**1. Introduction**

A current significant challenge for industrial production is vision substantiation of a future factory within the concept of Industry 4.0, with the aid of the Internet of Things (IoT) concept. The implementation of this concept into production requires high flexibility and adaptability of production lines and their smaller units. The deployment of this concept in practice is widely supported in Germany, and the first solutions in isolated production systems were developed there [1,2]. Qin et al. [3] note that the Internet of Things is a well-known concept that represents the next generation of products and communication among them. It has a direct correlation with the Industry 4.0 standard, where the existence of smart products is one of the prerequisites for intelligent manufacturing implementation [4].

In a smart factory, individual customer orders determine manufacturing processes and the associated supply chains. This results in the need for high production flexibility with shorter production times, which require the implementation of measures to improve production efficiency, often at a low cost associated with solving these problems. The new term "smart factory" is introduced here to refer to the current trend of automation and data exchange in manufacturing technologies. It includes cyber-physical systems, the Internet of Things, and cloud computing, as declared by Hermann et al. [5], Jasperneite [6], and Kagermann et al. [7]. One of the important indicators for such production

is decentralized decision making, that is, the ability of equipment to make its own decisions, which are the most independent inside closed production complexes. In the case of essential decisions and conflicts, it is clearly necessary to assign tasks to the next level of production control [5]. Closed production complexes require continuity of individual manufacturing operations, synchronization of individual production facilities, and the possibility of rapid adaptation and production time changes of individual production facilities, depending on the control structure of such complexes.

An alternative deployment area of robotic manipulators with a potential requirement for synchronization of their activities is healthcare, particularly that related to the rehabilitation of patients. Techniques dealing with robot-assisted therapy are included, for example, in the works of authors Chang and Kim or Yoon et al. [8,9].

Various methods of synchronization and control of individual production operations and processes are currently used. It is necessary to differentiate whether these are autonomous mobile systems or manipulators and what usage is expected of them. Many published works deal with the synchronization of mobile platform activities. Rubenstein et al. [10] offer a solution for the synchronization of a large group of mobile robots as an open-source, low cost robot, designed to test collective algorithms on hundreds or thousands of robots accessible to robotics researchers. This solution allows for easier testing of algorithms designed to control robot groups because these control algorithms, due to their cost, time, and complexity, are confirmed only through simulations. Popular synchronization methods are also based on the observation of nature, for example, fireflies as presented by Werner-Allen et al. [11]. A modification of this approach is intended to operate on systems that use a communication channel where contention and delays are possible. In addition, the coordination mechanisms that enable the execution of cooperative tasks with modular robotic systems are presented in the contribution of Baca et al. [12]. They describe the implementation of a tight cooperation strategy through Intra M-Robot communication based on a closed-loop discrete time method and the remote clock across the robot configuration enables proper coordination inside the colony.

The work of Chung and Slotine [13] presents a new synchronization tracking control law that can be directly applied to the cooperative control of multi-robot systems and oscillation synchronization in robotic manipulation and locomotion, where a common desired trajectory can be explicitly given.

Rodriguez-Angeles et al. [14] describe a controller utilization that solves the problem of position synchronization of two (or more) robotic systems, under a cooperative scheme, in the case when only position measurements are available. In the work of Yasuda [15], a Petri-net-based prototyping tool is presented to implement the control flow of parallel processes in multiple robot systems. The next variant of synchronization is presented by Markus et al. [16], where the coordination control of two flexible joint robotic manipulators using flat outputs is implemented by means of simulations. The differential flatness technique of trajectory generation enables easy estimation of synchronization parameters and trivializes stabilization of these trajectories around predefined points. Bouteraa et al. [17] describe a new adaptive algorithm, which was proposed for synchronization and trajectory tracking of multiple robot manipulators. The same authors also discuss other techniques in this problem area. They describe the possibility of designing decentralized control laws to cooperatively command a team of general actuated manipulators in the article "Distributed synchronization control to trajectory tracking of multiple robot manipulators" [18], or an approach to position synchronization of multiple robot manipulators based on emergent consensus algorithms [19]. Synchronization of activities in task-oriented robotic rehabilitation training using iterative learning synchronization (ILS) and immediate error correction (IEC) techniques is addressed by Duschau-Wicke et al. [20].

In the above-mentioned cases, the authors based their solutions on complex mathematical algorithms and derivation of complex relations, or by exploring new approaches using specialized hardware. These methods require investment in hardware infrastructure, which is their major disadvantage.

In some cases in homogenous production complexes, it is possible to use tools that are directly implemented in control systems, such as the MultiMove option of ABB robots [21,22] or the RoboTeam software of KUKA [22,23].

However, if limited possibilities exist to upgrade infrastructure and it is desired to use available hardware that may not be from the latest production series, there could be a serious problem with the use of the previously described solutions. We used an alternative approach in a production or technical process of a simple solution implementation even in production facilities without the possibility of using sophisticated methods of robotic set or robotic cell synchronization. Our aim is the simplest solution possible in terms of computing power demands. The basis is to design the least complex algorithms that can be easily implemented in existing controllers in their native programming languages. Therefore, our goal is to develop a solution in which synchronization algorithms can be performed directly on the control units of robotic manipulators. This is based on the utilization of our rich empirical knowledge and experiences in algorithms, in addition to the implementation of various tasks in the field of robotics or modeling and visualization of processes.

The main idea does not lead to a specific use or solution to a precisely specified problem, for example in a production process. The aim of the proposed solution can have a wide range. From the analyzed areas it is possible to use this solution either in production or in healthcare in rehabilitation. Another area of application could be the control of collaboration-oriented workplaces with a master manipulator connected to movements of a human as a cell control element. Finally, it could be used in presentation events oriented to Industry 4.0 or the latest trend, Internet-of-Robotic-Things (IoRT).
