Agent-based automation

Abstract

What is proposed is a plurality of field components (1, 2, 3, 4, 5, 6) of a process oorr a manufacturing plant, which field components (1, 2, 3, 4, 5, 6) comprise respectively a computer- implemented agent (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12) which is designed for: checking the operational requirements of the respective field component (1, 2, 3, 4, 5, 6), especially limit violations or deviations of setpoint, generating aa request aabboouutt tthhee present needs of the respective field component (1, 2, 3, 4, 5, 6), sending the request about tthhee present needs of the respective field component (1, 22,, 3, 4, 5, 6) to other field components (1, 2, 3, 4, 5, 6), using a communication connection (1, 2, 3, 4, 5, 6), forwarding requests received from a further field component (1, 2, 3, 4, 5, 6) to a still further field component (1, 2, 3, 4, 5, 6) or to a plurality of field components (1, 2, 3, 4, 5, 6), performing reactions to requests the computer-implemented agent has received from other field components (1, 2, 3, 4, 5, 6), if the field component (1, 2, 3, 4, 5, 6) has the capability to do so, and confirming a fulfillment of requests which the computer- implemented agent has received from other field components (1, 2, 3, 4, 5, 6) if the field component (1, 2, 3, 4, 5, 6) has the capability to do so, wherein the field components (1, 2, 3, 4, 5, 6) further comprise respectively a computer-implemented behavior agent (A1 A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12) which is specific for a type of the respective field component (1, 2, 3, 4, 5, 6).

Description

[0001] Description

[0002] Agent-based automation

[0003] The present disclosure is directed to a plurality of field components of a process or a manufacturing plant according to claim 1. Furthermore, the present disclosure is directed to a method of automating at least a part of a process or a manufacturing plant according to claim 12. Still further, the present disclosure is directed to an engineering station server according to claim 13. Still further, the present disclosure is directed to a method of generating an automation for at least a part of a process or a manufacturing plant according to claim 16. Still further, the present disclosure is directed to a control system for a process or manufacturing plant according to claim 17.

[0004] Process plant automation may typically involve recording measurement signals from the physical process, evaluating automation logics (closed- and open-loop controls, including sequences and interlocks) , and sending the corresponding commands to the process plant's actuators. The automation software may be run on a centralized automation server.

[0005] In traditional process plant automation, the automation software may be engineered manually by correspondingly experienced persons or automatically by using appropriate engineering tools. In case of manual engineering, significant engineering efforts may be required. Even in case of automated engineering, certain tasks, e.g., connecting sensors and actuators to a control loop, may still need to be carried out manually. Furthermore, if the topology of the physical plant is changed, e.g., by adding or removing certain sensors or actuators, the engineering of the automation software must be changed accordingly. The same may apply if the requirements for automatic plant operation like recipes are modified. Additionally, the current engineering may not be valid anymore if certain sensors or actuators drop out. Thus, significant engineering efforts may be required in the beginning, as well as for any change in plant configuration later. Additionally, plant flexibility and availability may be limited.

[0006] WO 2022 / 207107 Al there is disclosed a general concept for an agent-based automation. This basic concept of agent-based process automation uses the swarm intelligence of many different agents to automatically generate a basic automation structure. In a process plant many different components are included e.g., valves, pumps, different sensor types. Each of these components are represented by an agent. In the previous concept each component was assigned to a special agent e.g., the pump agent represented the pump and the pressure sensor agent the pressure sensor. Therefore, many different agents are needed to realize the agent-based automation of WO 2022 / 207107 Al. In addition, this concept is rather complex and an extension to all existing field components (e.g. tanks, heaters) is very difficult. Additionally, the concept of WO 2022 / 207107 Al is not very flexible to changing conditions e.g., changing flow direction.

[0007] It is an objective of the present invention to provide an improved automation of a process plant.

