-
1
A Smart Machine Supervisory System Framework
Sri Atluru1, Samuel H. Huang1, and John P. Snyder2
1School of Dynamic Systems University of Cincinnati Cincinnati,
OH 45221
2Techsolve Inc.
6705 Steger Drive Cincinnati, OH 45237
Abstract
Machine tools and machining systems have gone through
significant improvements in the past several decades. Recent
advance in information technology made it possible to collect and
analyze a large amount of data in real-time. This brings about the
concept of a smart machine tool, enabled by process monitoring and
control technologies, to produce the first and all subsequent parts
correctly. This paper presents a system framework for a smart
machine supervisory system. The supervisory system integrates
individual technologies and makes overall intelligent decisions to
improve machining performance. The communication mechanism of the
supervisory system is discussed in detail. Its decision-making
mechanism is illustrated through an example that integrates process
planning, health maintenance, and tool condition monitoring.
Keywords: Smart Machine, Supervisory System, Communication,
Decision Making 1. Introduction
With growing technological advancements in the manufacturing
world, there has been an emergence of various control systems and
technologies that would help increase the efficiency of the machine
tool. However, most of these technologies are disparate in the
sense that their specialization was confined to the optimization of
only one component of the machining process. A general consensus
has recently emerged that the effectiveness of automation lies not
only in the technical capabilities of individual process monitoring
and control systems, but also in the ability of a machine tool to
coordinate among all the individual technologies and control
systems to deliver an overall optimal performance. The ability to
monitor and control multiple process modules forms the basis of the
next-generation machine tool, the Smart Machine, which will result
in higher productivity, better quality, and prognostic capability
for near-zero breakdown performance in the machining process.
-
2
However, it is important to understand that coordination between
individual control systems and technologies can potentially lead to
a number of problems including conflicting outputs from different
systems and the ordering of priority for individual process
adjustments. Hence, in order to realize the idea of the smart
machine, there is a need for a closed-loop supervisory system to
coordinate individual process modules for real-time adjustment,
conflict resolution, and priority assignment. The motivation for
the development of a supervisory system was identified in the Smart
Machine Platform Initiative [3]. The smart machine supervisory
system is the manufacturing expert system which works like the
brain and nervous system of a smart machine. It collects
information from individual components of the smart machine and
makes hierarchical decisions based on a set of predefined
manufacturing rules and logics. Hence, it addresses the need for an
all-encompassing system to enable the First Part Correct philosophy
[1].
This paper describes the design and implementation of a smart
machine supervisory system. It focuses on the technical definition
and architecture of the supervisory system as an over-arching
functional area over other components of a smart machine. It then
proceeds to illustrate how to implement a supervisory system with
an emphasis on integrating tool condition monitoring, pre-process
planning, and machine health and maintenance.
2. Literature Review
It is well established that multiple-process monitoring and
control improves productivity and reduces machining time [16].
However, existing technologies related to supervisory control have
been limited to regulating a single process using a single process
variable [7, 14]. Additionally, it was observed that within the
existing machining applications, there are no established
procedures or standards to implement effective process control
[20]. Most of these applications use propriety software and
hardware that are bundled together, making them incompatible with
other applications.
The technologies related to process monitoring and control in a
smart machine can be classified into specialized thrust areas based
on their functionalities, briefly described as follows:
Tool condition monitoring: It allows detection of cutting tool
conditions, including wear, breakage, missing tools, and collision,
in the machining process. It can also function as an excellent
process monitoring mechanism with additional capabilities such as
metalworking fluid flow monitoring, spindle health and maintenance
monitoring, and adaptive control. Technologies that come under this
thrust area include popular systems used to monitor cutting tool
conditions during the cutting process, such as Caron Engineering
TMAC, Artis, and Techna-Tool.
-
3
On machine probing: It concerns with technologies that allow the
inspection of a work piece on a machine tool with minimal
peripheral equipment or personnel. It also facilitates the
verification of machined parts on the machine tool mitigating the
use of a Coordinate Measuring Machine (CMM) to determine
geometrical tolerance accuracy. On machine probing is usually
accompanied by the use of an on-machine probe similar to the probes
used on a CMM. The on-machine probe is used to accurately determine
the coordinates at pre-determined locations to support the
verification process.
