Description

The invention relates to a system for automated handling of at least one working step of a workflow in an automation and/or electrical engineering project, where the system comprises a data input interface for reading in at least one input value belonging to the at least one working step of the workflow of the project, a processing unit connected to the user interface for generating at least one output value from the at least one input value, where the at least one output value represents a result of the at least one working step and a data output interface connected to the processing unit for transmitting any data read in by the data input interface or generated by the processing unit to at least one of a data repository, another processing unit and/or a display unit.

Automation projects or electrical engineering projects are projects which are aimed at the bidding, planning, designing, installation, commissioning and servicing of technical installations or systems for automating and/or supplying power to a technical process or facility, where the technical installation or system can be for example a process automation system for a factory or a control system for a power plant or a power supply system for a building complex. Such a project usually undergoes various phases, which may occur subsequently or in parallel, such as tendering, collection of requirements, planning of information signals, planning of power cabling and/or communication networks, engineering of control logic, configuration of human machine interfaces (HMIs) and system integration. The phases themselves can again be divided into a certain number of working steps. A working step is understood to comprise at least one activity which generates at least one output information and/or at least one physical result. Inside a workflow of a project, the at least one output information and/or physical result represents at least one input information and/or prerequisite for a subsequent working step, respectively. Examples for working steps are the planning of cross controller communication during the planning of information signals, the designing of process graphics during the configuration of HMIs, the implementation of sequence control logics during the control logic engineering or the parameterization of controllers during system integration and sub-parts of these activities.

Nowadays, it is an aim to perform at least some of these working steps automatically by computer tools. Examples for currently known automated working steps are the configuration of devices based on templates, the auto-generation of control code for I/O boards and the auto-generation of order requests based on material lists. It is a further aim for the future that more and more of these computer tools are enabled to communicate with each other so that not only a single working step but a whole sequence of working steps can be performed automatically.

During execution of the phases and their corresponding working steps, a considerable number of different professions, different computer tools and different types of information have to go hand in hand and need to be coordinated in order to ensure that the resulting technical installation or system functions properly. Some working steps cannot be performed before other working steps are successfully finished or before certain input information is available.

These interrelations between working steps can be visualized by a workflow diagram, such as depicted in Fig. 3, where the numbered rectangles 101 to 113 each represent a working step. The four input circles on the left hand side illustrate external inputs which are required for performing the working steps 101 , 102 and 104 connected by arrows to the input circles. Working steps 101 and 104 each receive two external inputs. On the right hand side, the external outputs generated by the working steps 1 10, 1 12 and 113 are depicted by three output circles connected again by arrows to the corresponding working steps. Working steps 103, 105 to 109 and 11 1 only have internal input and output connections, i.e. they receive their inputs from one or more preceding working steps of the workflow and deliver their outputs to one or more subsequent working steps of the workflow. In general, inputs to and outputs from working steps can be on the one hand data, information or decisions and on the other hand physical objects, such as documents, tools or installed equipment, no matter whether they are external or internal inputs or outputs. At least some of the working steps of Fig. 3 may be performed automatically by using a system as shown in Fig. 1. A first computing device PC1 contains a data input interface 1 for reading in input information of a working step, where the input information is delivered by either a second computing device PC2, which may have performed one or more preceding working steps, by a first data repository 12 located for example on a central data server, or by interaction with a user via for example an acoustic input device 10, a pointing device 1 1 and/or a keyboard. The input information is represented by one or more input values, where the input values may be constants or variables, predefined or continuously changing and where the input values may be given directly as numbers or in text form which is later transformed into numbers. The first computing device PC1 further comprises a processing unit 3 for generating the output information of the working step by processing the input information during performance of a predefined task allocated to the working step, and a data output interface 2 for transmitting input as well as output information to further devices for further processing and/or data storage and/or acoustic or graphic visualization. The further devices may be a third computing device PC3 which is arranged for performing subsequent working steps, a second data repository 13 which may be located on the same central data server as the first data repository 12 or on another storage device, a graphic display unit 14 or an acoustic display unit 15. The output information is represented by one or more output values, which again may be constants or variables and may be given in text form or directly as numbers. The second and third computing devices, PC2 and PC3, both may contain the same components as the first computing device PC1 , i.e. a data input interface 4 or 7, a processing unit 6 or 9 and a data output interface 5 or 8, respectively. As becomes clear from Fig. 3, a workflow of an automation and/or electrical engineering project may contain a considerable number of closely meshed working steps where each working step may only be performed when all necessary inputs are available and where a change in one external input or in the output of a preceding working step may affect a comparatively large number of subsequent working steps.