[0008] The problem is solved by a plurality of field components of a process or a manufacturing plant according to claim 1. Furthermore, the present disclosure is directed to a method of automating at least a part of a process or a manufacturing plant according to claim 12. Still further, the present disclosure is directed to an engineering station server according to claim 13. Still further, the present disclosure is directed to a method of generating an automation for at least a part of a process or a manufacturing plant according to claim 16. Still further, the present disclosure is directed to a control system for a process or manufacturing plant according to claim 17. Before undertaking a detailed description, certain terms used in the description are explained below.

[0009] A "field component" refers to a device or part of the process plant which interacts with the process. A field component can be, for example, a sensor, an actuator (e.g., pump, valve) , a pipe branch, and process equipment (e.g., tank, heat exchanger, reactor, crystallizer) , among others. Also, a pipe node can be a field component in this disclosure. Thereby, a pipe node can be an intersection of pipes, or an inlet and / or an outlet of certain process components.

[0010] An "agent" represents one field component of the process plant by way of a piece of software, which can be executed to carry out defined agent functions. The piece of software may run on the field component itself, if the field component has a required processing capability, or may run on another device, such as a nearby field component, or on a centralized computing system, such as an engineering station server.

[0011] A "request" refers to information about a certain need of an agent in the process plant. A request is generated during a pairing run and may carry, among other information, a request type, information about the requesting agent, a priority value, and a penalty value.

[0012] "Neighboring" agents refers to agents representing field components that are adjacently connected topologically, e.g., via a respective pipe.

[0013] According to the invention, a plurality of field components of a process or a manufacturing plant comprises respectively a computer-implemented agent which is designed for checking the operational requirements of the respective field component, especially limit violations or deviations of setpoint, generating a request about the present needs of the respective field component, sending the request about the present needs of the respective field component to other field components (more precisely : to the agent ( s ) of the other field components ) , using a communication connection, forwarding requests received from a further field component to a still further field component or to a plurality of field components , performing reactions to requests the computer-implemented agent has received from other field components , i f the field component has the capability to do so , and confirming a ful fillment of requests which the computer-implemented agent has received from other field components i f the field component has the capability to do so .

[0014] Thereby, the field components further comprise respectively a computer-implemented behavior agent which is speci fic for a type of the respective field component .

[0015] During operation of the process or manufacturing plant each computer-implemented agent is designed for checking its requirements , especially at every time step . This check can include an examination of limit violations or deviations of a setpoint .

[0016] Further, the computer-implemented agent is designed for generating a request i f his requirements are not ful filled any longer . In this request all information needed to find the best suiting control circle can be included, e . g . the ID of the creating agent , the request type ( e . g . , change in flow is requested) or the priority of the request ( e . g . , a limit violation can be more important than a setpoint ) .

[0017] Still further, the computer-implemented agent is designed for sending an according request generated by his own to all physically connected components defined in the agent ' s behavior .

[0018] Still further, the computer-implemented agent is designed for forwarding a received request to the connected agents defined in the agent behavior. Preferably, the request is only forwarded if the request is not already known by him.

[0019] If a request is sent to the agent, he will process this request to decide if he is able to reach the asked changes. During these processing he can check some preconditions (e.g., if the request is already known by him) , he can update request specific values included in the request (e.g. penalties, indicating the "distance" to the request sending agent) and he will check if his own behavior allows him to react on the request. If this is possible, he preferably transmits all relevant data to a superimposed pairing service of the process or manufacturing plant.

[0020] Still further, the computer-implemented agent is designed for confirming a fulfillment of requests.

[0021] The invention discloses only one common computer-implemented agent for all field components. The individual behavior of the field components is expressed by a computer-implemented behavior agent. In difference to the state of the art, using this new concept all agents are represented by one and the same general agent and not by individually agents for each agent type (e.g., pump, valve, pressure sensor) . Therefore, only one general agent is necessary. Hence, the complexity of the concept and the programming code is reduced significantly. This leads to reduced maintenance costs and reduced error- proneness. Additionally, the transferability to different field components and the integration of further new components are now possible, as well as the integration of environmental conditions, e.g. changing flow directions, which was not addressable in the already known concepts. Therefore, the concept of the agent-based process automation is getting more reasonable, and a widespread usage in different plant topologies and different complex processes is getting possible. Preferably, the computer-implemented behavior agent comprises information about a flow direction o f physical intersections of the respective field component . Thereby, it is known which physically intersection is an inlet , and which one is an outlet . In some processes this matrix can change .

