1
DIAGNOSTIC NETWORK WITH
AUTOMATED PROACTIVE LOCAL
EXPERTS
CROSS-REFERENCE TO RELATED 5
APPLICATION
This application claims priority back to Provisional Application Ser. No. 60/134,982, filed May 20, 1999 now expired, the entire disclosure of which is incorporated herein by reference.
TECHNICAL FIELD
The present invention relates generally to predictive maintenance, and more particularly to a diagnostic network is and method which employs local experts to automatically monitor, diagnose and take action in connection with different machines included within a system.
BACKGROUND OF THE INVENTION 20
Vibration is produced by moving parts of machinery due to causes like unbalance, misalignment of shafts and worn out bearings. Every machine has reasonable levels of vibration that are characteristic of its operation. Excessive levels of vibration indicate machine problems which may lead to 25 unsatisfactory operation and eventual breakdown.
Machine malfunction can sometimes be detected by a change in the vibration pattern of that machine. In today's plants where machines are complex and/or large in number, 3Q engineers rely on more advanced methods of identifying abnormal levels and patterns of vibration in order to determine the condition of a machine. It is now common to measure the level and rate of the vibrations of machines and use this data to determine which machines require servicing. 3J This monitoring technique is called predictive maintenance.
To properly employ predictive maintenance, vibration data is typically taken and recorded on a scheduled basis during normal use of the machinery. There are a number of systems available for this purpose. Many of such systems are 40 designed around a pool of portable data collectors. Each individual data collector is designed to be transported to a machine to be tested. Vibration data from the machine is collected, and the vibration data is subsequently uploaded to a host computer. The host computer analyzes the vibration 45 data in order to provide a system operator with complex diagnoses. See, for example, U.S. Pat. Nos. 4,885,707 and 4,612,620.
Although such conventional systems have been useful in the past, there have been a number of drawbacks associated 50 with such systems. For example, the portable data collectors have been limited in their ability to analyze the vibration data locally. The system host computer was required to carry out complex analysis of the data subsequent to receiving the data from the data collector. This could result in analysis 55 delays, and could further result in missed opportunities for avoiding impending machine failure.
Furthermore, such conventional systems have tended to rely heavily on the host/client relationship between the host computer and the individual data collectors. In addition to 60 performing the complex data analysis, the system host computer was required to provide schedule information and/or lists of machines which require predictive maintenance to the individual data collectors. If the system host computer were to fail for any reason, the predictive main- 65 tenance system as a whole could be effectively shut down. In addition, there may be instances where a communication
2
link between the host computer and one or more of the data collectors cannot be established (e.g., due to limited accessibility to telephone lines, failed communications equipment, etc.). The dependency on a system host computer thus becomes an impediment to a smoothly functioning system.
Portable data collectors also have required considerable labor costs in the past. Several man hours are spent visiting different sites and collecting vibration data from the machines at each site. When problems are detected, oftentimes service personnel and/or more experienced data collector operators must be dispatched to the particular site to determine if in fact there are problems. Again, several more man hours can be lost which may otherwise be attributable to operator error in the original collection of data.
Some "on-line" systems have been developed recently which allow for data to be collected from machines via permanent wiring. Such systems at least partially eliminate the need for portable data collectors. Nevertheless, the systems provide little more than on-line data collection and in some instances, failure probabilities. Analyses of the data and/or advanced diagnoses based on the data are reserved typically to some central location such as the system host.
In view of the aforementioned shortcomings associated with conventional data collectors and systems, there is a strong need in the art for a data collector and system which is no longer dependent on a system host computer or portable data collectors. There is a strong need for a system which is capable of performing its own complex analyses of vibration data. Moreover, there is a strong need for a system which can operate without a system host.
SUMMARY OF THE INVENTION
The present invention provides an entirely new approach to conducting predictive maintenance. The present invention provides a system in which localized experts are provided at various machine sites. The localized experts are automated data collector/analyzers which are configured to acquire and analyze vibration data substantially continuously from one or more machines located at the site. The vibration data is analyzed in real-time and the condition of the machines may be ascertained without reliance on portable data collectors and/or analysis by a system host computer.
