EP2633374A1

EP2633374A1 - Maintenance information apparatus, state sensor for use therein, and method which can be performed thereby for making a decision for or against maintenance

Info

Publication number: EP2633374A1
Application number: EP11776134.6A
Authority: EP
Inventors: Gerrit Schramm; Arnd Schirrmann; Klaus Zoller
Original assignee: EADS Deutschland GmbH
Current assignee: Airbus Defence and Space GmbH
Priority date: 2010-10-28
Filing date: 2011-10-11
Publication date: 2013-09-04
Also published as: DE102010049909A1; US9477222B2; US20140365178A1; WO2012055699A1

Abstract

The invention relates to the area of state-based maintenance where a state of a technical system (12) is monitored by a sensor (16). In order to allow the use of relatively simple sensors (16) with relatively simple evaluation logic units, the invention proposes a method and an apparatus for assisting decisions on the basis of hyperordinal signals from nonperfect hyperordinal state sensors (34).

Description

Maintenance information device, condition sensor for use therein, and method of decision making for or against maintenance that can be performed thereby

The invention relates to an apparatus and a method for supporting a decision as to whether or not maintenance or servicing of a system whose condition is monitored by means of a condition sensor is to be performed. Moreover, the invention relates to a condition sensor usable in such a device or such a method.

The invention is in the field of condition-based maintenance (CBM). In short, CBM means that maintenance or servicing is then carried out when the need arises. Maintenance or servicing is performed when one or more indicators indicate that the monitored equipment or system will fail, or that operating characteristics, technical characteristics, functions of the monitored equipment or the monitored technical system will deteriorate.

For the purpose of describing the invention and its advantageous embodiments, the terms maintenance and servicing are used synonymously. They form a generic term for any measure that is useful for maintaining or improving the operational readiness of the monitored technical system, such as replacement of components or wear parts; Cleaning of contaminants, replacement of operating fluids, lubrication; Replacing or cleaning filters; Removal of waste products; Repair of damaged areas etc. Condition-based maintenance has been introduced to try to maintain or maintain the correct equipment at the right time. CBM is based on the use of real-time data to prioritize and optimize maintenance resources. Monitoring the state of the system is known as "state monitoring." A condition monitoring device will detect the "health" or "health" of the monitored equipment and act only when maintenance is actually necessary has resulted in extensive metrology of technical systems and equipment, and this, coupled with improved tools for analyzing condition data, has made it more than ever before that maintenance personnel are capable of deciding the right time to perform maintenance Ideally, condition-based maintenance allows maintenance personnel to do just the right and necessary things, minimizing parts replacement costs and system downtime during maintenance and maintenance time.

There are some literature on condition-based maintenance that represents the current state of the art in this field and explains the terms, measures, and devices used. These include:

• Jarrell DB, Sisk DR & Bond LJ: "Prognostics and Condition-based Maintenance (CBM) - a Scientific Crystal Ball", published by the Pacific Northwest National Laboratory, Richland, Washington, No. PNL-SA-36771 2002 with further references;

· International Atomic Energy Agency IAEA's "Implementation Strategies and Tools for Condition Based Maintenance at Nuclear Power Plants, IAEA-TECDOC-1551, May 2007, with further references;

• Memorandum for Secretaries of the Military Departments, by Deputy Under Secretary of Defense, USA, November 25, 2002: "Policy for Department of

Defense condition-based maintenance ". It is apparent from the above publications, inter alia via the Internet, that different state sensors can be used to monitor quantities that can indicate load or wear of a component of a technical system. For example, vibrations, material expansions, temperatures, viscosities or other properties of operating fluids, etc. can be monitored.

Examples from the patent literature on condition-based maintenance can be found in DE 103 32 629 A2, DE 101 44 076 A1, EP 08 95 197 B1, DE 31 10 774 A1, DE 10 2005 012 901 B4, DE 102 22 187 A1, DE 101 48 214 C2.

The respective monitored technical systems can be quite different; Condition-based maintenance has advantages in many technical areas. Examples can be found in the above references and patent documents. The condition-based maintenance is particularly interesting in the field of aircraft technology, as it is particularly maintenance-intensive due to the safety relevance of the technical systems used. In particular, there is a need to use condition-based maintenance on all possible parts of an aircraft that require maintenance.