[0022] According to one aspect of the invention, the computer- implemented behavior agent comprises information about a convenience value , which influences a decision on which the respective field component react . The convenience represents the ability to react to di f ferent request types ( e . g . pressure , flow) with di f ferent ef ficiency e . g . , a convenience value of 0 corresponds to having no influence on this request type .

[0023] Advantageously, the computer-implemented behavior agent comprises information about a penalty, which reflects the ability of the reacting field component to influence the request sending field component , wherein the computer- implemented behavior agent is designed for changing the penalty if the field component might reduce the ef fectiveness of the reacting field component to influence the needs of the request sending further field component . The penalty ( indicating the "distance" to the request sending agent ) has to be changed during forwarding individually for each agent e . g . , the penalty change regarding a temperature request is 0 i f the agent has no influence on the temperature at all .

[0024] Preferably, the computer-implemented behavior agent is designed for sending an information about its ability to react on di f ferent request types to a computer-implemented pairing service .

[0025] According to one aspect of the invention, the computer- implemented behavior agent is designed for changing a direction of a control reaction of the field component . In some agents the direction of the control reaction is changing, e . g . a control valve has to be closed to increase the pressure upstream . Preferably, the computer-implemented behavior agent is designed for sending different request types to further field components. According to their abilities, the behavior agents can send different request types, e.g. a pump will send a pressure request while a passive component is not able to send any request.

[0026] Advantageously, the computer-implemented behavior agent is designed for forwarding a request received from a further field component to one or more specific field components. In some cases, a forwarding to all possible connections is not reasonable due to request type (e.g., temperature) or setup of components (e.g., tank) .

[0027] According to one aspect of the invention, the computer- implemented behavior agent is designed for generating a request about the present needs of the respective field component with a specific priority. Request creating behavior agents can assign a priority value to their request according to the reason of their request.

[0028] Preferably, the computer-implemented behavior agent comprises one or more matrices for storing the information in connection with the computer-implemented behavior agent. Thereby, the special behavior of the respective field component can be adjusted easily to changing processes.

[0029] Advantageously, the computer-implemented agent is designed for :

[0030] - receiving a target operation range for an operation of the respective field component,

[0031] - receiving a setpoint range for the operational setpoints of the respective field component,

[0032] - in case, the target operation range is not reached yet, creating an optimization request, and sending it to the neighbored field components, - receiving an optimization request of another field component and verifying if it could compensate a specific change in the operation point of the sending field component.

[0033] For example, the optimization request concerns an energy optimization. All behavior agents can send an optimization request if the unit is in steady state and the operating point of the component is not in its optimal (e.g. energy) range. This request will be forwarded and handled by the other agents as described above. If another agent is able to compensate a potential change in the operating point of the asking agent, it can change its own operating point. The agent can also send an answer to a computer-implemented pairing service of the process or manufacturing plant. The computer-implemented pairing service can decide which agents asking for an optimization are allowed to change their operating point based on the degree of freedoms in the unit, priorities, and feasible solutions. If an agent is allowed to change its operating point, he will do so based on his allowed step size and limitations .

[0034] After the according field component is back in steady state again, the procedure can be repeated as long as the limits of the relevant field components are not violated and as long as no changes in the requirements (e.g., setpoints) or configuration occurs. If after the operation point was changed limitations in the field component are exceeded, the last optimization step can be reversed, and the affected agents will not be allowed to be part of further optimization steps within the same optimization procedure.

[0035] In difference to known methods, no models of the plant or process are needed, as well as no measurement data for any kind of training or data-based modeling. As the agents do not need any kind of mathematical model and the optimization algorithm is very simple, the complexity as well as the costs for the optimization can be significantly reduced. Furthermore , an expert knowledge is not necessary to use this kind of optimi zation as only some easy-understandable parameters have to be set .