In a preferred embodiment, one or more local experts are provided at a given site (e.g., building, ship, boiler room, power station, etc.). Each local expert is computer-driven and is configured to receive vibration data from up to thirty-two different machines located at the site. The local expert includes an expert system and database which allows the local expert to analyze the vibration data locally for each particular machine to detect incipient failure, diagnose wear, etc.
Each local expert includes an interface which is linked to a system network such as a local-area-network (LAN) or wide-area-network (WAN). Furthermore, each local expert may communicate with a remote location via a dial-up network, the Internet, etc. The diagnoses performed by each local expert can be used to alert appropriate personnel of impending failure, wear (such as bearing wear), necessary maintenance, etc. The system enables a local expert to initiate electronic mail (email) messages and/or electronic pages to appropriate personnel indicating a machine failure or impending failure, a need for maintenance, etc.
One or more local experts typically will be sufficient to perform predictive maintenance for an entire site (e.g., an entire building) having several different machines. The
database within each local expert contains information specific to the machines connected thereto, as well as information relating to machines of the same type. An expert system within each local expert draws on such information as part of a rules database and/or knowledge database also included 5 therein. Based on such information, the local expert ascertains which machines are operating normally, or are experiencing minor, moderate or severe problems. Depending on the severity of any problems identified by the local expert as well as other possible factors, the local experts are configured to contact the appropriate personnel by email, pages, causing a network event (e.g., setting a flag), etc.
In this manner, the local experts in the present invention not only collect vibration data, but also analyze the data in order to ascertain the condition of the machine and diagnose
15
specific problems such as thrust bearing wear, coupling misalignment, gear problems, etc. The local experts can therefore not only alert the appropriate personnel of a problem, but also can inform the personnel of the particular cause of the problem. In addition, the local experts can recommend action needed to be taken. For example, the local experts may identify which bearings need to be replaced on a machine, the specific part numbers of the bearings, and may provide instructions for enabling a maintenance engineer to make any necessary repairs. ^
Thus, at the same time a person is notified that there is a problem with a particular machine, that person is also notified in many instances of the particular cause of the problem and the maintenance needed to be performed. No longer does a predictive maintenance system require the 3Q sizable delays associated with data being collected by portable data collectors and/or waiting for the collected data to be analyzed. Moreover, considerable cost savings are presented as data is collected and analyzed automatically. Oftentimes the solution to a problem can be provided at the 3J same time as a notice of the problem, thus reducing the amount of time spent analyzing and/or processing the vibration data in order to ascertain the problem using a system host computer.
Vibration data from a machine is collected and analyzed 40 by a local expert substantially continuously. There are no voids in monitoring which are frequently found in systems requiring personnel to collect data periodically during scheduled rounds using portable data collectors.
Another feature of each local expert in accordance with 45 the present invention is an optional connection to a central monitoring location via the dial-up network or Internet, for example. The central monitoring location may be equipped to do special analysis and/or offer user support. Each local expert is designed to be able to replicate its data within its 50 database into a central monitoring location database. Such replication preferably is carried out automatically by the local expert such that the data is available for review at the central monitoring location substantially as soon as a problem is detected. 55
In addition, the local expert is configured to permit the central monitoring location or another location on a local or remote network to control the local expert via the dial-up network or Internet connection. Consequently, the central monitoring location or other location can control the local 60 expert from a remote location to perform special testing and analysis in real time. This enables someone at the central monitoring location or elsewhere to look effectively "under the hood" of the particular machines within the network without necessitating sending personnel directly to the site. 65
According to one particular aspect of the invention, a method is provided for invoking condition monitoring
among a plurality of machines, comprising establishing a network of automated local experts at generally fixed locations and interconnected by at least one network connection, configuring each of the local experts to receive vibration data from at least one of the plurality of machines which is located in relative physical proximity thereto and configuring each of the local experts to analyze the received vibration data and to diagnose a condition of the machine providing the received vibration data based on the received vibration, and configuring each of the local experts to transmit diagnostic information relating to the condition of the respective machines via the at least one network connection.