In real life execution of automation and/or electrical engineering projects, the collection of required input information can be a cumbersome task and may lead to delays. In addition, facts and data as well as the physical environment of a project are usually not static, which leads to frequent changes. These changes result in further delays since a number of already finished working steps have to be performed again. It is the object of the present invention to suggest a system and method for automated handling of at least one working step of a workflow in an automation and/or electrical engineering project with which the above named side effects of input collection and changes during project execution can be reduced.

This object is achieved by a system and a method according to the independent claims.

In the system according to the invention the data input interface is arranged to read in at least one input uncertainty characterizing a possible or allowed range of the input value, the processing unit is arranged to calculate from the at least one input value and the at least one input uncertainty at least one output uncertainty characterizing a possible or allowed range of the output value of the at least one working step, by taking into account a parameter and/or a technical condition of the working step As already described above, the term working step is used for at least one activity performed during the execution of the project, where the at least one activity generates at least one output information. The at least one output information is most preferably an indication of a quantity, where the quantity may for example be given as number of devices or number of communication signals, or as a length of cable or size of a cabinet. In the alternative, the output information may also be given as an indication for a specific type of equipment which may be selected from a couple of choices, for example a specific type of field bus, controller device, signal transmission protocol or software tool to be used.

Parameters or technical conditions of the working step are those parameters and conditions which have the highest influence on the magnitude of the output uncertainty. For example, a certain type of technical equipment may be specified to be able to handle a predefined optimum amount of a physical entity during normal operating conditions and up to a maximum or down to a minimum of this physical entity connected with a degradation in the operating conditions, i.e. only during a shorter period of time or at the cost of operational speed etc. When this certain type of technical equipment plays a role during a working step, i.e. when the processing unit has the task to determine how many devices of this technical equipment are needed to handle an uncertain amount of the physical entity, the above named specification is taken into account by the processing unit in order to evaluate whether the input uncertainty can be handled by the specific number of devices without needing to specify an uncertainty for the number of devices. By allowing for input values to be accompanied by a corresponding uncertainty value and by taking the uncertainty into account when performing the task allocated to the working step so that the output value or values are generated together with an uncertainty value as well, it becomes possible to execute or perform working steps even when not all of the required input information is available. In such cases, input values may be created from available knowledge, such as estimations or heuristics or experiences from previous projects which may be laid down in formulas or tables.

An example from the field of automation and/or electrical engineering projects, is the external input information about the number of information signals to be transmitted inside the technical installation or system. When this number can not be given exactly but at least with an uncertainty value lying within a reasonable limit, i.e. 5000 information signals with 5% uncertainty, it becomes possible to plan already the number of required computer servers and/or controller devices without having to fear that these numbers will have to be corrected. Accordingly, the working step of planning the computer and/or control equipment can be performed earlier, even when not all external input information is present, resulting in a saving of time.

A further example for a working step requiring external input values is the planning of storage capacity for an information management system belonging to the technical installation or system. External input values are here the number, type and sample frequencies of the data signals to be stored in for example a historian server of the information management system. When upper and lower boundaries of these input values are known, estimations for the required storage capacity can be made early in the project which may then allow for an iterative adjustment of the desired number and sample frequency.

Subsequent working steps may also be performed without having all external input values available exactly, since the subsequent working steps receive as their input values the output values of the preceding working step or steps together with the corresponding output uncertainties which become the new input uncertainties.