[0036] Due to the simplicity the results of the optimi zation are very fast available . Opposing goals like quality, productivity or energy requirements can be easily solved by setting up target ranges for these values . Therefore , the hurdle to use an optimi zation algorithm can be reduced by this advantageous aspect of the invention .

[0037] The problem is also solved by a method of automating at least a part of a process or a manufacturing plant , wherein the process or manufacturing plant comprises a plurality of field components according to one of the preceding claims , comprising :

[0038] Establishing a communication network or using an existing communication network between the plurality of field components ,

[0039] Each field component forwards requests which it has received from a further physically connected field component to further, physically connected field components ,

[0040] Each field component performs reactions to requests it has received from other field components , i f the field component has the capability to do so ,

[0041] Each field component confirms a ful fillment of requests which it has received from other field components i f the field component has the capability to do so .

[0042] The term „automating" is understood as a computer-controlled operation of the plurality of field components to achieve a distinct operational behavior of the process plant .

[0043] The problem is also solved by an engineering station for an automation of a process plant , comprising a computer- implemented engineering program, which is designed for generating the automation of a plurality of field components of a process or a manufacturing plant , wherein the engineering program comprises respectively a computer-implemented agent for each field component which is designed for, within the engineering program : checking the operational requirements of the respective field component , especially limit violations or deviations of setpoint , generating a request about the present needs of the respective field component , sending the request about the present needs of the respective field component to other field components (more precisely : to the agent ( s ) of the other field components ) , using a virtual communication connection, forwarding requests received from a further field component to a still further field component or to a plurality of field components , performing reactions to requests the computer-implemented agent has received from other field components , i f the field component has the capability to do so , and confirming a ful fillment of requests which the computer-implemented agent has received from other field components , i f the field component has the capability to do so , wherein the field components further comprise respectively a computer-implemented behavior agent which is speci fic for a type of the respective field component .

[0044] Preferably, the computer-implemented agent is designed for, within the engineering program :

[0045] - receiving a target operation range for an operation of the respective field component ,

[0046] - receiving a setpoint range for the operational setpoints of the respective field component ,

[0047] - in case , the target operation range is not reached yet , creating an optimi zation request , and sending it to the neighbored field components ,

[0048] - receiving an optimi zation request of another field component and veri fying i f it could compensate a speci fic change in the operation point of the sending field component , - sending an according message back to the field component of the optimi zation request ,

[0049] - changing the operation setpoint of the field component .

[0050] Most preferably, the engineering program comprises a computer- implemented pairing service which is designed for coordinating a communication between the agents of the field components within the engineering program . The pairing service can also be designed for identi fying a fitting partner ( e . g . sensoractor ) within a control loop .

[0051] The problem is also solved by a method of generating an automation for at least a part of a process or a manufacturing plant , using an engineering station as described above .

[0052] The problem is also solved by a control system for a process or manufacturing plant , comprising at least one operator station for monitoring and controlling the proces s or manufacturing plant , and a plurality of field components as described above and / or an engineering station as described above .

[0053] Features of examples of the present disclosure will become apparent by reference to the following description of exemplary embodiments of the invention . The figures are showing :

[0054] FIG 1 a control system for a process plant according to a first aspect ;

[0055] FIG 2 a control system for a process plant according to a second aspect ;

[0056] FIG 3 a first method of automating a process plant , illustrated in a flow chart ; and

[0057] FIG 4 a second method of automating a process plant , illustrated in a flow chart . FIG 1 shows a pipe and instrumentation diagram (P&ID) of a process plant, comprising a plurality of field components. There are shown four valves 1, 2, 3, 4, a pump 5 with a motor and a tank 6. There are four valve position sensors V01, V03, V04, V06 to determine a position of the valves 1, 2, 3, 4. In addition, there is one motor sensor MOI to determine a rotation speed of the pump motor. In addition, there is one pressure sensor P02 between the pump 5 and the two valves 2, 3. In addition, there are two temperature sensors T06, T07 in an inflow of the tank 6 and within the tank 6 itself. In addition, there is a level sensor L07 which determines a level of a fluid within the tank 6.