However, this is a long-term goal, which until now can not be realized because of the weaknesses of the previously known devices and methods for condition monitoring and maintenance.

In aircraft technology, there is also the problem of an expected increase in maintenance work, as more and more new materials, such as power flow-compatible fiber composite materials, are used in aircraft construction to save weight. However, in the case of fiber-reinforced composite materials in particular, wear and tear and invisible internal damage may occur under heavy load, so that an exchange of components made of fiber-reinforced Composite materials according to specific operating scopes and load levels before these components fail. It would therefore be desirable to be able to wait as many state-based as possible for such components. However, the main disadvantage of CBM is that the initial cost of CBM is very high. CBM needs improved metrology for the monitored equipment. In most cases, the costs of adequate measurement technology are very high; This is especially true for equipment that has already been installed. But also in the new development of technical equipment and systems with the previously known state sensors and evaluation techniques result in high additional costs, so that condition-based maintenance could previously only be realized on a few main parts of technical systems.

The invention is based on the object to obtain a decision support for maintenance planning based on signals from sensors that can provide information about a state of health of the monitored system, but still easier and cheaper to manufacture and integration into a measurement technology than before Condition sensors used at CBM. There is literature on condition monitoring (CBM) with perfect sensors. However, condition monitoring with perfect sensors is only conditionally applicable to real systems. In order to enable condition-based maintenance (CBM) of a system, the system must be equipped with sensors that are able to determine its state of health as far as possible in real time (see DIN 44300). In classical CBM implementations, the sensors determine an expected remaining life (RUL) and, if appropriate, a confidence level. The calculation of the RUL is very complex depending on the application and drives the implementation costs.

It is known, for example from the aforementioned literature and patent literature, that through the development of CBM sensors and above all the algo- These algorithms calculate the remaining lifetime from high physical data of the sensors - high to prohibitive costs. This is due to the fact that the sensor tries at any time between commissioning and failure to determine the most accurate residual life possible. The designer of the sensor is not aware of how good the calculation of the remaining lifetime at certain intervals really must be, so that he is given no opportunity for optimization. So far, there is no product or literature on decision support with non-perfect sensors. The invention seeks to provide discrimination support based on imperfect sensors.

This object is achieved by a maintenance information device having the features of claim 1, a condition sensor having the features of claim 9 and a method according to claim 10.

Advantageous embodiments are the subject of the dependent claims.

According to a preferred embodiment, the invention provides an advantageous algorithm for decision-making with which the technical problem presented above can be solved.

Decision support for maintenance planning with non-perfect sensors allows a better prediction of the future wear behavior of condition-monitored components or systems. The improved forecast can be used for a better decision support algorithm.

In one aspect, the invention provides a maintenance information device for providing condition-based maintenance information of a technical system, comprising:

a, preferably ordinal, sensor with a sensor element for monitoring at least one load or wear size of the technical system tems as an indicator of a maintenance-relevant state of the system, wherein the sensor is designed to provide a, preferably ordinal, state signal,

an operation amount detecting means that detects an accumulated amount of operation of the system as a value of an operation amount parameter, expectation delivery means for providing a deterioration function representing the expected deterioration characteristic of the condition detectable by the load or wear amount depending on the operation amount parameter;

a residual life determining device, the occurrence

a) at least one predetermined degree of deterioration of the state compares the achieved actual value of the operating volume parameter with the value of the operating volume parameter expected for this predetermined degree of deterioration according to the expected deterioration curve

and or

b) at least one predetermined value of the operating volume parameter then compares the current degree of deterioration of the condition detected by the sensor with the value of the operating volume parameter expected in this degree of deterioration in accordance with the expected deterioration profile, and

based on the current value in the case a) of the operating volume parameter and in case b) of the degree of deterioration provides a recalibrated expected deterioration of the state over the operating volume parameter, and a maintenance recommendation generating means for generating a maintenance recommendation for a future planning period based on the recalibrated expected deterioration course.

According to a further aspect, the invention provides a hyperordinal status signal for monitoring the status of at least one component of a technical system for determining a recommendation as to whether maintenance is indicated within a specific planning period or not. According to a further aspect, the invention provides a hyperordinal condition sensor for monitoring a status of at least one component of a technical system for determining a recommendation as to whether maintenance is indicated within a certain planning period or not.