For some of the subsequent working steps the uncertainty may not influence their outputs at all, as long as the output uncertainty of the preceding working step lies within certain limits. An example for such a preceding working step in an automation and/or electrical engineering project is the planning of cross controller communication, i.e. of signal communication between controller devices. An output value is the figure for the expected number of signals to be communicated between controller devices. Even if this figure can be generated only as an approximation, i.e. as an output value with corresponding output uncertainty, this may be sufficient to design, in a subsequent working step, the topology of the communications network. A re-design of the network topology will in most cases not be necessary after the exact figure becomes available, since one and the same network topology is usually applicable for a wide range of actual communication implementations. In this way, time can again be saved, since more working steps can be performed in parallel even if preceding working steps are not absolutely complete or if required input information is not yet available with absolute certainty. Not only single working steps may be performed earlier due to taking into account uncertainties, also whole phases of automation and/or electrical engineering projects can be performed completely allowing for the execution of subsequent phases. For example, the planning of power cabling may generate a preliminary figure for the expected length of power cable required for a subsequent equipment installation phase, where this preliminary figure is again accompanied by an uncertainty value. Even though this figure is not yet exact, it gives the possibility to at least order a minimum length of power cable so that the subsequent installation phase may begin much earlier compared to waiting until the input information is complete. On the other hand, if the common approach was to order a generous quantity of material and equipment in order to be prepared for changes, the invention allows for limiting this quantity in accordance with the maximum quantity given by the output uncertainty. This results in a reduction of hardware costs of the project.

As was mentioned, the decision about whether a working step can be performed despite its input values being not exact depends on the value of the respective uncertainties. If a given uncertainty exceeds a reasonable limit, for example becomes larger than 30 % of the corresponding input value, it may be more suitable to wait until the uncertainty decreases below this limit or until the exact input information is available. Accordingly, the invention provides an additional means to automatically manage the timely execution of the project, where the uncertainty information helps in taking a reasoned decision about which working step is to be performed when.

A further advantage of the invention is that a change in an external input value as well as in a parameter or a technical condition of a working step does not automatically result in the re- performing of all subsequent working steps. The change in the parameter or technical condition leads to a change in the output uncertainty of the working step. For those working steps which are subsequent to the changed external input value or changed output uncertainty it has to be checked simply, whether the change falls within the input uncertainty range which was considered earlier when performing the respective subsequent working step at an earlier point in time, and whether at this earlier point in time the determined output uncertainty of the subsequent working step was zero or not. A zero output uncertainty indicates that as long as the input uncertainty range is not exceeded, the output value of the subsequent working step remains unchanged, i.e. for such a subsequent working step an update is not required. Accordingly, change requests in the automation and/or electrical engineering project can be processed and executed with less effort due to the invention.

An example for an unnecessary update of a working step is the planning of controller cabinets, i.e. of the housings within which the controller devices are to be installed. Input values to the working step are for example the number and type of controllers and the number and type of information signals. Output values are the number as well as the size and thereby the capacity of the controller cabinets. These can be planned so that a certain amount of empty space is left inside the cabinets in accordance with the upper limit of the input uncertainty corresponding to at least one of the input values. As long as this input value remains below its upper limit, changes in the input value do not affect the output values so that the working step of planning of controller cabinets needs to be performed only once.

The method according to the invention contains all above described steps performed by one of the elements of the system, i.e. by the data input interface, the processing unit and the data output interface.

In an embodiment of the invention, the workflow contains at least two working steps, the at least one output value represents an input value of at least one of the at least two working steps and each of the at least two working steps is defined to generate at least one corresponding output value, and where the processing unit is arranged to predict whether the at least one input uncertainty propagates through the at least two working steps of the workflow or not.

In other words, the processing unit is arranged to check whether an input uncertainty influences the output variable or variables of working steps having this input uncertainty assigned to one of their input values, i.e. whether the output variables can also be given with upper and lower boundaries only. The processing unit performs this checking working step for working step following the logical sequence of the workflow until the output uncertainties will be zero. During this checking, the working steps themselves are not executed. Instead, the processing unit predicts the propagation of the input uncertainty using one or more of look-up tables, heuristics, formulas, neural networks or fuzzy logic. The quantitative statement generated as a result of this embodiment may be provided by the processing unit via the data output interface to a reasoning engine where it is decided whether some of the working steps of the workflow may be started to be executed. These working steps will preferably be those working steps which are unaffected by the input uncertainty. The embodiment can be further extended by arranging the processing unit to be able to predict the extent to which the at least one input uncertainty propagates through the at least two working steps of the workflow by calculating a corresponding output uncertainty for each of the at least two working steps, where again the above named methods may be used. Instead of a qualitative statement as above, a quantitative statement is generated here which may help to decide for which working steps of the workflow the corresponding input uncertainties are small enough to start performing. In this way, more working steps may be executed at an earlier point in time compared to the generation of q mere qualitative statement of the propagation of the input uncertainty. The processing unit may also be arranged to predict whether and/or to which extent a change in the at least one input variable propagates through at least two working steps of the workflow by comparing the changed input variable with the previously given input uncertainty and/or by predicting a change in the output values of each of the at least two working steps. Those changed output values which form input values of at least one of the at least two working steps are also compared to the previously determined input uncertainties. In those cases, where the changed input values do not exceed their corresponding previously given input uncertainty, the processing unit sends a message to the data output interface stating that the change does not affect these working steps, i.e. that they do not need to be re- performed.