To the accomplishment of the foregoing and related ends, the invention, then, comprises the features hereinafter fully described and particularly pointed out in the claims. The following description and the annexed drawings set forth in detail certain illustrative embodiments of the invention. These embodiments are indicative, however, of but a few of the various ways in which the principles of the invention may be employed. Other objects, advantages and novel features of the invention will become apparent from the following detailed description of the invention when considered in conjunction with the drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a system view illustrating a diagnostic network with an automated local expert in accordance with the present invention;
FIG. 2 is a system view illustrating a diagnostic network including a plurality of local experts in accordance with the present invention;
FIG. 3 is a block diagram of a local expert in accordance with the present invention;
FIG. 4 is a flowchart suitable for programming the operation of the local experts in accordance with the present invention; and
FIG. 5 is a block diagram illustrating in more detail the expert analyzer and maintenance manager in accordance with the present invention.
DESCRIPTION OF THE PREFERRED
EMBODIMENTS
The present invention will now be described with reference to the drawings, wherein like reference numerals are used to refer to like elements throughout.
Referring initially to FIG. 1, a diagnostic network in accordance with the present invention is generally designated 10. In the exemplary embodiment of FIG. 1, the network 10 is focused around a particular site or sites (represented by phantom line 12). The site 12 may be an office building, manufacturing facility, power plant, etc., or basically any location(s) having one or more machines which are to be monitored for predictive maintenance. Such machines may be engines, turbines, compressors, generators, motors, or any other type of machine for which predictive maintenance is useful. The site 12, as exemplified in FIG. 1, includes a plurality of machines 14. The machines 14 may each be the same kind of machine or different machines as will be appreciated.
The network 10 further includes at least one local expert 16 coupled to the machines 14 for receiving predictive maintenance information such as vibration data. As will be discussed in more detail below in relation to FIGS. 3-5, each local expert 16 in the exemplary embodiment is a computerdriven data collector and analyzer. The local expert 16 receives vibration data from each of the machines 14 or a subset thereof. The local expert 16 includes an expert system for analyzing the vibration data in relation to a rules database and/or knowledge database in order to evaluate the condition 5 of the respective machines 14. The local expert 16 includes a database which stores information on the particular machines 14 from which it receives vibration data as well as a library of data on machines of the same type. In addition, or in the alternative, the expert system may include a neural network, fuzzy logic, or other adaptive and/or predictive analyzers. By applying an appropriate set of rules and knowledge facts, for example, the local expert 16 analyzes the vibration data and diagnoses the condition of the machines 14. For example, the local expert 16 may diagnose particular bearing wear on a given machine 14. The network 10 can be distinguished over conventional on-line systems in that the local expert 16 is continuously able to collect vibration data as well as configured to analyze the data to provide diagnostic information and intelligently notify the appropriate personnel as is discussed more fully below.
As represented in FIG. 1, each machine 14 includes one or more sensors 18 (e.g., accelerometers) for providing a respective vibration data signal to the local expert 16. In the exemplary embodiment, each local expert 16 has a 32 2J channel input and is thus capable of receiving vibration data signals from up to 32 different sensors 18. Of course, each local expert 16 could be designed to have some other number of channels without departing from the scope of the invention. Each machine 14 may have one or more sensors 3Q 18 mounted thereon. The output from each sensor 18 is coupled to the local expert 16 via a hardwired cable, for example. Alternatively, the output of each sensor 18 may be transmitted to the local expert 16 via an optical cable link, or wirelessly via an infrared link, radio link, etc. In this 3J manner, the local expert 16 may be mounted on a wall or some other location within the site 12, away from the vibration, dirt, debris, etc. which may be found near the machines 14. At the same time, a single local expert 16 can be used to monitor up to 32 different machines (or some 4Q other number).
