Sažetak
Mobile robots take a great part in the robotics research space. Today researches are carried out in the area of robot design, robot control, interaction between a robot and its environment, path planning, sensors, etc. The least researched problem in the area of mobile robots is related to robot programming languages. Most programming languages for mobile robots are oriented to robot movement, not to the robot goal. It means that the programmer explicitly defines the robot kinematics (position, velocity, orientation) which works fine in a well-defined environment, but it is limited in flexibility and is not user-friendly. This paper describes the research that has a goal to define a minimal set of robot commands that would enable flexible control of the robot in a poorly defined environment. Robot commands are task- oriented rather than kinematics-oriented. To enable the goal of task orientation, a vision system mounted on the robot is used besides internal robot sensors (encoders, range finders and timers). Vision system is a separate application that recognizes objects by their color and sends information to the program for the robot control. All algorithms for the object recognition and robot control are tested using the eMIR mobile robot that works within an appropriate environment.
Ključne riječi
mobile robot programming language robot vision color tracking