Welcome to MGSL!
Turtlebot 4 Launched! Preorder yours now!
D'Kitty: Non Assembled
D'Kitty is platform introduced by project-ROBEL (RObotics BEnchmarks for Learning) for studying and benchmarking locomotion. It is a twelve degree of freedom quadruped capable of agile locomotion. It consists of four identical legs mounted symmetrically on a square base. The feet are simple 3d printed parts with rubber ends. D'Kitty is symmetric along all three axes and can also walk normally when upside down.   The Assembled version of the D'Kitty platform ships completely assembled with the servos ID'ed, it is ready out of the box. The kit version of the D'Kitty contains the Robotis parts necessary for the build, you will have to supply the 3D printed / Laser cut parts. Walking in clutter A Sim2Real policy trained via Natural Policy Gradient on MuJoCo simulation with randomized perturbations learns to walk in clutter and step over objects Walking through obstacles A Sim2Real policy trained via Natural Policy Gradient on MuJoCo simulation with randomized perturbations learns to walk in clutter and step over objects Gravel & branches A Sim2Real policy trained via Natural Policy Gradient on MuJoCo simulation with randomized height field learns to walk outdoors over gravel and branches Slope & grass A Sim2Real policy trained via Natural Policy Gradient on MuJoCo simulation with randomized height field learns to handle moderate slopes     Avoid Moving Obstacles Policy trained via Hierarchical Sim2Real learns to avoid a moving block and reach the target (marked by the controller on the floor) Push to Moving Goal Policy trained via Hierarchical Sim2Real learns to push block towards a moving target (marked by the controller in the hand) Co-ordinate Policy trained via Hierarchical Sim2Real learns to coordinate tow D'Kitties to push a heavy block towards a target (marked by two + signs on the floor)   Includes: DYNAMIXEL XM430-W210-R x 12 DYNAMIXEL U2D2 x 1 6 Port RX/EX Power Hub x 1 HN12-I101 Set x 12 FR12-H103GM Set x 12 FR12-S102k Set x 9 Turbo Lock Nuts & Bolts for assembly Please Note: D'Kitty will require additional 3D printed / laser cut pieces to assemble, for more information please refer to the D'Kitty Parts and Assembly Files

LoCoBot PX100
LoCoBot PX100
The LoCoBot is an ROS research rover for mapping, navigation and manipulation (optional) that enables researchers, educators and students alike to focus on high level code development instead of hardware and building out lower level code. Development on the LoCoBot is simplified with opensource software, full ROS-mapping and navigation packages and modular opensource Python API that allows users to move the platform as well as manipulator in as few as 10 lines of code.   The rover is built on the Yujin Robot Kobuki Base (YMR-K01-W1) and powered by the Intel NUC NUC8i3BEH Core i3 w/ 8gig of ram and 240g HD. An Intel® RealSense™ Depth Camera D435 sits atop an independently controlled pan / tilt mechanism (2XL430-W250-T) at the top of the platform which allows mapping and scanning. The (optional) 360 degree LIDAR can further improve both mapping and scanning for more advanced projects. There are 3 different robot arms that can be added to the LoCoBot platform, offering 4, 5 and 6 degree of freedom options. All of the arms are based off of the X-Series servos from Robotis Co., specifically the XM430-W350-T and XL430-W250- which offer easy hardware setup, available software API’s and other ‘smart servo’ benefits.Interbotix products are aimed at getting users up and developing on ROS as quickly as possible, which is why we've designed our platforms to ROS ready in under an hour. The robot comes pre-assembled with only minimal hardware setup and software is loaded with a few terminal commands allowing for ROS development in under an hour from receiving the LoCoBot.   