In another aspect, the invention provides a method for deciding whether to maintain a technical system, comprising the steps of:

a) monitoring the system with a sensor element and supplying a status signal and the present value of a current operation parameter indicating the extent of operation of the system when the status signal is detected,

b) determining an expected degradation history for the condition representing an expected degradation of the condition over the operating environment parameter from experience and / or simulation values for the system,

c) determining a residual life expectancy distribution from the state signal supplied by the sensor element and the expected degradation function,

(d) Classification of residual life expectancy into one of several given classes and

e) Generating a maintenance recommendation based on the class received. Preferably, the invention provides discrimination support with data from imperfect hyperordinal sensors.

Hyperordinated sensors provide the ability to blur the state of the system. A sensor is hyperordinal when it classifies the state of health of the system it is monitoring. For example, the following classes can be used:

• Green: nominal operating state • Yellow: measurable wear but no danger of system failure

• Red: shortly before system failure

• Black: System Failure While there are no perfect sensors in reality, classical sensors are more elaborate compared to the hyperordinal sensors proposed according to a preferred embodiment of the invention and are associated with correspondingly higher development and operating costs. A hyperordinal sensor generates statements with significantly reduced entropy, but remains largely precise in the key statements.

In one aspect, the invention provides "decision support with data from imperfect hyperordinal sensors," which is an algorithm that enables reliable residual lifetimes to be determined from economic hyperordinal state information.

Such an algorithm allows a reliable determination of the remaining life based on very fuzzy statements made about the health of a system.

Embodiments of the invention are explained in more detail below with reference to the accompanying drawings. Showing:

1 shows a schematic block diagram of a maintenance information device and a method for decision-making for or against a maintenance of a monitored technical system that can be carried out therewith;

2 shows a graph with an example of a calculation of a degradation model (degradation model) on the basis of a first exemplary case; and Fig. 3 is a graph comparable to that of Fig. 2 with an example for calculating a deterioration model based on another example case. FIG. 1 schematically shows a maintenance information device 10 and a flow chart for its function in the form of an algorithm shown as a block diagram. The maintenance information device 10 serves to provide decision-making information as to whether a state-monitored technical system 12 should be maintained or not.

The engineering system 12 may be any equipment or component that is to be maintained. Examples of such technical systems can be found in the literature and patent literature on status-based maintenance cited above. In a preferred embodiment, the monitored technical system 12 is a component 14 or equipment of an aircraft. Examples of such components are structural elements of the fuselage or of wings or of tail units, engines or engine parts, air conditioning systems, cooling systems, life support systems, rescue systems, landing gear and chassis parts, etc.

The maintenance information device 10 includes a sensor 16, an operating circumference detection device 18, an expected value delivery device 20, a remaining life determination device 22, and a maintenance response determination device 24.

The sensor 16 is a health condition detection means 25 which detects a health condition of the technical system 12.

The sensor 16 in the illustrated example is configured to detect a size or operating parameter of the technical system 12 that may serve as an indicator of a condition of the system in which maintenance or servicing would be required. The sensor 16 is a non-perfect sensor in the illustrated example, which performs ordinal state measurement as ordinal sensor. The states are preferably scaled on an ordinal scale, so that a degree of deterioration of the state can be determined. The degree of deterioration can range, for example, from the state "NEW VALUE" (eg, as new as new standard state / completely unloaded, unworn state) to the state "SYSTEM FAILURE".

The sensor 16 has for this purpose a sensor element 26 which is designed to monitor and detect at least one load or wear size of the technical system, which is a measure of a load or wear of at least a portion of the system 12 and thus an indicator of a state of health can. For example, static or dynamic loads, strains, vibrations, temperatures, viscosities or other properties of operating fluids, etc. are measured.

Some examples of load or wear parameters to be monitored can be found in the literature and patent literature on CBM mentioned in the introduction. Examples of sensor elements 26 that can also be used for condition monitoring in the maintenance information device 10 described here are described and shown in the following documents, the contents of which are hereby incorporated by reference: WO2009 / 062635A1, WO2009 / 087164A1,

WO2009 / 071602A2, DE10004384C2, DE10053309A1, DE10153151A1,

DE10236051 A1, DE102006060138A1, DE102008017175A1, as well as the unpublished DE102010032093.5-22. It is expressly made to the cited references for more details on the structure of possible sensor elements 26.