[0058] FIG 2 shows, within the P&ID of FIG 1, agents which are used by an engineering station to generate an automation for the process plant. The agents are computer-implemented within an engineering program of the engineering station. The agents are designed for, within the engineering program: checking the operational requirements of the respective field component, especially limit violations or deviations of setpoint, generating a request about the present needs of the respective field component, sending the request about the present needs of the respective field component to other field components, using a virtual communication connection, forwarding requests received from a further field component to a still further field component or to a plurality of field components, performing reactions to requests the computer-implemented agent has received from other field components, if the field component has the capability to do so, and confirming a fulfillment of requests which the computer-implemented agent has received from other field components, if the field component has the capability to do so, wherein the field components further comprise respectively a computer-implemented behavior agent which is specific for a type of the respective field component.

[0059] In the following, the computer-implemented agent and the behavior agent are expressed by the common word "agent" to make it easier to understand the description of the figures.

[0060] In FIG 2 agents Al, A6, A7, A9 represent the valves 1, 2, 3, 4. Furthermore, there are agents A2 for the pump 5, agents A4 for the pressure sensor P02, agents All for the temperature sensor T07 within the tank 6, agents A10 for the level sensor L07 and agents A3, A5, A8, A12 for pipe intersections.

[0061] For validation of the inventive concepts three test scenarios have been simulated. In a first scenario only a setpoint for the level sensor L07 of the tank 6 was set while the setpoints of all the other sensors P02, T07 remained empty. At the beginning all valves 1, 2, 3, 4 were closed and the pump 5 was not working. FIG 3 shows the simulation results. Therein, in the first part 7 the measured level and its setpoint are plotted (X-axis: time in sec, Y-axis: level in %) .

[0062] In the second part 8 of FIG 3 (X-axis: time in sec, Y-axis: values of variables) the manipulated variables of the active control circles as well as the readback from the actors can be found. As it can be seen in FIG 3, the valves 1, 2, 3, 4 ("VI, V2, V3, V4") are opened one after the other until the pump 5 ("P") j_sswitched on. After the pipe is open and the pump 5 is running, the level in the tank 6 can be changed only by using valve 4.

[0063] In the third part 9 of FIG 3 (X-axis: time in sec, Y-axis: pairing partners) the corresponding pairing partners are plotted. In a first step the level sensor L07 is paired to valve 4 (agent A9) . After reaching its limits, valve 4 is additionally paired with valve 2 (agent A6) . Reaching its limits, valve 4 is additionally paired with valve 1 (agent Al) . Valve 1 itself is paired with the pump 5 (agent A2 ) after the limits are violated again. For a short time, the pump 5 is also paired to valve 3 (agent A7 ) .

[0064] In a second scenario a setpoint is assigned each of the level sensor L07, pressure sensor P02 and temperature sensor T07. The simulation results can be found in FIG 4. The three parts 10, 11, 12 show results analogously to the structure of FIG 3. The temperature is controlled with the help of valve 3 (agent A7 ) , the pressure with the pump 5 (agent A2 ) and the level with a control circle with valve 4 (agent A9) . During the short time the valves 3, 4 being at their limits valve 2 (agent A6) and valve 1 (agent Al) are used additionally. The control problem can be solved .

[0065] Due to the separation of individual behavior and functionality, only one general agent is necessary any longer. Additionally, the special behavior and the environmental conditions are stored in matrices and therefore are well known and can be easily adjusted to a changing process. Hence, the complexity of the concept and the programming code is reduced significantly. This leads to reduced maintenance costs and reduced error- proneness. Additionally, the transferability to different field components and the integration of further new components are now possible as well as the integration of environmental conditions, e.g. changing flow directions which was not addressable in the already known concepts. Therefore, the concept of the agent-based process automation is getting more reasonable, and a widespread usage in different plant topologies and different complex processes is getting possible. The invention can e.g. be easily applied for creating the automation logic in a distributed control system (DCS) . Furthermore, for simulation testing of different plant topologies during plant design the automation can be created automatically. This enables the process engineer to simulate the controlled process directly without any further effort which generates high benefits as this simulation procedure is highly iterative.

