This paper describes one of the possible ways of implementing a cost-effective bellow type pneumatic muscle in a complex system like Stewart's platform. Two prototypes of the platform 3D model were created and the construction of both was described in detail. In order to control the platform, additional research was conducted, and algorithms were created for control systems. Algorithms were tested for the closed-loop control of the movement of bellow type pneumatic muscle. Detailed laboratory tests of the response of two types of pneumatic systems, with indirect and direct vacuum control, were carried out when controlling the movement of the pneumatic muscle. The results of the response to the step function of PID and LQR controllers are graphically presented and their characteristics such as steady-state error and rise time are analysed in detail. To ensure spatial position feedback of the platform, several sensors such as accelerometers and gyroscopes were tested, but without success. Due to the high noise in the readings from the sensors, it was not possible to carry out the necessary numerical integration processes to get the distance or the rotation angles. Few attempts were made by implementing several simpler ways of filtering the signal, but they did not give satisfying results either. Algorithms for open-loop controlling the motion of the platform were developed and tests procedures were carried out. Camera readings were used for the open-loop control process. The results of maximum displacements in a particular direction and the validation of the open-loop control algorithm are presented graphically and tabularly. The obtained results were analysed and the characteristics of Stewart's platform with pneumatic muscles were described.