The operation amount detection means 18 detects the accumulated amount of operation of the monitored technical system 12 and supplies the measurement as a value of an operation amount parameter. In most cases, the operating time of the monitored system 12 is simply detected as the operating parameter. In the case of vehicles or the like could be given as an operating parameter and the total with the vehicle or one of its components (eg motor) covered route, since the distance covered in such vehicles can have a greater impact on wear and load than the actual operating time. In technical systems where deterioration could occur independently of actual operation, eg due to material fatigue due to aging phenomena or environmental influences such as wind, weather, seawater, the cumulative time could be measured, or just the time that the component to be monitored could be measured Exposed to environmental influences. All of these are examples of possible operating parameters; and, depending on the application, a suitable operating parameter detection device 18 is provided, such as a clock or other time measuring device for measuring operating time or for measuring other time durations relevant to the state of health, or a distance counter such as an odometer, etc.

Hereinafter, for ease of illustration, the maintenance information device 10 will be explained using the example that the operation time of the monitored technical system 12 is detected for the operation amount parameter.

In the example shown, the operating circumference detection device 18 is simply a part of the sensor 16. In this example, in addition to the sensor element 26 for detecting the load or operating variable, the sensor 16 also has the operating circumference detection device 18 for detecting the cumulative value of the operating circumference parameter, viz here the accumulated operating time, up.

Due to the detection of the at least one load or wear quantity and the value of the operating circumference parameter present at the time of detection, the sensor 16 forms an ordinal status signal 27, which will be explained in more detail below with reference to FIGS. 2 and 3. As illustrated in FIGS. 2 and 3 with reference to exemplary curves S for the ordinal state signal 27, in the example shown here, the sensor 16 supplies, as an ordinal state signal S, a degree of deterioration j determined on the basis of the detected value of the amount of load or wear depending on the operating scope parameter - here the operating time s-hrs in operating hours. It suffices in the procedure described here for the actual degree of degradation j detected by the sensor 16 to be stated only relative or ordinal. For example, the degree of degradation is normalized in the graphs of FIGS. 2 and 3, where 100 as 100% degradation indicates the system failure and 0 as 0% degradation indicates the mint condition.

Referring again to FIG. 1, the expected value delivery facility 20 includes a system database 28 that accumulates based on experience with legacy systems that are comparable to the technical system 12 and / or on the basis of simulation data Degradation paths included. In particular, the database includes experience and / or simulation values that include expected states per value of the operating scope parameter. For example, the system database contains expected values for state / operating hours or state / operating cycles. As a concrete example, e.g. expected function values for the load or wear parameter as a function of the operating parameter specified. Furthermore, the system database 28 receives newer values by receiving the status signal 27 from the sensor 16, which can additionally be used for the later determination of expected values.

The expected value delivery device 20 is designed to calculate from the data of the system database 28 an accumulated degradation model 30 of the technical system 2. Such an accumulated degradation model 30 may be represented as an expected degradation history of the condition monitored by the sensor 16. Examples of such accumulated degradation models can be found in the respective curve M in the graphs of FIGS. 2 and 3 for two concrete example cases (Case A in Fig. 2 and Case B for Fig. 3) of a system monitoring.

Referring again to FIG. 1, the remainder of the life determining means 22 is connected to the ordinal sensor 16 for receiving the ordinal state signal 27 and is connected to the expected value delivery means 20 for obtaining accumulated degradation models 30. The remaining life determination device 22 calculates a residual life distribution from accumulated degradation models 30 of the system database 28 and the measured ordinal state signal 27 of the technical system 12. From the calculated residual life distribution, a hyper-ordinal state signal 32 is then generated which indicates a hyper-ordinal system state.

Thus, a hyper-ordinal state sensor 34 is formed. In the example shown here, the latter has the ordinal sensor 16, the expected value delivery device 20 and the remaining life determination device 22. Here, sensors are named as hyperordinal, which classify health statuses of the systems they monitor. For example, the following classes are used:

»GREEN: nominal operating condition

»YELLOW: measurable wear but no danger of system failure

• RED: imminent system failure

• BLACK: System Failure Examples of classification, calculation of residual life distribution, and a prediction of the remaining lifetime thereof will be explained in more detail below with reference to Example Cases A and B shown in Figs. 2 and 3, so that the nature of hyperbaric ordinal condition sensor 34 becomes more understandable.

