Cps in Production Powered by Diffused Reasoning Capability

The fourth industry revolution is a concept that upset the conventional view of automation and digitization of factories. In fact, the aim of Industry 4.0 is the realization of a Smart-Factory, where the key stone is represented by CPS in production. In the last section, we have seen how this technology is able to reproduce the factory into the virtual space in order to control and optimize processes second the rate of autonomy of the system. Internet of Things is considered the basis for the accumulation and transmission of information and knowledge among all parties inside and outside the boundaries of the factory. Finally with all features of Big-Data, data storage and data processing, CPS in production is able to identify, and forecast optimal way of working of each module or parties of the factory. In this section, we are going to study a smart-factory embedded of CPS in production realized second three parties that we will see during this paragraph. The revolution is that each physical element inside a Smart-Factory can be considered as an alone entity, with thinking ability through embedded computational power that permits it to be autonomous in its operation. Therefore, during the working day it is monitored by widespread controlling operations. Thus, each element in the physical space has autonomous intelligence and goes to the direction of continuous improvement, in order to fast react to possible uncertainties and troubles.

We are going to analyze the architecture of a smart production plant based on a CPS in production, the pillar is a modular system based on different layers that communicate among them. Starting from physical modules composed of equipment, AGVs, and product parties, the system realizes a cyber-display in order to collect information about them in terms of movements, working progresses, tasks, and skills during the manufacturing operations. Possibilities inside cyber space are infinite, but mainly it connects all parties of the physical space and permit them to communicate whatever modules are collocated. This feature permits a comprehensive and right view of the production plant above mentioned, in addition convert it in a flexible system able to be responsive to troubles. In this section, we want to focus on the composition of a smart production plant and the different stratum that compose it. They are mainly three material, logical and interaction stratum. All those layers communicate among them, exchanging huge amount of information and knowledge.

Starting from the first mentioned stratum, it can be considered as the set of elements that conduct the production processes and all movements of resources required in order to accomplish various tasks. Result easy to understand that material stratum is made of different components, first of all equipment and all machines that realize physically product and product parties during the manufacturing process. Second there are automated robots with the aim of manage and realize the movement of material resources when acquired and transported to the first production phase and finally store the finished product into warehouse. Third, AGV technology that will be studied in deep in the next paragraph, that moves material resource and semi-finished products among the entire manufacturing process realizing a link among the components mentioned above. Given an overall overview of the physical layers assembly shop-floor, we can say that operations (manufacturing, transportation, shifting, storage, and distribution) are settled by receiving and sending real time information and data among entities in order to realize a responsive system. Summarizing, performances of physical entities are mainly two, realization of what was designed and acquisition of information. The second stratum consists in a virtual representation of all components mentioned above of the material layer. In fact, in this layer is expected the presence of a supervisor LU for each physical component, LU has different roles but as we can imagine the first role is supervisor with the aim of improving and link different elements of the physical space. In this space, we have to consider logical unit as a single entity that have strong calculation capacity and thus an intrinsic thinking capacity that is exploited through communication and network among each LU. In fact they extrapolate information from MCs and use those data to generate control interfaces. For instance, LU organize the manufacturing planning for each equipment and the transportation planning for all automated vehicles in order to move workpieces among machines. Thus, LU composed of embedded thinking and planning time capabilities improve and organize tasks utilizing information and knowledge exchanged through communication linkage.

The interaction stratum is realized to permit interaction among the other two stratum. In fact it is based on two gates, The first one take and exchange information in material stratum, while the second gate is exploited to link the logical stratum. The technology is based on WIFI connection and commute information in order to make accessible information of one stratum to the other one. For instance take in consideration the case that material stratum updates the logical stratum about new manufacturing conditions, in this case it is a C-L task to communicate the notice and apply translation process. After that Logical unit elaborates information, and realize a new production template and communicates it through C-L to material stratus. Thus, result of primary importance the linking role of C-L.