Intelligent process planning: It generates, verifies, and
optimizes tool paths, automatically selects the most-suitable
cutting tools, and optimizes cutting conditions. In addition,
optimum cutting parameters (speeds and feeds, etc.) are generated
based on overall machining performance requirements, including
surface roughness, cutting forces, material removal rate, and tool
life. Most Computer-Aided Design/ Manufacturing technologies
(CAD/CAM) are drawn under the purview of this thrust area.
Additionally, various Computer-Aided Engineering (CAE) software
solutions that enable users to analyze machining in 2D and 3D
environments by predicting performance indicators are also
classified under this thrust area.
Machine tool metrology: It identifies the differences in the
reported and actual position of a cutting tool. Sources for these
differences can be errors built into the machine such as
straightness, linearity, square-ness, pitch, roll, yaw, or dynamic
error sources such as thermal growth and cutting force tool
deflection.
Machine health and maintenance: It assesses the health condition
of the machine tool (in the areas of availability and utilization).
Valuable data, such as controller signals and sensor measurements,
are analyzed using appropriate prognostics algorithms that allow
for machine condition assessment, as well as prediction of
performance degradation, so that equipment can be repaired before
component failures actually occur.
Supervisory system: it is in charge of coordinating technologies
resulted from all the other thrust areas to provide an overall
solution to improve machine tool performance.
Research in these thrust areas over the past decades has
resulted in a number of commercial products and promising new
technologies. These products and technologies are summarized in
Table 1.
Table 1: Products and technologies related to smart machine
thrust areas
SMPI Thrust Area Products/Technologies Reference
Tool Condition Monitoring
Caron Engineering TMAC http://www.caron-eng.com/ Blum Laser
Measurement Tooling http://www.blum-novotest.com/
Artis
http://www.artis.de/en/competences/monitoring-solutions/tool-monitoring/
-
4
Nordmann http://www.nordmann.eu/usa/aeltere_neuigkeiten.html
Techna-Tool http://www.techna-tool.com/
On-machine Probing PC-DMIS NC http://pcdmis.com/pc-dmis-nc
Intelligent Process Planning
ThirdWave AdvantEdge Production Module
http://www.thirdwavesys.com/products/advantedge_production_module.htm
Esprit CAM http://www.dptechnology.com/ CimSkil
http://www.cimskil.com/ Master CAM http://www.mastercam.com/
Vericut http://www.cgtech.com
Machine Tool Metrology
INORA http://www.inora.com/ Remmele http://remmele.com/
Health and Maintenance
WatchDog Agent http://www.imscenter.net/
Freedom E-log
http://www.infimatic.com/products/freedom-elog-products.html
Supervisory System
GE Fanuc Oi http://ge-fanuc.com/ NI LabVIEW DSC
http://www.ni.com/labview/labviewdsc/ I/Gear DTU
http://www.igearonline.com/Products/DTU/
Siemens SINUMERIK 840Di
http://www.sea.siemens.com/us/Industry_Solutions/Machine-tools/Products/CNC/Pages/SINUMERIK_840Di.aspx
KEPServerEX OPC Server
http://www.kepware.com/Products/kepserverex_features.asp
B2D Solution Manufacturing
http://www.b2dsolutions.com/Solutions_HTML/industry.html MTConnect
http://www.mtconnect.org
With regards to the incorporation of a supervisory control
mechanism for individual
technologies, it is notable that current R&D efforts in
academia and industry is related and directed towards the
development of Open Architecture Systems viz. Open Modular
Architecture Control (OMAC) technologies group, Open System
Architecture for Controls within Automation (OSACA) systems, Japan
FA Open systems Promotion (JOP) group and STEP-NC [10-13]. These
open architecture systems are expected to address the limitations
posed due to the lack of a standard communication protocols among
individual technologies.