[0066] In a third scenario the pump 5 (more precisely: the agent A2 of the pump 5) sends an optimization request. Dependent on the plant requirements (active setpoints) an optimization procedure is executed. For an easier understanding of the algorithm in the following an example of only one active setpoint is used. The level is the control variable of the only control loop. FIG 5 shows the simulation results in an overview. Therein, in the first subplot 13 the level and its setpoint can be seen (X-axis: time in sec, Y-axis: level in

[0067] In the second subplot 14 (X-axis: time in sec, Y-axis: pressure and rotational speed in arbitrary units) the pressure (circles) and the rotational speed (crosses) of the pump 5 are plotted.

[0068] In the third subplot 15 (X-axis: time in sec, Y-axis: valve position) the valve positions of the valves 1, 2, 3 not involved in the final control loop are shown. In the last subplot 16 the valve position (X-axis: time in sec, Y-axis: valve position) of the valve 4 included in the final control loop can be seen. In the first 300 seconds the whole unit is automatically started up via the agent system. When the unit has been started up and is in a steady state (see dotted black line L) the optimization procedure is started.

[0069] The optimization is started at the black dotted line L after steady state is reached. The pump 5 is allowed to reduce its rotational speed as the valve 4 is not at its upper limit and the level of the tank 6 is within its setpoint range.

[0070] After changing the working point the next optimization step will be done if the unit is back in steady state again. However, in the third optimization step (the rotational speed is reduced for the third time) the valve which compensates the pump 5 reduction is at its artificial limit (set here to 30% to show the behavior of the algorithm) and the last optimization step is reversed to fulfill the requirements.

[0071] Using the agents to generate the automation structure, the optimization goes on by using the other valves 1, 2, 3 not included in a control loop and not being at their upper limit. In this case valve 1, valve 2 and valve 3 could also be more opened .

[0072] As no modelling or no plant data are necessary, this optimization method is very easy to use. No modelling or optimization experts are required as only some easy- understandable parameters have to be set. Therefore, the invest in the optimization is very low regarding other optimization technics. Due to the simplicity of this method the hurdle to use an optimization algorithm is very small. The computational time to solve the optimization problem is very fast as no complex algorithms have to be solved. Opposing goals like quality, productivity and energy requirements can easily solved by setting up target ranges for these values.

[0073] The agents can also be used during a runtime of the technical plant to update / optimize the automatisation, e.g. the control loops .

[0074] While several examples have been described in detail, it is to be understood that the disclosed examples may be modified. Therefore, the foregoing description is to be considered nonlimiting .

Claims

Claims1. A plurality of field components (1, 2, 3, 4, 5, 6) of a process or a manufacturing plant, which field components (1,2, 3, 4, 5, 6) comprise respectively a computer-implemented agent (Al, A2, A3, A4, A5, A6, A7, A8 , A9, A10, All, A12) which is designed for: checking the operational requirements of the respective field component (1, 2, 3, 4, 5, 6) , especially limit violations or deviations of setpoint, generating a request about the present needs of the respective field component (1, 2, 3, 4, 5, 6) , sending the request about the present needs of the respective field component (1, 2, 3, 4, 5, 6) to other field components (1, 2, 3, 4, 5, 6) , using a communication connection (1, 2, 3, 4, 5, 6) , forwarding requests received from a further field component (1, 2, 3, 4, 5, 6) to a still further field component (1, 2, 3, 4, 5, 6) or to a plurality of field components (1,2, 3, 4, 5, 6) , performing reactions to requests the computer-implemented agent has received from other field components (1, 2, 3, 4, 5, 6) , if the field component (1, 2, 3, 4, 5, 6) has the capability to do so, and confirming a fulfillment of requests which the computer-implemented agent has received from other field components (1, 2, 3, 4, 5, 6) if the field component (1, 2, 3, 4, 5, 6) has the capability to do so, wherein the field components (1, 2, 3, 4, 5, 6) further comprise respectively a computer-implemented behavior agent (Al, A2, A3, A4, A5, A6, A7 , A8 , A9, A10, All, A12) which is specific for a type of the respective field component (1, 2,3, 4, 5, 6) .

