contact    site map  
    polski   english

Mobile battle robot

The platform was designed in the Institute of Applied Computer Science, Lodz University of Technology, by a research team: Marcin Bąkała, Sylwester Błaszczyk, Wojciech Dadan, Piotr Duch, Rafał Jachowicz, Roman Krzeszewski, Maciej Łaski, Piotr Ostalczyk and Adam Wulkiewicz. The team, headed by Prof. Dominik Sankowski, developed the current version of the platform in cooperation with the companies: Sochor and GreenPoint. The platform features several innovative solutions and it has attracted a considerable interest during one of the world's major fairs of military equipment - 21st International Defense Industry Exhibition which that took place in Kielce, Poland, in September 2013.

The robot, depicted in Fig. 1, is composed of six-wheeled body comprising electrical locomotion units, sensors, on-board computer and communication modules, and the manipulator arm, mounted on the body. The manipulator has 4 degrees of freedom, and in the current version of the platform, it hosts the multi-camera observation head.

Fig.1. Mobile platform ROBOKIS-II

Robust design of the locomotion unit combined with advanced control strategies enable reliable platform operation in natural terrain. Modular architecture of the robot allows for its quick reconfiguration and extension, so that it can be easily adapted to suit various user's needs. The observation arm can also be easily replaced with a manipulator or appended with additional actuators.

Both arm-motion control and platform-motion control are based on entirely new mathematical concepts. Platform is driven by six independent electric motors that operate as a single motor unit in a closed-loop control mode. To enable handling of such a complex system, algorithms based on fractional-order differential calculus were developed. The fractional-order derivatives and integrals can be viewed as an extension to classical PD controllers. The control strategy applied is adaptive and, unlike in classical PD controllers, control variables to be tuned are fractional orders of differentiation and integration. Since the adopted, cutting-edge control strategy is still open and not fully explored, optimization of control algorithms is subject to on-going research.

The platform is equipped with various sensors, including visual, infrared and near-infrared cameras, laser, ultrasound sensors, accelerometers and GPS. Signal acquisition and signal processing units are modular. The central processing unit is based on multi-core microprocessor systems and DSPs, enabling tasks from low-level computations through communication to execution of complex data analysis algorithms. The powerful central unit provides a basis for implementing autonomous, 'intelligent' functionalities. The platform is capable of path-planning, getting to an operator-defined target, avoiding collisions with still and moving obstacles. It can provide remote operator with fused information from different types of cameras, it stores and interprets the path followed, in order to share crucial information with other similar units. Finally, it is capable of visual object detection, recognition and tracking, and it can cooperate with other units, including aerial devices, such as quadcopters. The platform can also operate in the remote control mode, offering simple and intuitive manipulation of both its motion and its arm movements.

return to the top

Piotr Ostalczyk
Piotr Duch

Last modification:
2014-03-10 11:00:18, Maciej Łaski