FRANKA RESEARCH 3

FRANKA RESEARCH 3

Franka Research 3 is a research robotics system that references world-class, sensitive forces and provides researchers with easy-to-use robotics features and low-level access to robot control and learning capabilities.

Category:

Description

characterization

  • Flexibility: 7 degrees of freedom for flexible adaptation to various work environments and tasks
  • High precision: Highly accurate motion control and positioning capabilities for precise work operations
  • Safe and reliable: Built-in safety sensors and anti-collision system to ensure the safety of human-machine collaboration
  • Easy to program: uses simple programming language and interface for rapid development and debugging
  • Highly scalable: supports the installation of a wide range of accessories and tools that can be applied to automation tasks in different fields

Relevant information

Product Brochure
data sheet

Preferred platform for cutting-edge AI and robotics research

Frank Research 3 is a research robotic system that cites world-class, sensitive force.

Provide researchers with easy-to-use robot functionality and low-level access to the robot's control and learning capabilities.

Artificial Intelligence and Robotics Toolset

robotic system

The Franka Research3 robotic system includes the Armand control system. Equipped with torque sensors at each joint, this sensitive and flexible robotic arm features industrial-grade attitude repeatability of +/-0.1mm and virtually no path deviation, even at high speeds. It has a 3kg load capacity, 855mm working radius and 94.5% workspace coverage.

FCI (Franka Control Interface)

The FCI list is the ideal interface for exploring low-level programming and control scenarios, providing the current state of the robot and enabling it to be commanded at 1kHz. On top of the C++ interface library franka, integration with the most popular ecosystems ROS, ROS2 and MATLAB & Simulink is available!

DESK

Desk is a browser-based user interface that provides quick robot control options and the ability to drag and drop applications into sequences to quickly create entire tasks. Ideal for rapid prototyping of robot behavior, setting up experiments, and conducting simple human-robot interaction studies and demonstrations.

Watchman

Easy to use and quick to implement security. Thanks to the browser-based user interface Watchman, the often complex security setup is greatly simplified to ensure that your lab and lab personnel are protected.

The right interface for each use case

The three access levels address different needs and skills and are applicable to the entire field of robotics studies.

DESK

Simplicity of use and minimal programming time make Desk the most suitable interface for rapid prototyping, simple human-computer interaction studies and presentations.

RIDE

It allows researchers to fully integrate the Franka Emika system into experimental setups and to utilize its integrated high-performance controller. It is also an excellent teaching tool for elementary robotics.

FCI

FCI bypasses robot control and allows researchers to run their own control algorithms at 1kHz on an external real-time viable personal computer. This is an ideal interface for exploring low-level planning and control schemes.

Technical Parameters

robot arm
(number of) degrees of freedom (physics)7connector

-Ethernet (TCP/IP) for visual intuitive programming

-Externally enabled devices with desktop security input
-2 configurable safety inputs for emergency stops, protection devices or other guards (connected via external OSSD converter)

-Hardware preparation: 2 DI and 2 DO (24 V, isolated, conforming to EN 61131-2 type 3 characteristics, 100 Hz sampling rate)

-Control Connector

-End effector connector

load3kg
Maximum extension distance855mm
Force/Torque SensingChain-side torque sensors on all seven axes
Joint Position LimitationsA1, A3: -166/166 degrees A2: -105/105 degrees A4: -176/-7 degrees A5: -165/165 degrees A6: 25/265 degrees A7: -175/175 degrees
Mounting FlangeDIN ISO9409-1-A50
mounting positionriserThe user interface is located on the arm driver's grip-Guided enable switch with integrated safety ratings
-Guide button
-Oriented mode selector
weights~17.8kg
protection classIP40
environmental temperature+5°C to ~+45°C
air humidity20-80% non-condensingUser interface on Arm's driver's disk-Status indicator
-Flight Mode Selector
-Arrow keys, Learn, Confirm, Delete
containmentperformances
Controller size (19 inches)355×483×89mm (L×W×H)campaigns
Supply Voltage100-240V ACJoint speed limitationa1-a4:150°/s a5-a7:301°/s
main frequency50-60Hz 
power wastage~80WCartesian speed limitUp to 2 m/s transfer speed
Active Power Factor Correction (PEC)beStance Repeatability<+/-0.1mm (ISO9283)
weights~7kg 
protection classIP20each other
environmental temperature+5°C to ~+45°Cchannelling power~2.5N
air humidity20-80% non-condensingAdjustable translational stiffness10-3000N/m
Permissible mounting directionlevel (of achievement etc)Adjustable rotational stiffness1-300Nm/rad
connector-Ethernet (TCP/IP) for Internet/workshop connections -Power connectors IEC60320C14 (V-Lock) -Arm connectorsmonitoring signalJoint positions, velocities, moments, Cartesian positions and forces.

From machine learning, robot control and motion planning, to manipulation and reinforcement learning.

For researchers at the forefront of artificial intelligence and robotics, FRANK RESEARCH3 provides a reference force-sensitive robotics platform and powerful control interfaces for rapid results and publication. The platform also provides a low-barrier entry for researchers seeking to automate experimental setups and supports teaching courses on robot control and automation.

Inducing Uncertainty Awareness to Policy Optimization

Action Reasoning for Goal-Based Imitation Learning

RL Bench: Robot Learning Benchmark

Constrained probabilistic motion principles for robot trajectory adaptation

Reinforcement learning for robotic rock grasping learning in an off-earth space environment

Ability to learn generalized Coupled terminology for obstacle avoidance via low-dimensional geometric descriptors

6 Degrees of Freedom for Target-Driven Object Manipulation in Cluttered Environments

Demonstrable safe and efficient motion planning with uncertain human dynamics

A novel adaptive controller for robot manipulators using active reasoning.

A teleoperator interface for MOCA to control robot motion and manipulation.

Planning Maximum-Maneuverable Cutting Paths - R RT*-R MM

Online planning for partially observable task and action problems in belief space

Goal-oriented task and motion planning in dynamic environments

Build to Learn: Learning How to Build Robotics Manipulation Skills Development

Learning to generate reachability-aware 6-degree-of-freedom grasping poses.

Learning to pre-grip Manipulating objects from ungraspable poses

Interaction Force Calculation Estimating Sensorless Robot Applications Using Environmental Stiffness

Feedback-based folding of fabric strips

Physical reasoning for describing physics: force-based sequential operation planning

Deep visual reasoning: learning to predict action sequences for task and action planning from images

Competency-aware role assignment methods for industrial assembly tasks

A Framework for HCI User Research

Search-Based Task Planning and Learning Skill Effectiveness Modeling for Lifelong Robot Operations

Shared Autonomous Reconfigurable Control Framework for Remote Operation of Multi-Arm Systems

Related products