven Object Manipulation in Clutter
Provably Safe and Efficient Motion Planning with Uncertain Human Dynamics
A novel adaptive controller for robot manipulators using active inference
A Teleoperation Interface for Loco-manipulation Control of MOCA
Planning Maximum-Manipulability Cutting Paths - RRT*-RMM
Online Replanning in Belief Space for Partially Observable Task and Motion Problems
Object-Centric Task and Motion Planning in Dynamic Environments
Scaffold Learning: Learning to Scaffold the Development of Robotic Manipulation Skills
Learning to Generate 6-DoF Grasp Poses with Reachability Awareness
Learning Pregrasp Manipulation of Objects from Ungraspable Poses
Interaction Force Computation Exploiting Environment Stiffness Estimation for Sensorless Robot Applications
Feedback-based Fabric Strip Folding
Describing Physics For Physical Reasoning: Force-based Sequential Manipulation Planning
Deep Visual Reasoning: Learning to Predict Action Sequences for Task and Motion Planning from Images
A Capability-Aware Role Allocation Approach to Industrial Assembly Tasks
A Framework for Human-Robot Interaction User Studies
Search-Based Task Planning with Learned Skill Effect Models for Lifelong Robotic Manipulation
A Shared Autonomy Reconfigurable Control Framework for Telemanipulation of Multi-arm Systems