A reference model based approach for the evaluation of industrial plant service and asset management tools.

Holm, Timo ; Luschmann, Christian ; Delgado, Antonio 等

1. INTRODUCTION

The most important goal for process industries is the return on investment (ROI). One way to raise this value is to reduce total cost of ownership (TCO) by efficient plant services including integrated maintenance management and execution, asset management and process optimization. This applies to operators of plants as well as to service providers. To achieve efficient plant service, supporting software systems are needed to handle the large number of special tasks involved, for instance outage optimization or knowledge based maintenance.

In contrast to plant operators and service providers the suppliers of SIS run a product business. They have to sell a product, which meets the domain specific challenges and needs of their customers. So for sustainable business success it is most important to know the requirements of customers, for example the special needs of the maintenance engineers who plan plant outages.

To be able to innovate SIS in a planned and systematic way in a first step the challenges posed by the industrial service business have to be identified and structured for application. In (Amberg et al. 2008) a concept for the characterization of SIS has been introduced. In this concept challenges and structures as well as a knowledge base of best practices suitable to address any single challenge are presented. The methodology was developed at the Department for Systems Engineering, Siemens AG (CT SE 5). It is based on research work conducted and project experience gained within the context of the industrial solution, service and plant engineering business at Siemens AG. Supportive research work was done by the Department of Information Systems 3 (WI3) at Friedrich-Alexander-University Erlangen-Nuremberg (FAU). An evaluation of a large integrated plant asset management (PAM) and field device management system (FDM) was prototypically executed by the Department of Chemical- and Bioengineering--Chair of Fluid Mechanics (LSTM), Bioprocess Automation Research Group--at FAU to verify the practical applicability.

[FIGURE 1 OMITTED]

2. CHALLENGES IN INDUSTRIAL SERVICE BUSINESS

2.1 Three Types of Challenges

Along the life cycle phases of a plant three different types of challenges can be identified: Engineering Challenges (based on Lowen et al., 2005), Project Challenges (Svensson et al., 1999) and Service Challenges (Amberg et al., 2008).

In engineering many types of risks and technical dependencies have to be managed in several disciplines. Part of service engineering for instance is the parameterization of SIS for PAM as well as process monitoring and optimization systems.

Project challenges address issues based on concurrent and cross project engineering, for example change management, multiple language support and user management.

Service Challenges are derived from service execution, for example maintenance, which includes inspection, monitoring, compliance test, function checkout, routine maintenance, overhaul, rebuilding, repair, temporary repair, fault diagnosis, fault localization, improvement, and modification (EN 13306, 2001). In (Ehrenspiel et al., 2007) maintenance for industrial plants is defined as "all technical measures to obtain or restore the functional state" of an industrial plant. Challenges of service execution result mainly from the existence of a plant, for instance the need to respond to urgent malfunctions and the plant itself as data source of SIS (for example supported by asset management systems in practice).

2.2 Structuring Challenges

Every challenge of industrial service business is described in detail by a number of so called subchallenges, which describe a single determinant of the efficiency of supporting SIS regarding the industrial service business. For every determinant described by one subchallenge five different best practices, ordered from Class 0 to 4, have been gathered (example and meaning of classes see Tab. 1), which reflect the corresponding determinant within the SIS (meta-model see Fig. 1). For every challenge, the arithmetic mean of the performed classifications is calculated and finally visualized in a kiviat graph, forming a characteristic evaluation footprint of the SIS (Fig. 3).

2.3 Service Challenges

Service challenges have been indentified analyzing multiple SIS, a number of expert workshops (see Amberg, 2008) as well as know-how gathered in several projects by CT SE 5.

During evaluation, the mapping of concepts implemented in SIS to challenges is a complicated step which is up till now done intuitively producing subjective results. Therefore a generic architecture of SIS is introduced as guideline for the evaluator, contributing to the formalization of the evaluation process by providing a mapping-oriented reference model (Fig. 2).

The reference structure of a SIS and the mapping to the challenges are derived from the plant service process. The model shows all relevant layers of an SIS from processing of field data from the plant up to possibilities to connect to ERP systems. For an efficient service execution there is a need to store, process and change large amounts of datasets coming from the plant. These challenges are grouped via the service challenge data handling. The challenge data processing deals with this data and characterizes e.g. the level of automation that is provided by the SIS during further processing and data consistency checks that might be executed.

