**3. Real-Time Results for A/D/RML Control Based on SHPN Model**

The SHPN model is transposed via the SCADA platform from Siemens into a real-time application, obtained by interfacing the SHPN model with synchronized signals taken from the real process by means of PLC and sensors [28,30,31].

Following implementation, real-time results within the laboratory setup are shown in Figures 25–28, for continuous and discrete places associated with displacements of ARS and FC, for later comparing and validating data with the simulation framework results as presented in Section 2.2.

**Figure 25.** Results for continuous and discrete places associated with displacements of FC with IRM for Assembly WP1.

**Figure 26.** Results for discrete places and transitions on Assembly WP2.

**Figure 27.** Results for continuous and discrete places associated with displacements of ARS and FC with IRM for Disassembly.

The synchronization signals, used in the real-time control application, validate certain transitions into the SHPN model [32]. These transitions are conditioned by the associated signals for releasing recovered workparts on FC trays or on the ML storage units by the ARS. Synchronization will lead to initializing the robot and to monitoring/controlling assembly/disassembly/repair operations with the ARS. Discrete time and sliding-mode control, in trajectory tracking, based on a kinematic and dynamic model, is used to control WMR. In this way, both ARS and the A/D/RML are controlled so as to achieve a minimum assembly and disassembly time cycle.

**Figure 28.** Results for continuous and discrete places associated with displacements of ARS and FC with IRM for Repair.

In order to grab the recovered workparts and place them on the dedicated storage positions, the ARS PeopleBot is equipped with 7-DOF Cyton RM with a gripper paddle and HD video camera on the end effector (Figure 8), connected both via USB with the Remote PC. The gripper is positioned by VSS so as to grab the disassembled component and transport and place it into the dedicated warehouse.

Figure 29 shows the desired and real trajectories of the ARS PeopleBot obtained with the TTSMC in a closed loop control to move from the FC to the storage unit from the mechatronics line and back to the FC in the desired time.

**Figure 29.** Full disassembly process. Desired and real trajectories of ARS PeopleBot based on TTSMC: (**a**) Cylinder 1, (**b**) Cylinder 2, (**c**) Top and (**d**) Body.

In Figure 30 presents X and Y axis trajectories for a complete disassembly process, both desired and the real one so that the differences between them can be easy distinguished.

Figure 31 illustrates X and Y axis tracking errors in absolute coordinates for the disassembly process as well, where ADRML is served/assisted by ARS for robotic pickand-place operations for the recovered/dismantled components and transporting them back to the storage units.

**Figure 30.** Full disassembly process. Desired and real trajectories of ARS PeopleBot based on TTSMC: (**a**) X axis, (**b**) Y axis.

**Figure 31.** Full disassembly process. Trajectory tracking errors in absolute coordinates.

#### **4. Discussion**

The paper proposes an extension both in hardware as well in software, which allows the implementation of a flexible and multifunctional technology able to manufacture different products (Figure 9) and to disassemble (Figure 10), recover components or to repair products that do not correspond to the desired quality (Figure 11). All these functionalities are made with high precision due to the integration of an industrial robotic manipulator (ABB 120 IRM), an autonomous robotic system equipped with a mobile visual servoing system and by using a multi-agent control strategy and communication structure between

the flexible cell and the mechatronics line that allows synchronizations of the requested operations. Therefore, the master PLC synchronizes with subsystems PLCs to automate their respective areas and for operating and controlling their local IO devices, after confirmation from the main control unit is applied (Figure 24).

Modeling of the system using hybrid Petri Nets, in which A/D/RML is a hybrid SHPN model having the Hera&Horstmann mechatronics line with discrete states and transitions and the ARS subsystem with continuous dynamics, is presented in the paper and represents only an intermediate stage (Figure 12). Due to the dynamic nature of the system, analytical methods can be used, but are limited; therefore, task scheduling and the simulation of the model, which tackle the compatibility between the two subsystems, is used for studying the evolution of the discrete states of A/D/RML with the physical constraints and continuous states of the ARS.

The real-time research and implementations is followed, comparing and validating data with the simulation framework results. SCADA environment is developed (Figure 20) so that the entire system works autonomously, fully automated to meet the actual industry requirements. The actual results also showed that the actual manufacturing line implementation satisfied the design target. By using the smart and autonomous technologies to operate in a seamless and secured way, this meets also the new requested requirements standards of Industry 4.0, increasing the degree of integration and compatibility with the actual industry needs

The control of the robotic arm Cyton 1500, for handling and precise positioning operations, when gripping or releasing the part, is based on the inverse kinematic model and is robust, having the desired behavior even in presence of uncertainties and external disturbances. Cyton 1500 manipulator is equipped with an anti-collision map tested on the simulation stage; the robotic arm moves until the object detected is in the center, and if the time needed to get to the object is higher than expected or a supplementary torque is detected, for example an obstacle has been placed in the trajectory, the robotic arm returns to the home position and notifies the user that the trajectory path following has been unsuccessful. The implementation of multifunctional flexible manufacturing technology in a laboratory system, to be as close as possible to the real industrial world, draws some limitations as well; we could not gather consistent data regarding the performance of robotic arm in the presence of noise for applying methods to overcome that. A more sophisticated approach has not been implemented, as the deviations that appear are just on the X or Y axis, due to the complex autonomous robot transportation errors.