Software: Fully Configured Navigation Stack simultaneously localize and map out its environment (SLAM), generating a new map in the process, localize itself in a pre-generated map, create a 3D point cloud representation of its surroundings (viewable in Rviz), develop local and global costmaps to be used for obstacle avoidance and path planning in real time (viewable in Rviz), automatically move the base to follow the planned route, move to a desired spot specified by a mouse-click on a virtual map in Rviz Rviz Visualization Environment Get familiar with your LoCoBot platform by visualizing a model of it in Rviz. Using Rviz, you can also see sensor data, monitor your LoCoBot’s progress as it navigates or does perception, and view maps. This can be done on the LoCoBot itself or from a remote computer via ROS networking. Fully Configured Perception Pipeline - With the onboard Intel RealSense D435 camera and robotic arm, your LoCoBot can; find the transform to multiple small objects within the camera’s field of view, pick up a slew of objects in a specified order (based on distance), sort objects based on color / point cloud size / or their respective positions, allow for a researcher to dynamically tune various point cloud filter parameters using an intuitive graphical user interface. MoveIt Support - Abstract away motion planning, inverse kinematics, and collision checking by using MoveIt to manipulate the arm on your LoCoBot. Our custom Moveit package is already fully integrated to be used with ROS Control on both the physical and Gazebo simulated robots. It can also be simulated in Rviz using MoveIt’s Fake Controller Manager plugin. MoveIt Move Group Interface Demos - See how to use MoveIt’s Move Group Python or C++ interfaces with the arm on your LoCoBot. Pose the arm’s end-effector using a custom graphical user interface built on the Move Group’s C++ API. Or launch an interactive tutorial that moves the arm to different poses based on the Move Group’s Python API. Hardware: Kobuki YMR-K01-W1 - The Kobuki platform is a low-cost mobile research base designed for education and research on state of art robotics. With continuous operation in mind, Kobuki provides power supplies for an external computer as well as additional sensors and actuators. Its highly accurate odometry and calibrated gyroscope enables extremely precise navigation. Intel NUC NUC8i3BEH Mini PC - 8th Gen Intel Dual-Core i3, 8GB DDR4 Ram, 240GB Solid State Drive (SSD), Intel Iris Plus Graphics 655, Wifi, Bluetooth 5.0, Gigabit Ethernet, 4k Support, Card Reader, Dual monitor Capable, HDMI, USB, Thunderbolt 3, Ubuntu 20.04. Intel® RealSense™ Depth Camera D435 - is part of the Intel® RealSense™ D400 series of cameras, a lineup that takes Intel’s latest depth sensing hardware and software packages them into easy to integrate products. Perfect for developers, makers, and innovators looking to bring depth sensing to devices, Intel® RealSense™ D400 series cameras offer simple out of the box integration and enable a whole new generation of intelligent vision equipped devices. Pan & Tilt - The DYNAMIXEL 2XL can control two axis with a single module, which serves as the pan & tilt for the camera of the LoCoBot. This servo allows for scanning or object detection whether the platform is in motion or stationary. The 2XL series adopts new features that allow 360 degrees control mode with its contactless magnetic encoder and hollow back case assembly structure and Metal gear (Powder Metallurgy) is used for high durability. PincherX 100 Robot Arm - The PincherX 100 is a 4 degree of freedom manipulator with a maximum reach of 300mm and a working payload of 50g. The PincherX 100 is built using the DYNAMIXEL X Series servos from Robotis which feature high resolution (4096 positions), user definable PID parameters, temperature and positional feedback and much more. Packages: RPLIDAR A2M8 360° Laser Range Scanner (Optional) The RPLIDAR A2 is an indoor, 360 degree 2D LIDAR. Each RPLIDAR A2 can take up to 8000 samples of laser ranging per second thanks to its high rotation speed. The on-board system can perform 2D 360° scans within a range of 12 meters (18m with a bit of firmware adjustment). Additionally, the generated 2D point cloud data can be used in mapping, localization and object/environment modeling. Specfications: LoCoBot Base: WEIGHT - TBD PAYLOAD (KOBUKI)  -  2kg OPERATING TIME  -  90 Mins HEIGHT   -  63cm WIDTH   -  35cm LENGTH  -  35cm PinCherx 100 ARM: DEGREES OF FREEDOM   -  4 REACH  -  300mm SPAN  -  600mm REPEATABILITY  -  5mm WORKING PAYLOAD  -  50g WEIGHT  -  1.7lbs 1.7lbs​​ Drawings: Documents: Documentation Github MESH / STL Files URDF Troubleshooting STEP File Technical Dr