As shown in FIG. 1, the maintenance recommendation generating device 24 determines from the hyper-ordinal status signal 32 a maintenance response. Recommendation 36 according to one of the preceding classifications. Thereby, a decision support function for maintenance actions on the system 12 is obtained. In particular, it is interesting here whether maintenance on a planning scale is required.

Aircraft are used, for example, in short-distance, medium-haul or long-haul operation. For example, it is interesting to know whether a maintenance would be required within a planned short-haul, middle-distance or long-haul, which would therefore be carried out beforehand. It may be that maintenance for a short-haul route is not yet required, but a need for long-haul maintenance is expected. Before the long haul would you perform the maintenance thus, before the short haul not yet. It may also be that you should schedule a maintenance at the destination of the home airport and should carry appropriate material.

The maintenance recommendations 36 could therefore be one of the following, depending on the planning horizon considered:

• Immediate repair

NO-GO

»Schedule replacement during the planning period (for example, take a replacement part so that it may be exchanged at the destination before a return flight);

• no action (everything in the GREEN area). More than the output of such classified maintenance recommendations is not provided in the embodiment of the device 10 and the method that can be carried out here.

The maintenance personnel will then follow the repair recommendation and plan and complete the maintenance execution 40. In the following, the classification and estimation of the residual service life will be explained in more detail with reference to two exemplary cases A and B, which are reproduced in FIGS. 2 and 3. These figures represent graphs depicting the degree of deterioration j of the monitored condition over the operating range parameter, in our example the service hours [s-hrs].

As already mentioned above, the curves S represent the measured ordinal status signals in the form of a function of the degree of deterioration depending on the value of the operating volume parameter at which such deterioration has been measured ngs-grad. The curves M1 and M2 represent both cases the accumulated degradation models 30 provided by the expected value delivery means 20 as functions of the expected deterioration values j as a function of the values of the operating volume parameter where these deterioration values j are expected.

Based on the deterioration values j, a classification is provided. In the illustrated examples, it is defined that sensor signals indicative of deterioration values j between 0 and 30 are classified into a class A. If sensor signals are within this class A, the health status is in the green range, no action or maintenance is required.

Sensor signals indicating measurable wear but where there is still no risk of system failure are classified in Class B. In the case of example given in FIGS. 2 and 3, this ranges from the degree of deterioration j = 30 to the degree of deterioration j = 60.

Another class C, indicating that a system failure is imminent, ranges from the degree of degradation j = 60 to the degree of degradation j = 90 in the examples given. Sensor signals indicating a degree of degradation of j> 90 are ranked N-NO in the example. Continued operation without prior maintenance is prohibited when this class occurs. In particular, it is interesting to estimate the number of operating hours s-hrs to reach NO-GO class N. For example, with the aircraft it would be important to know: can I safely fly a short, medium or long haul? Let us now look at the example in case A reproduced in FIG. The accumulated degradation model 30 initially indicates, according to the curve M, a value of about 1900 s-hrs operating hours until the NO-GO class is reached. A sensor signal transition from class A to class B is assumed to be at or about the degree of degradation j = 30. However, the number of hours of operation that accumulated until reaching this second class B is only about 1300 s-hrs, while according to the degradation model 30 about 1600 s-hrs have been estimated until reaching class B. This gives rise to a drastic revision 1 of the initially estimated degradation model 30. Revision 1 is indicated as a left shift P1 of the remaining degradation path of the degradation model 30. The result is a first-time revised degradation model according to revision 1 as indicated by the curve R1.

A second revision 2 takes place at the degree of deterioration j = 60 if the condition for class C is fulfilled. Here, the number of operating hours attained for j = 60 by the measured state signal 27 is slightly larger than the number of operating hours estimated by the revised expected deterioration function - curve R1. A (slight) right shift P2 of the remaining degradation model path indicates a (moderate) increase in the useful life of the monitored element. The correspondingly revised The remaining degradation degradation path is represented as Revision 2 by the curve R2.