A recent effort to develop the communication standard between
multiple process controls is the MTConnect initiative, which was
intended to help realize the "seamless manufacturing pipeline" from
design to production [17]. The goal of this pipeline approach is to
allow for universal capture of data from the machine tool and then
transfer this captured data to other control systems; thereby
facilitating a seamless method for managing and analyzing data for
process and product optimization. MTConnect is based on the
eXtensible Markup Language (XML), which provides for exchange of
semi-structured machine-readable
-
5
data. It is also expected to account for the seamless
connectivity between various components and systems of the machine
tool, right from the lowest end of the process chain to the highest
level. Additionally, MTConnect is also expected to deliver on its
goal of interoperability, which will enable third party solution
providers to embrace the standard in their products [2]. MTConnect
specifications were formulated after extensive review and analysis
of various other standards including OMAC, Cam-X, and OPC. The XML
based approach accounts for hierarchical levels inside the data
transferred. It is widely supported by various software and
hardware systems, which implies that it can be adopted relatively
easily by the manufacturing industry.
In addition to communication, there is a need to develop sensor
fusion technologies, as well as systematic design approaches to
intelligently construct and implement multi-process control modules
in the manufacturing industry [19]. The research is this area had
largely been based on ad-hoc construction of various process
controllers for specific manufacturing systems [7, 9]. The effort
for developing an integrated multiple process control technology is
limited. This paper presents a systematic framework to develop an
integrated smart machine supervisory system to bridge this
technology gap.
3. System Framework
The smart machine supervisory system is defined as a system that
integrates and coordinates individual process monitoring and
control modules such that a globally optimal machining solution
could be delivered real-time to achieve desired quality and maximum
productivity. A schematic framework of the supervisory system is
shown in Figure 1. The major functions of the supervisory system
are communication and decision making. The following subsections
describe the communication and decision functions, along with an
illustration of the relationship between the supervisory system and
individual control modules.
(Insert Figure 1 here)
3.1 Communication Function
Communication, including sending control signals from the smart
machine supervisory system to the machine and reading the machine
information into the supervisory system, is a key function of the
supervisor system. The communication function is intended to be
implemented complying with the MTConnect protocol. Implementation
of MTConnect on any non-compliant system or legacy machine requires
the deployment of an adapter and an agent system, technically
referred to as agent core. The MTConnect compliant data is then
output by the agent, which can be utilized by external applications
for further processing and analysis.
-
6
An MTConnect adapter was built for the Fanuc oi-mc controller.
It was developed in C++ programming language. It is a simple
adapter built to accept one connection on its socket server. The
Fanuc adapter is in turn, programmed to use the FOCAS (Fanuc Open
CNC API Specifications) libraries which enable the reading of CNC
and PMC data from the machine controller via the Ethernet. The
FOCAS libraries (Fwlib32.dll and Fwlibe1.dll) contain numerous
functions that reference the data window functions on the Ethernet
board of the CNC to access the data available about the machine.
The adapter essentially serves as a routing channel between the
controller and the MTConnect agent, while also assuming the
responsibility for translating raw data from the individual FOCAS
functions into structured data that can be easily comprehended by
the MTConnect agent. The agent was developed on the Ruby on Rails
platform and uses a SQLite3 database.
3.2 Decision Function
One of the features of the supervisory system is to have the
ability to make decisions based on the data supplied by the thrust
area technologies. The supervisory system needs to monitor, in real
time, the inputs from various technologies and must be able to
effectively process multiple process signals simultaneously to make
the necessary decisions. However, to initiate any adaptive action,
there is a need for the supervisory system to communicate back to
the CNC. Additionally, in case of an emergency, there is a need for
the supervisory system to ensure that the machine responds to the
supervisory system with a higher priority than the current NC code
being processed.
A solution to address the aforementioned issues is to
communicate efficiently with the CNC in real-time, using
methodologies such as Interruption Type Custom Macros, which
implement the ability to read inbuilt data window functions within
the CNC machine controller. This is explained in detail below and
can be controlled through USB control switches on the PC.
(Insert Figure 2 here)
In Fanuc controllers, when a program is being executed, it is
possible that another program can be called by inputting an
interrupt signal (UINT) from the machine. This function is referred
to as an interruption type custom macro function (Figure 2). The
format is as follows:
M96 Pxxxx; enables the macro interrupt
M97; disables the macro interrupt.
When M96 Pxxxx is specified in a program, subsequent program
operation can be interrupted by an interrupt signal (UINT) input to
execute the program specified by Pxxxx.