#### **5. Conclusions**

The presented research is still in progress; it is a place for further improvements and fine tunings, and the important benefit and contribution of this research is the implementation of manufacturing technology assisted by autonomous robotic systems at the laboratory level, which works in real-time and which, if used industrially in the real world, would increase efficiency, reliability and precision. This research aimed for a dual purpose, one educational and another to implement, test and adapt this technology to be as close as possible to the real industry world requirements.

The educational goal aims to familiarize the system designer with everything that defines new industry architecture, including Industry 4.0 concepts, and to try to improve the actual technology design with the integration of all new, state of the art aspects of production and engineering, including smart manufacturing products and intelligent material handling systems and technologies.

Regarding the correspondence with the real industrial world, most manufacturing industrial technologies are served by robotic systems that have a fixed position (robotic manipulators). Through this study, we extended the degree of automation and efficiency of these production lines by using new technologies such as autonomous robotic systems equipped with manipulators and visual servoing systems. The goal is to adapt this technology to meet as much as possible the actual industry needs to confirm the feasibility of

the line and to keep up the rhythm with the technology development. Therefore, the final purpose is to develop a fully automated multifunctional flexible manufacturing technology without the intervention of the human operator for a predefined production volume with the recovery of components of bad assembled products that did not pass the quality tests and integrating new emerging technologies such as SCADA, IIOT and MQTT protocols for Cloud interface.

Although this is a technology that has been used at the level of a laboratory, it can be extended further to real industry, where high accuracy and positioning are needed. Multispectral video sensors, providing new imaging capabilities without adding size or weight, can be used in order to reduce errors in reflectance estimation for remote sensing on production line inspection or workparts validation and quality checking to more strongly demonstrate the reliability and to increase speed and efficiency by integrating with the ARS of the presented manufacturing line technology, especially for recovery and accurate positioning operations.

The implementation of robust control architectures to uncertainties will be further considered for all systems: ARS, FC and the mechatronics line. As a result, this increases the reliability, flexibility and robustness of the technology to the uncertainties that might come from the sensors from the ARS and VSS.

The presented control architecture is a hybrid structure, multi agent-based control. Unlike using this control strategy, the system can be enhanced with artificial intelligence (AI), which is a combination of situational awareness and creative problem solving, to identify and fix potential assembly problems much faster and can diagnose and prevent further issues by directly alerting through SCADA systems when anomalous units are identified.

Additionally, we will also focus on the time study system performance evaluation and optimization methods of the complete production process to improve the performance and support better product quality [33]. Efficiency requirements are one of the key factor nowadays; therefore, optimization in what concern costs, energy and time will be one of the further purposes of development for manufacturing lines using ARS equipped with robot manipulators and visual servoing systems.

**Author Contributions:** Conceptualization, G.S., A.F. (Adrian Filipescu), D.I., R.S, ., D.C., E.M. and A.F. (Adriana Filipescu); methodology G.S., A.F. (Adrian Filipescu), D.I., D.C., R.S, . and E.M.; software, G.S. and D.I.; validation, A.F. (Adrian Filipescu), R.S, ., D.C. and E.M.; formal analysis, G.S., D.I. and A.F. (Adriana Filipescu); writing—original draft preparation, G.S. and A.F. (Adrian Filipescu), D.I.; writing—review and editing, A.F. (Adrian Filipescu), R.S, . and D.C.; supervision, A.F. (Adrian Filipescu); project administration, A.F. (Adrian Filipescu); funding acquisition, A.F. (Adrian Filipescu), G.S. and D.I. All authors have read and agreed to the published version of the manuscript.

**Funding:** This article (APC) will be supported by Doctoral School of Fundamental Sciences and Engineering, "Dunărea de Jos" University of Galat,i.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data availability is not applicable to this article as the study did not report any data.

**Acknowledgments:** The results of this work will be also presented to the 10th edition of the Scientific Conference organized by the Doctoral Schools of "Dunărea de Jos" University of Galat,i (SCDS-UDJG) http://www.cssd-udjg.ugal.ro/ (accessed on 10 May 2022), that will be held on 9th and 10th of June 2022, in Galat,i, Romania.

**Conflicts of Interest:** The authors declare no conflict of interest.
