Vision 60 Robots
The HDSRL lab has two very advanced 3D quadruped robots, referred to as Vision 60, designed and manufactured by Ghost Robotics. Vision 60 is a mid-sized tele-op and autonomous all-terrain ground drone. It weighs approximately 32 kg. Vision 60 has 18 DOFs, of which 12 leg DOFs are actuated. The robot can transverse a range of unstructured terrains and substrates.
A1 Robots
The HDSRL lab includes three A1 quadrupedal robots. The A1 robot is made by Unitree for agile locomotion and has a maximum continuous outdoor running speed of 3.3 (m/s). It has foot pressure sensors, intelligent depth cameras, LiDAR, and NVIDIA TX2 for Visual SLAM.
Go2 Robots
The HDSRL lab also has two advanced Go2 robots (model Go2 Edu-U3) made by Unitree. Similar to the A1 robots, the Go2s have 18 DOFs with IMUs, joint and foot sensors, and torque control capabilities at 1 kHz. Additionally, the robots are equipped with 3D LIDAR Mid-360 and RealSense D435i depth cameras for navigation algorithms. They feature 100 TOPS of computing power with AI algorithms and can communicate with external computers.
Minitaur
The HDSRL lab also includes Minitaur, designed and manufactured by Ghost Robotics. Minitaur is a 6.5 kg 2-DOF leg 8-motor (2 per hip) quadrupedal robot platform. The high-torque brushless motors, high-resolution encoders, and specialized leg design let Minitaur run and jump over difficult terrain. 2DOF legs are good for more than just running; they can sense and exert precise forces as well as manipulate. The robot is capable of running at 2 m/s and turning at 1 rad/s.
Kinova Arms and Motion Capture
The HDSRL lab includes two Kinova robotic arms. The Kinova arm is a 4.8kg and 4-DOF arm mounted with a 3-finger end-effector and a payload capacity of 4.2kg. These features make the Kinova arms suitable for manipulating a variety of objects. In addition, it is capable of running low-level controllers at 500Hz and high-level controllers at 100Hz. The Kinova arm is capable of RS-485 communication protocol, and its SDK is compatible with Ubuntu OS. The HDSRL lab also has an OptiTrack motion capture system consisting of twelve Prime 13 cameras (240 FPS at 1.3 MP resolution).
Jackal UGV
The HDSRL lab includes Jackal, an unmanned ground vehicle (UGV), made by ClearPath Robotics. The robot includes an onboard computer, GPS, and IMU and the ability to add a variety of custom sensing options. Jackal’s comprehensive ROS API provides end-to-end visibility of its operation.
Franka Emika Panda Arm
The HDSRL lab includes one Franka Emika Panda collaborative robotic arm. It is a 7-DOF arm with torque sensors at each joint, industrial-grade pose repeatability of +/- 0.1 mm, and negligible path deviation even at high velocities. It is mounted with a Franka Hand gripper and has a maximum payload of 3 kg. It has a reach of 855 mm and a workspace coverage of 94.5$\%$. It provides a browser-based interface offering quick robot control options using drag and drop sequences, which is ideal for rapid prototyping of robot behaviors. It has a C++ interface libfranka enabling direct torque control at 1 kHz. Additionally, it also offers integration with popular ecosystems such as ROS, ROS2, and MATLAB $\&$ Simulink. This robotic arm finds applications in various robotics tasks ranging from benchmarking of control and machine learning algorithms, assisting humans in collaborative tasks, physical haptic feedback devices for virtual reality-based teleoperation and industrial assembly tasks.
HaptX Gloves and VR-HTC Vive Pro Eye
The HDSRL lab has HaptX DK2 gloves with tactile and force feedback. It achieves true-contact haptics through patented microfluidic technology. It features over 130 discrete points of tactile feedback that can cause over 2 mm of physical displacement of the user's skin. It also features force feedback with exotendons applying up to 40 pounds of dynamic force per hand (8 lbs./35 N per finger). It has a proprietary magnetic system that captures 30 degrees of freedom per hand with sub-millimeter precision. It comes with a Windows-based SDK that provides low-level C++-based control of the actuators. Further, it has Unity and Unreal game engine plugins that can be integrated directly with a virtual game environment. It works in combination with the HTC Vive ecosystem and SteamVR. These gloves find applications in training, gaming, and teleoperation. the lab also has access to a HTC VIVE Pro Eye, which includes a VR headset, two base station cameras, and two hand controllers.