FIG. 1 illustrates how the network 10 further includes a system backbone 20 (such as an Ethernet or Token Ring bus) to which each local expert 16 is coupled. The system backbone 20 typically will be part of a local area network 45 (LAN) or wide area network (WAN) found at the site 12 which provides for everyday communications. Nevertheless, the system backbone 20 may be part of a dedicated network for use in accordance with the present invention.
A number of other devices can be coupled to the system 50 backbone 20 for use in accordance with the present invention. Communications between the devices and the local expert(s) 16 may be based on serial or parallel data communication or a combination thereof, and are carried out using conventional network communication protocols such 55 as Ethernet or Token Ring, as will be appreciated. Thus, additional details regarding the particular communication protocols are omitted herein for sake of brevity.
In the exemplary embodiment, the network 10 includes devices coupled to the system backbone 20 such as one or 60 more web-based clients 22 for requesting information from the local experts 16 in a web-based format (e.g., html files). Such information includes information relating to the collected vibration data for a given machine 14, diagnostic information provided by the local experts 16, etc. 65
One or more "object linking embedding (OLE) process control" (commonly known as "OPC") clients are coupled to
the backbone 20 for providing OPC control and review of the local experts 16. For example, an OPC mimic display client(s) 24 is provided on the backbone 20 and permits mimicking of the OPC control of the local expert(s) 16 anywhere within the network.
An email server 26 is also coupled to the backbone 20 and functions to control the routing of email messages provided by the local experts 16 within the network site 12 to their specified addresses within the network. One or more email clients (not shown) are coupled to the system backbone 20 (either directly or indirectly) and represents the addresses of personnel identified as being appropriate to receive emails from the local experts 16 regarding the condition of a machine 14.
One or more control panel OPC interfaces 28 are also provided on the system backbone 20 for allowing personnel located remote from the local experts 16 to control the local experts and/or other conditions involving the machines 14, as well as view data collected and/or analyzed by the local experts 16 as is discussed in more detail below.
One or more pager clients 29 also may be coupled to the system backbone 20. As discussed below, each local expert 16 has a pager server on board for generating and sending pager messages. The pager messages are used to send alerts as to machine condition as determined by the local experts 16. The pager client 29 shown in FIG. 1 may be a local pager facility for transmitting pages to the appropriate individual pagers within the site 12, or a link to a standard commercial paging facility which transmits pages over a wide region.
The local expert(s) 16 are each able to communicate with such devices coupled to the system backbone 20 by employing known network communication protocols as noted above. The backbone 20 may be a hardwired medium or wireless medium such as infrared, RF, etc., as will be appreciated. Likewise, connections with other networks or devices within the network system 10 may include hardwired links, wireless links, etc. Communications may take place via modem, radio transmission, satellite communications, etc., as will be appreciated.
As is discussed in more detail below, a local expert 16 which diagnoses a problem with one of the machines 14 it is monitoring is designed to notify the appropriate personnel via email, pager, and/or causing a network event, for example. Thus, if the local expert 16 diagnoses a problem it is configured to generate an email or page, for example, which is provided to the email server 26/pager client 29 via the backbone 20. The email server 26 in turn routes the email and/or pager message to the addressee. The content of the email is generated by the local expert 16, and includes the diagnoses ascertained by the local expert 16. For example, an email and/or pager text message indicating "Machine ID XYZ has minor/advanced/extreme bearing fault" is generated by the local expert 16 based on its analysis. The local expert 16 has an address book stored therein identifying the email/pager addresses of the appropriate personnel to be notified in the event of problems with a given machine 14. Such personnel may be operations and/or maintenance personnel for the particular type of machine involved. In this manner, the maintenance personnel are not required to periodically visit each machine 14 to evaluate its condition. Operations personnel are thereby notified as to whether machinery lineup changes, configuration changes, etc., are necessary. Thus, the operations and maintenance personnel can attend to other matters unless and until they receive an email or other notification identifying a problem with a particular machine 14.
|