The processing unit may further be arranged to evaluate whether and/or to which extent the propagation of the change through the at least two working steps results in at least one of the output values of the at least two working steps to exceed their corresponding output uncertainty which was calculated before the change. The result reflects an increased effort caused by the change since the extension of an output uncertainty means that all subsequent working steps have definitely to be re-performed. A reasoning engine may then decide whether the increased effort is acceptable with respect to predefined project constraints or not, i.e. whether the change request may be accepted or not. In those cases where the change in the at least one input variable is due to a changed request by the customer of the project and where the processing unit predicts that output values are definitely changing, a report about the implications of the change request may be generated automatically, either by the processing unit or by an external device connected via the data output interface. The changed request may for example result in a reduced amount of required material or equipment, where the material or equipment may have already been ordered based on previously planned numbers. The report may then reflect both the undesired surplus in material and equipment as well as the increased effort for the project execution. This report forms then the basis for further decisions to be taken together with the customer.

The data input interface may further be arranged to read in at least one limitation value which characterizes a possible or allowed range of at least one predefined output value and where the processing unit is arranged to check whether the output value including its corresponding output uncertainty falls within the range defined by the at least one limitation value or not. A limitation value may reflect a constraint on the automation and/or electrical engineering project or a possible fluctuation of the output value known from experience. The processing unit is therefore arranged to provide decision support on the quality of the output value by being arranged to check whether the limitation value is exceeded or not.

Even further, the processing unit may be arranged to combine the at least one output uncertainty with an external output uncertainty of another automation and/or electrical engineering project in order to calculate a number of hardware components which is required to be held in stock, where the number of hardware components is less than the sum of the at least one output uncertainty and the external output uncertainty.

For example, for three different projects it may have been determined that each project needs to install 10 ± 3 controller devices. Regarding each project separately, it may be advisable to have 3 controller devices on stock to save time in case that indeed all of the maximum 13 controller devices are needed. However, when all three projects are executed in parallel, it is not necessary to have a total of 9 controller devices on stock. Instead a lesser number, for example 5, may be sufficient and reduce the overall costs.

As described above, the processing unit is arranged to transmit its processing results to the data output interface from where they are transmitted to further devices units. One of these further units may be a reasoning engine which may derive information needed to take decisions with respect to the execution of the project. It is also possible to arrange the processing unit itself for determining, based on the at least one input and/or output uncertainty, the risk for a successful execution of the project. The data input interface may be connected to a user interface, which can be a visual, acoustic and/or haptic interface, where the user interface is arranged to query a user for inputting the at least one input uncertainty. In order to support a user in performing the at least one working step or in taking decision for subsequent working steps, the processing unit may be arranged to generate a visual and/or acoustic representation, such as a color code, a graphical marking or symbol, an animation, a sound etc., of at least one of its processing results. The invention and its embodiments will become apparent from the examples described below in connection with the appended drawings which illustrate:

Fig. 1 a system for executing working steps of a project,

Fig. 2 one working step with input and output values,

Fig. 3 an example of a project workflow,

Fig. 4 the working steps of Fig. 3 which are theoretically affected by a change in an external input value,

Fig. 5 the working steps of Fig. 3 which are practically affected by the change of Fig. 4 due to the invention,

Figs. 6a, b an example of a graphical representation of the exceeding of a limitation value, Fig. 7 the amplitude over time diagrams for rotational speed and torque of the motor of

Fig. 1 , at different points in time.