In the above paragraph, we have explained the architecture of a smart production plant, that consists in physical elements which perform vital activities for production process. After that, there are logical units embedded of computational power, that are dislocated control interfaces that interact among them and regulate the behavior of MCs and AGVs. Features mentioned above are the basis for the development of a Smart-Factory, in unpredictable context. In fact, it needs to be a flexible system, founded on the capabilities of self-regulation and self-adaptation in order to not suffer uncertainties and troubles. The final aim is to realize an intelligent production plant, realized on autonomy and thinking capabilities, which can be summarized inside a production system called NEIMS based on the functioning of a biological system. Inspired by the above mentioned biological system, we are going to propose a production paradigm based on the functioning of human neuro and hormone setting in order to react to troubles and uncertainties of production context. The system based the regulation and the productivity of entities inside the plant through implied commands typical of biological regulation. As above explained even this system is based on diffused control interfaces, embedded of thinking capabilities and computational power. Monitoring physical elements understand disturbances in the context in which operate and self-regulate the system independently. Now we are going to explain how the neuro control and hormone regulation system work inside a production plants based on the architecture explained above. During the regular proceeding of the production process the system uses a normal biological monitor, therefore in case of unexpected and critical situation in order to react to it, the NEIMS utilizes the hormone regulation bringing thinking capabilities to each LUs, which communicate, cooperate, and take a decision to self-react to disturbances. Think about the cycle of client order. All start from an order on the web-site through a smart-phone, this one arrives to the production plan and it has to be elaborated. LU analyze the order and find all the needed functions and manufacturing processes in order to make it. Second the date, that the product must be delivered to the client, LU organize with its embedded data processing the plan template for the production of the product. Taking in consideration the case of an order with no critical delivery date. The system put in list the order, and through ordinary biological control paradigm organize and establish the perfect production schedule. Therefore, can happen that the company was not able to manage the order of the client, and it is translated in an urgent order, because there is few time in order to produce and delivery at the established date. In this case, logical units upset the normal order of work and regulate the new way of working through hormone regulation. In this way, LUs store real-time data and elaborate them, in order to re-organize and re-schedule the production template and be able to face rush-orders or whatever difficulty. In the following figure, it is represented a possible case of unforeseen challenge and difficulty on the functioning of a machine. Starting from this figure, we want to analyze as the NEIMS system face problems keeping a huge level of productivity and efficiency.

From the picture above, result easy to understand that there are problems on the functioning of the machine number three, those data about malfunction are transferred to respective logical unit. In this phase, result essential a re-programming of the normal way of working, in order to complete and restore the affected product in an efficiently manner. Production schedule is re-planned and production process is left to machine one and consequently, movements of workpieces are re-scheduled by the logical unit of AGV number one that transmit the task to the AGV1 that executes the transportation. In conclusion, the product was created by machine one, making possible the production without losing time and respecting the delivery date. Through the examples show before, we can say that systems based on biological control and regulation improve the capabilities of the production plant in terms of reasoning, intelligence, and reactivity. Therefore, NEIMS is not finite in managing rush orders or machines faults, rather it can be used to explore and develop future production solutions able to react promptly to challenges and unexpected disturbances of market context.

In this section, we are going to analyze a practical example of the Cyber-Physical System in production developed before. The aim is to understand if the system can be considered achievable and if it leads to incredible results mentioned above. The architecture of the system remains the same, divided in three layers respectively physical, logical, and communication. Therefore, in order to go in deep with this experiment we are going to consider a series of suppositions and rules that cannot be overcome. Starting from the first:

• Transportation robots move one product or pieces of product for each delivery;
• Automated guided vehicle remains at the same point, after that it concludes the movements of workpieces. Waiting for future orders;
• Equipment inside the production plant executes one task for each time scheduled;
• Product processing needs a lot of tasks in order to be finished. Each task has been done by a different MC at the corresponding scheduled time;
• Before the execution of work, machines do not require any type of regulation reducing at minimum rump-up time.

The system is composed of 2 transportation vehicles and 4 machines dedicated at the manufacturing process, each of this machine executes its own task. After that, we distinguish the following table 4 for ordinary order, while Table 5 for unexpected orders, finally in the last table transportation schedule of automated guided vehicles.

During the practice was introduced an unexpected order in the production plant at time seventies. The system immediately react to the challenge, in fact all logical units started to interact and exchange information through data accessing and data processing in order to reschedule the production process. LUs communicate the new production template to MCs that execute the operation and indicate the new production circumstances to logical units. All this interactions are possible through the communication layer that keep in contact each element inside the factory homogenizing programming language among different layers. The work time template for re-programmed activities are illustrated in the following picture.

Through this experiment, we can easily conclude that Smart Factory can be realized bringing intelligence and thinking capabilities to each element of the production plant. Cyber Physical System in production is the main pillar in order to reach and realize an intelligent smart factory based on the features of flexibility, continuous optimization, and self-adaptation to the dynamic context in which operate. Result of primary importance, the realization of a modular system based on diffused control interfaces that analyze the situation, and are embedded of computational power, thus decision-making capability.