2. A plurality of field components (1, 2, 3, 4, 5, 6) according to claim 1, wherein the computer-implemented behavior agent comprises information about a flow direction ofphysical intersections of the respective field component (1, 2, 3, 4, 5, 6) .

3. A plurality of field components (1, 2, 3, 4, 5, 6) according to claim 1 or claim 2, wherein the computer- implemented behavior agent comprises information about a convenience value, which influences a decision on which the respective field component (1, 2, 3, 4, 5, 6) will react.

4. A plurality of field components according to one of the preceding claims, wherein the computer-implemented behavior agent (Al, A2, A3, A4, A5, A6, A7, A8 , A9, A10, All, A12) comprises information about a penalty, which reflects the ability of the reacting field component (1, 2, 3, 4, 5, 6) to influence the request sending field component (1, 2, 3, 4, 5, 6) , wherein the computer-implemented behavior agent is designed for changing the penalty if the field component (1, 2, 3, 4, 5, 6) might reduce the effectiveness of the reacting field component (1, 2, 3, 4, 5, 6) to influence the needs of the request sending further field component (1, 2, 3, 4, 5, 6) .

5. A plurality of field components (1, 2, 3, 4, 5, 6) according to one of the preceding claims, wherein the computer- implemented behavior agent (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, All, A12) is designed for sending an information about its ability to react on different request types to a computer- implemented pairing service.

6. A plurality of field components (1, 2, 3, 4, 5, 6) according to one of the preceding claims, wherein the computer- implemented behavior agent (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, All, A12) is designed for changing a direction of a control reaction of the field component (1, 2, 3, 4, 5, 6) .

7. A plurality of field components (1, 2, 3, 4, 5, 6) according to one of the preceding claims, wherein the computer- implemented behavior agent (Al, A2, A3, A4, A5, A6, A7, A8,A9, A10, All, A12) is designed for sending different request types to further field components (1, 2, 3, 4, 5, 6) .

8. A plurality of field components (1, 2, 3, 4, 5, 6) according to one of the preceding claims, wherein the computer- implemented behavior agent (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, All, A12) is designed for forwarding a request received from a further field component (1, 2, 3, 4, 5, 6) to one or more specific field components (1, 2, 3, 4, 5, 6) .

9. A plurality of field components (1, 2, 3, 4, 5, 6) according to one of the preceding claims, wherein the computer- implemented behavior agent (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, All, A12) is designed for generating a request about the present needs of the respective field component (1, 2, 3, 4, 5, 6) with a specific priority.

10. A plurality of field components (1, 2, 3, 4, 5, 6) according to one of the preceding claims, wherein the computer- implemented behavior agent (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, All, A12) comprises one or more matrices for storing the information in connection with the computer-implemented behavior agent.