Figs. 1 and 3 were already described above with respect to the state of the art. An example of a working step 16 which is performed in accordance with the invention is shown in Fig. 2, where the task assigned to the working step 16 is the determination of the type and number of controller devices. Input values and corresponding input uncertainties 17 to 19 are the following:

• the number of information signals 17 to be transmitted is ns ± u_ns = 5000 ± 500, · the percentage of these information signal 18 which can be transmitted via one and the same signal transmission path, i.e. cable, is pt ± u_pt= 40 % ± 5 %,

• the desired spare capacity 19 for the handling of signals per controller is sc = 20%. Further, a technical condition or parameter 22 is given for the working step which states that up to nc=200 information signals can be handled per controller device of type xyz.

As described above, the input values and uncertainties 17 to 19 are read in via data interface unit 1. The same is true for the technical condition or parameter 22 which may have been read in at an earlier point in time as a predefined value for the working step 16. These data are processed by processing unit 2 in order to determine the output value 20 and 21 , the output values being the number nc of controller devices and their corresponding type xyz, where the output value nc is accompanied by a corresponding output uncertainty u_nc with an upper limit u_nc,up and a lower limit u_nc,low. The processing can for example be based

The resulting number of controllers is therefore estimated to be nc ± u_nc = 18 ± 3. In addition to what is shown in Fig. 2, the technical condition or parameter 22 may be given with an uncertainty as well, for example as nc=200±60, where this uncertainty is also taken into account by the processing unit 2 when generating the output uncertainty u_nc.

Fig. 4 shows the same workflow as already described above with respect to Fig. 3. When external input value A is changed from 12±4 to 10^ , all working steps receiving this input value directly, i.e. working steps 101 and 104, have to be re-performed when no

uncertainties were taken into account during performance of these working steps before the change. The same is true for all working steps which are subsequent to working steps 101 and 104, i.e. working steps 103, 106, 108, 109, 1 11 , 1 12 and 113. In total, nine working steps have to be re-performed.

As a result and advantageous effect of the invention, this number can be considerably reduced. As is indicated in Fig. 5 in comparison to Fig. 3, the output value of working step 101 is certain, i.e. 3, as long as the input value A does not exceed the previously considered upper limit of 16 and lower limit of 8. Since the changed input value A is still within this range, working step 101 does not need to be re-performed. And since the output value of working step 101 remains unchanged also the subsequent working steps 103, 106, 108 and 11 1 do not have to be executed again. This reduces the number of working steps to be repeated to 4, i.e. working steps 104, 109, 112 and 1 13, which means a considerable saving of project execution time. Figs. 6a and 6b show an example for how a result generated by processing unit 3 and transmitted via data output interface 2 to graphical display unit 14 can be visualized.

In Fig. 6a, two controller cabinets are shown where their number and contents correspond to output values of different working steps. The color or color pattern, respectively, of the cabinet indicates whether these output values fall within the range of corresponding predefined limit values, i.e. a predefined project constraint, or not. The cabinet to the left is white, indicating the all output values fall within the range, while the cabinet to the right is gray, indicating that at least one of the output values is getting close to its corresponding limit value. This gives an information to a user or decision maker that the project execution can still be continued but that it is getting close to a risk zone.

Fig. 6b shows the same two cabinets after a next working steps of the project was executed. The cabinet to the left turned gray, since at least one of the newly generated or updated output values is now being close to its limit value. This color change in color gives a helpful information to for example the engineer who supervised the performing of the last working step that something in the working step should be optimized. This could for example be parameters of the working step or the formulas used. If no improvement is possible there, the color change of the cabinet can be taken as an indicator for a possibly necessary request to change external input values of the project.

The cabinet to the right in Fig. 6b is visualized with a striped pattern, indicating that at least one of the limit values is exceeded by the newly generated or updated output values.

Accordingly, the project is at risk and a decision needs to be made for how to bring the project execution back into the safety zone. This decision can be taken automatically by a reasoning engine, or the reasoning engine could process all available project information in order to make a proposal to a human operator for how the project and in particular its external input values or its constraints, like time or costs, need to be changed in order to be able to guarantee success of the project again.

Fig. 7 differs from Fig. 6b only in that the cabinet to the right is illustrated in its opened state, showing two rows of control input/output boards (I/O boards). The left hand row is depicted with the striped pattern in order to clarify that this is the part of the cabinet where one or more limit values, for example the maximum allowed number of I/O boards, is exceeded. This simplifies the task for the decision making instance to find the project part where changes or adjustments need to be made in order to fall back into the limit value range.