Research Robot Servosila "Engineer"

The Servosila Engineer robot was originally designed for disaster response, search and rescue missions, and public safety applications, but has since found a new role as a robotics research platform.

The robot features a powerful onboard Linux computer capable of massively parallel computations, a rich set of sensors, and many degrees of freedom that make it ideally suitable for cutting edge research in robotics and artificial intelligence.

Due to its heritage, the robot has a very rugged design capable of withstanding a service life as a disaster response robot, but the robot still very lightweight, and can be transported by a single man in a backpack. The chassis of the robot is capable of negotiating obstacles found in urban environments. The robot climbs stairs, is capable of penetrating buildings.

It is water tight and operates in the rain and in the snow. A powerful onboard computer install in the robots head is passively cooled. These features make the robot well suited for programmers working on outdoor robotics projects.

The robot features between five to seven degrees of freedom, depending on a chosen configuration. Its robotic arm is capable of lifting heavy loads, and reaching substantial heights.

The robotic arm enables conducting research in topics such as motion planning, inverse kinematics, self-collision avoidance, grasping, usage of tools, learning from experience, learning from interaction and many others.

Autonomous door opening is an important research subject, and the robot is a suitable platform, capable of reaching door handles and various heights.

The robot is equipped with a laser scanner device, an optical zoom camera, a pair of stereo vision cameras, an inertial measurement unit, a rear view camera, and a GPS/GLONASS receiver.

The powerful suite of sensors can be used for research in simultaneous localization and mapping, or SLAM. The GPS/GLONASS receiver enables extending the SLAM algorithms into outdoor environments.

Machine vision is enabled by several cameras installed in the head of the robot. The cameras and a powerful computer enable research in machine vision topics, such as object recognition, visual object manipulation, 3D scene reconstruction and visual SLAM.

The robot is controlled via a portable operator control unit, or OCU, equipped with virtual reality goggles and a joystick. The OCU is a battery operated computer, running Linux, with a radio modem built-in. Such an OCU design enables research in efficient human-to-machine interfaces for both indoor and outdoor environments.

The robots chassis and the head both have hardpoints and sockets for connecting external payload modules, such as software defined radios, ground penetrating radars, infrared cameras, drills, and so on. The sockets connect such external payload modules to an onboard power supply, onboard Ethernet bus, and an onboard CAN bus.

The chassis of the robot is a research platform in its own right, and is equipped with a built-in rotating servo mechanism, that can be used to mount rotating turrets, directional antennas or custom robotic arms.

Likewise, the manipulator arm and the robotic head can be used as a standalone system, not connected to the mobile base. Such a configuration is useful in research in robotic social interaction, or in table-top manipulation.

The robot has a modular construction, and is a very a reliable piece of equipment as it was subjected to the rigorous testing during its harsh life in the disaster response world.

Both the onboard computer and the OCU computer run either Linux or Windows, and are capable of executing ROS, MRPT or any other Linux or Windows compatible robotics software.