Let us now consider case B, as shown in FIG. Here it is assumed that a transition from class A to class B at the degree of deterioration j = 30 in the case of the measured status signal 27 as indicated by curve S is reached only after about 2400 s-hrs of operating hours. After the degradation path M according to the initial accumulated degradation model 30, such a transition would have already been expected at approximately 1600 operating hours. This gives rise to a drastic revision 1 of the initially estimated degradation model 30 for estimating the operating hours remaining until the NO-GO class N is reached. This is indicated by a right shift P1, resulting in the remaining revised degradation model path according to revision 1, indicated as curve R1. Another revision 2 occurs when the condition of class C has been reached. Again, measured state signal and degradation model are reconciled with respect to the accumulated number of operating hours. A slight right shift P2 of the remaining degradation model path indicates a further (moderate) extension of the useful life of the monitored unit. This is illustrated in the graph of FIG. 3 by the remaining curve R2 for revision 2.

The examples illustrate that it suffices only to recalibrate the expected degradation history of the state at some particular values of the state signal.

It is also sufficient if the status signal is placed in classes and the class is specified. As long as the class A is present, the hyper-ordinal state sensor 34 can indicate the signal GREEN as the hyper-ordinal status signal for the remaining time remaining.

For example, it is not until a certain threshold for the state of health has been reached, for example, when a sensor element 26 has exceeded it. woke load or wear size reaches a certain threshold, changed from a class A to a class B. The then achieved value of the operating volume parameter is used to recalibrate the expected deterioration process. With this, the distribution of the remaining remaining service life can be re-estimated and thus a corresponding service recommendation class (eg still GREEN or already YELLOW), depending on the planning period of interest, can be generated.

It is therefore sufficient to use non-perfect, ordinal sensors for condition monitoring and to equip them with a simple evaluation logic that provides simple classified signals for the states and the recommendations (only A, B, C or N or GREEN, YELLOW, RED or BLACK) output.

It is particularly interesting that due to the simple classification of state signals which are only ordinally arranged, it is possible to easily recalibrate the estimation of the remaining service life based on the information about the scope of operation of a class transition.

Deviations from the illustrated embodiment are obvious. For example, instead of recalibrating the residual life estimation model, it would be conceivable not to reach a certain transition for the load or wear quantity but to achieve a certain value of the operating parameter. The model 30 of FIG. 2 or 3 is in a different, not shown embodiment, not revised at a certain degree of deterioration j, but on reaching a certain number of operating hours, for example, every 500 hours of operation, based on the then present state signal. However, since the status signal provides a better basis for reliable information about an imminent warning, the variant illustrated in FIGS. 2 and 3 is preferred. LIST OF REFERENCE NUMBERS

10 Maintenance information device

12 technical system

14 component of an aircraft

16 sensor

18 operation scope detection device

20 expected value delivery device

22 remaining life determination device

24 maintenance recommendation generating device

25 health condition detection device

26 sensor element

27 ordinal state signal

28 system database

30 accumulated degradation model

32 hyper-ordinal state signal

34 hyper-ordinal state sensor

36 Maintenance recommendation

40 maintenance execution

S Curve for measured condition signal

M curve for initial accumulated degradation model

P1 first revision 1

P2 second revision 2

R1 curve for remaining course of the accumulated degradation model revised by revision 1

R2 Curve for remaining course of the accumulated degradation model revised by Revision 2

Claims

Claims 1. A maintenance information device (10) for providing condition-based maintenance information of a technical system (12), comprising:

a, preferably ordinal, sensor (16) with a sensor element (26) for monitoring at least one load or wear size of the technical system (12) as an indicator of a maintenance-relevant state of the system (12), wherein the sensor (16) for delivering a , preferably ordinal, state signals (27) is formed,

an operation amount detecting means (18) for detecting an accumulated amount of operation of the system (18) as a value of an operation amount parameter;

expected value delivery means (20) for providing a deterioration function representing the expected deterioration history of the condition detectable by the load or wear amount depending on the operation amount parameter,

a remaining life determination device (22) which, when it occurs

and or

b) at least one predetermined value of the operating volume parameter then compares the current degree of deterioration of the condition detected by the sensor (16) with the value of the operating volume parameter expected in this degree of deterioration according to the expected deterioration characteristic and based on the current value in case a) of the operating volume parameter and in case b) of the degree of degradation provides a recalibrated expected deterioration of the condition over the operating range parameter, and a maintenance recommendation generating means (24) for generating a maintenance recommendation (36) for a future planning period from the recalibrated expected deterioration history ,

2. Maintenance information device (10) according to claim 1,

characterized,

that it classifies the current state of health of the system (12) to be monitored into several classes.