This data has to be pre-processed appropriately before it can be presented to the user. Subchallenges addressing the visualization of service information, are subsumed by the service challenge view concept. The service challenge service know-how reuse deals with the necessity to formalize and reuse service know-how (Fig. 2--right).

2.4 Problem of Mapping Challenges to SIS Components

Especially if large integrated systems are evaluated, the problem occurs, that evaluators can't decide, which component of the SIS has to be assessed, solely based on the description of the challenges. This can be illustrated by the following example, taken from the evaluation of an integrated PAM and FDM system which was executed in cooperation by WI3 and LSTM. In this particular example different results would be reached by two possible component candidates.

The evaluated SIS, a PAM system, is embedded in a process control system and therefore has direct access to all field-devices. The PAM system uses, amongst other things, the process control system's integrated view component. The FDM system in this case reads the diagnostic data directly from the field-devices via a fieldbus or HART and acts as a gateway; simply writing data into a placeholder box surrounded by pictures defined in the SIS view component.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

In case of the subchallenge customization, which is assigned to the challenge view concept, either Class 0 or 4 can be matched (best practices see Tab. 1). Class 0 is applicable since the FDM system itself has no user definable views. The diagnostic values from a field device are always read completely by the gateway and can't be filtered. On the other hand Class 4 can be matched, because the powerful view component integrated into the PAM tool allows free engineering of views and also the integration of every type of data.

The systematic solution of this problem is the reference model (Fig. 2.), which maps generic components of SIS to challenges. In our example the FDM gateway belongs to the component Conditioning, which is evaluated by the challenge data processing. The considered subchallenge customization belongs to the challenge view concept matched to the component View of the reference model. Therefore the View component of the PAM system should be evaluated, eventually leading to a Class 4 rating for this subchallenge.

As shown in this example, the challenges combined with the reference model offer an abstract structure used to evaluate tools in a more objective and easy-to-use way.

3. SUMMARY AND OUTLOOK

The results of a SIS evaluation (Fig. 3) can help tool suppliers to specifically identify and evaluate requirements of future versions to match the needs of service engineers and personal to get the capabilities to reduce total costs of plant ownership by service optimization. The evaluation method based on a reference model approach leads to more objectified results by formalizing the evaluation process making design decisions more secure and riskless for SIS developers. It is an easy-to-use practical tool for product managers and service engineers, which closes the gap between plant service business and the product driven business of the service system suppliers.

The formulation and application of the developed methodology is currently being extended to other life cycle phases.

4. REFERENCES

Amberg, M.; Holm, T.; Maier, R. & Maurmaier, M. (2008). Evaluation von industriellen Service-Werkzeugen Proceedings of 10th Symposium on Entwurf komplexer Automatisierungssysteme (EKA 2008)

Ehrenspiel, K.; Kiewert, A. & Lindemann, U. (2007). Cost efficient design, Springer Verlag, Berlin

EN 13306 (2001) European Standard: Maintenance Terminology

Lowen, U.; Bertsch, R.; Bohm, B.; Prummer, S. & Tetzner, T. (2005) Systematization of industrial plant engineering, atp --Automatisierungstechnische Praxis, Vol 50, I. 4, pp 54-61

Svensson, D.; Malmstrom, J.; Pikosz, P. & Malmquvist, J. (1999) A Framework for Medelling and Analysis of Engineering Information Management Systems, In: Proceedings of ASME Design Engineering Technical Conference, Las Vegas

HOLM, T[imo]; LUSCHMANN, C[hristian]; DELGADO, A[ntonio] & AMBERG, M[ichael] *

* Supervisor, Mentor

Tab. 1. Structure of Subchallenge Customization.

 General Meaning Customization--Possibilities
 of Class for Views on Service Data

Class 4 Generic Support Definition of user-specific
 views with integration of data

Class 3 Explicit Support Definition of user-specific
 (High Level) views (more than spreadsheet)
 without integration of data

Class 2 Explicit Support Definition of user-specific
 (Low Level) spreadsheet views (e.g. queries)

Class 1 Implicit Support Limitation of directly supported
 views for specific users

Class 0 No Concept No support