11. A plurality of field components (1, 2, 3, 4, 5, 6) according to one of the preceding claims, wherein the computer- implemented agent (Al, A2, A3, A4, A5, A6, A7, A8 , A9, A10, All, A12) is designed for:- receiving a target operation range for an operation of the respective field component (1, 2, 3, 4, 5, 6) ,- receiving a setpoint range for the operational setpoints of the respective field component (1, 2, 3, 4, 5, 6) ,- in case, the target operation range is not reached yet, creating an optimization request, and sending it to the neighbored field components (1, 2, 3, 4, 5, 6) ,- receiving an optimization request of another field component (1, 2, 3, 4, 5, 6) and verifying if it could compensate aspecific change in the operation point of the sending field component (1, 2, 3, 4, 5, 6) ;- changing the operation setpoint of the field component (1,2, 3, 4, 5, 6)12. Method of automating at least a part of a process or a manufacturing plant, wherein the process or manufacturing plant comprises a plurality of field components (1, 2, 3, 4, 5, 6) according to one of the preceding claims, comprising:Establishing a communication network or using an existing communication network between the plurality of field components (1, 2, 3, 4, 5, 6) ,Each field component forwards requests which it has received from a further physically connected field component to further, physically connected field components (1, 2, 3, 4, 5, 6) ,Each field component performs reactions to requests it has received from other field components (1, 2, 3, 4, 5, 6) , if the field component has the capability to do so,Each field component confirms a fulfillment of requests which it has received from other field components (1, 2, 3, 4, 5, 6) if the field component (1, 2, 3, 4, 5, 6) has the capability to do so.

13. Engineering station for an automation of a process plant, comprising a computer-implemented engineering program, which is designed for generating the automation of a plurality of field components (1, 2, 3, 4, 5, 6) of a process or a manufacturing plant, wherein the engineering program comprises respectively a computer-implemented agent (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, All, A12) for each field component (1, 2,3, 4, 5, 6) which is designed for, within the engineering program: checking the operational requirements of the respective field component (1, 2, 3, 4, 5, 6) , especially limit violations or deviations of setpoint, generating a request about the present needs of the respective field component (1, 2, 3, 4, 5, 6) ,sending the request about the present needs of the respective field component (1, 2, 3, 4, 5, 6) to other field components (1, 2, 3, 4, 5, 6) , using a virtual communication connection, forwarding requests received from a further field component (1, 2, 3, 4, 5, 6) to a still further field component (1, 2, 3, 4, 5, 6) or to a plurality of field components (1,2, 3, 4, 5, 6) , performing reactions to requests the computer-implemented agent has received from other field components (1, 2, 3, 4, 5, 6) , if the field component (1, 2, 3, 4, 5, 6) has the capability to do so, and confirming a fulfillment of requests which the computer-implemented agent has received from other field components (1, 2, 3, 4, 5, 6) , if the field component (1, 2,3, 4, 5, 6) has the capability to do so, wherein the field components (1, 2, 3, 4, 5, 6) further comprise respectively a computer-implemented behavior agent (Al, A2, A3, A4, A5, A6, A7 , A8 , A9, A10, All, A12) which is specific for a type of the respective field component (1, 2, 3, 4, 5, 6) .

14. Engineering station according to claim 13, wherein the computer-implemented agent (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, All, A12) is designed for, within the engineering program:- receiving a target operation range for an operation of the respective field component (1, 2, 3, 4, 5, 6) ,- receiving a setpoint range for the operational setpoints of the respective field component (1, 2, 3, 4, 5, 6) ,- in case, the target operation range is not reached yet, creating an optimization request, and sending it to the neighbored field components (1, 2, 3, 4, 5, 6) ,- receiving an optimization request of another field component (1, 2, 3, 4, 5, 6) and verifying if it could compensate a specific change in the operation point of the sending field component (1, 2, 3, 4, 5, 6) ,- sending an according message back to the field component (1, 2, 3, 4, 5, 6) of the optimization request,- changing the operation setpoint of the field component (1, 2, 3, 4, 5, 6) .

15. Engineering station according to claim 13 or 14, wherein the engineering program comprises a computer-implemented pairing service which is designed for coordinating a communication between the agents (Al, A2, A3, A4, A5, A6, A7, A8, A9, A10, All, A12) of the field components (1, 2, 3, 4, 5, 6) within the engineering program.

16. Method of generating an automation for at least a part of a process or a manufacturing plant, using an engineering station according to one of claims 13 through 15.

17. Control system for a process or manufacturing plant, comprising at least one operator station for monitoring and controlling the process or manufacturing plant, and a plurality of field components (1, 2, 3, 4, 5, 6) according to one of claims 1 through 11 and / or an engineering station according to one of claims 13 through 15.

Images