3. maintenance information device (10) according to claim 2,

characterized,

that it classifies the state of health into at least two classes selected from a group of classes which

A nominal operating state,

«Measurable wear but no risk of failure of the monitored system (12),

· Imminent system failure and / or

β system failure

contains.

4. Maintenance information device (10) according to one of claims 2 or 3,

characterized,

in that the remaining life determination device (22) is designed such that it recalibrates the expected deterioration course in the case of an transition between two classes.

5. Maintenance information device (10) according to one of the preceding claims, characterized,

in that the expected value delivery device (20) is designed to set up an expected deterioration curve (M) from a database (28) with empirical and / or simulation values for the system (12) as a model for the deterioration.

6. maintenance information device (10) according to claim 5,

characterized,

the degradation profile of the load and / or wear quantity currently detected by the sensor (16) is entered into the database (28) via the operating volume parameter as further empirical values.

7. maintenance information device (10) according to one of the preceding claims,

characterized,

in that the remaining life determination device (22) is designed to determine a remaining life distribution from accumulated deterioration models, which are determined from a system database (28), and the measured ordinal status signal (27) of the system (12).

8. maintenance information device (10) according to claim 7,

characterized,

in that the residual life determination device (22) is designed to supply a hyperordinal system state signal.

A hyperordinal condition sensor (34) for monitoring a condition of at least one component of a technical system for determining a recommendation as to whether maintenance is indicated or not within a particular planning period.

10. Method for deciding whether to perform a maintenance of a technical system, comprising the steps of: a) monitoring the system (12) with a sensor element (26) and supplying a status signal (27) and the current value of an operating parameter specified on the extent of the operation of the system (12) when the status signal (27) is detected;

b) determining an expected degradation history for the condition representing an expected deterioration of the condition over the operating volume parameter from experience and / or simulation values for the system (12),

c) determining a remaining life expectancy distribution from the state signal (27) supplied by the sensor element (26) and the expected degradation function,

e) Generating a maintenance recommendation based on the class received.

11. The method according to claim 10,

characterized,

in that step a) contains at least one of the following steps:

a1) using a non-perfect ordinal sensor (16) for detecting load and / or wear magnitude indicative of health status of the monitored system (12);

a2) detecting at least one current value of the operating parameter at the time of measuring a predetermined value of the load and / or wear quantity; and or

a3) detecting at least one current value of the load and / or wear quantity upon occurrence of a predetermined value of the operating volume parameter;

and / or that step b) contains at least one of the following steps:

b1) querying a system database (28) with empirical or simulation values for which values of the operating volume parameter which degrees of deterioration are to be expected for the condition, and / or b2) creating an accumulated degradation model (30) for the system (12) from empirical or simulation values.

12. The method according to any one of claims 10 or 11,

characterized,

in that step c) contains at least one of the following steps:

c1) upon detecting at least one predetermined degree of deterioration of the condition monitored by the sensor (16): comparing the current value of the operating scope parameter reached in this detection with the value expected for that predetermined degree of deterioration according to the expected deterioration characteristic;

and or

c2) upon reaching at least one predetermined value of the operating scope parameter: detecting the actual degree of deterioration then present by means of the sensor (16) and comparing it with the value of the operating peripheral parameter expected in this degree of deterioration according to the expected deterioration course;

and

c3) recalibrating the expected deterioration profile over the operating parameter due to the value of the operating scope parameter achieved in step c1) and / or due to the degree of deterioration detected in step c2);

and

c4) determining the residual life distribution based on the expected deterioration course recalibrated in step c3).

13. The method according to claim 12,

characterized,

that step a) includes the step:

a4) classifying the measured state signal (27) into one of several classes and specifying the corresponding class; and that step c1) is performed on transition from one class to another class.

14. The method according to claim 13,

characterized,

in that step a4) contains at least the step:

a1a) Classifying the measured state signal, in particular the detected value of the load or wear quantity, into one of at least two classes selected from a group of classes which

A nominal operating state,

• measurable wear, but no risk of failure of the monitored

Systems,

• imminent system failure and / or

• System failure

contains.

15. The method according to any one of claims 10 to 14,

characterized,

in that step d) contains at least the step:

d1) classifying the residual lifetime distribution determined in step c) into one of at least two classes selected from a group of classes which

A nominal operating state,

Measurable wear but no risk of failure of the monitored system,

• imminent system failure and / or

• System failure

contains.