-
7
When the interrupt signal (UINT, marked by * in Fig. 2) is input
after M97 is specified, it is ignored.
The supervisory system stores relevant decisions, such as stop
machining when a broken tool is detected, in programs specified by
Pxxxx. During machining, it evaluates signals provided by
individual control modules in real-time, and triggers the interrupt
signal when necessary. A detailed example of the decision making
process is provided in Section 4.
3.3 Relationship between the Supervisory System and Individual
Control Modules
The supervisory system is responsible for communicating with
individual control modules, evaluating the information it obtained,
making appropriate decisions, and controlling the machining
process. Here we use machine health and maintenance to illustrate
the relationship between the supervisory system and thrust area
technologies. The health and maintenance thrust area is intended to
monitor the CNC machine, perform machine prognosis, and determine
the OEE (Overall Equipment Effectiveness) of the CNC machine on the
shop floor [5].
It is the responsibility of the supervisory system to provide
the over-arching functionality for health and maintenance
technologies to seamlessly communicate with the CNC machine to
monitor the required data and subsequently, use the prognosis
forecasted by these technologies to determine the suitability of
the CNC machine to do a certain job. Health and maintenance
technologies require the following data parameters from the machine
controller [15]:
- Actual spindle speed - Spindle status - Trigger to start
monitoring the tool assembly - Two macro-variables, which provide
information about the system parameters of the
controller - Actual feed rate - Spindle load (in % from the load
meter on the controller)
The MTConnect implementation within the supervisory system is
designed to provide
for seamless data transfer between the machine tool controller
and health and maintenance applications. The MTConnect adapter
built for Fanuc captured the machine PMC data and transferred it to
the MTConnect agent. This agent transformed the machine data into
MTConnect standards and relayed it onto a HTTP port, which was
accessed readily by the health and maintenance technologies. The
health and maintenance technologies had access to peripheral
sensors mounted on the CNC machine, thus enabling the accurate
prognosis of CNC machine tool health using sensory data as well as
the PMC data parameters supplied through the MTConnect Agent Core
implemented.
-
8
Figure 3 depicts the architecture of the MTConnect system
installed enabling communication between the supervisory system and
the health and maintenance technologies. This implementation also
addressed the issue of plug and play functionality across a variety
of machine tool controllers from different manufacturers. It
illustrated the concept of having multiple MTConnect agent cores
for various CNC machine tools that would communicate across the
shop floor network to various applications and technologies.
(Insert Figure 3 here)
4. System Integration: Process Planning, Health Maintenance, and
Tool Condition Monitoring
Now we focus on the decision-making capability of the
supervisory system when coordinating various thrust area
technologies. An integration interface was developed among three of
the thrust areas -- Intelligent Process Planning (IPP), Tool
Condition Monitoring (TCM), and Health and Maintenance (HAM) -- as
well as with the CNC Machine Tool. As there are a multitude of
technologies pertaining to each thrust area, one technology was
chosen to represent each of these thrust areas. The choice of these
technologies was based on detailed review of the technologies and
expert feedback. The thrust area technologies, thus selected, are
briefly described as follows:
- IPP deals with virtual simulation and the subsequent
generation of optimized tooling and tool paths necessary for
machining operations. In addition, cutting parameters (speeds,
feeds, etc.) are also optimized based on overall machining
performance requirements, including surface roughness, cutting
forces, material removal rate, and tool life. ThirdWave AdvantEdge
Production Module [18] was used by IPP in its efforts to simulate
machining process in order to generate NC programs based on
user-defined tool profiles. It then verifies the generated tool
path in the NC program based on its own machining performance
database related to force calculations, physics based material
models, and optimization speedups. Finally, it draws up an
optimized tool path and a new NC program in order to achieve
reduced cycle time, maximum machine utilization, and optimum
machining performance.
- TCM monitors the in-process condition of the cutting tools,
including wear, breakage, presence of tool, and unforeseen
collisions. Most of TCM technologies are power sensor based
applications that monitor the spindle power to determine and
predict the occurrences of tool wear and other tool defects. Tool
Monitoring Adaptive Control (TMAC) by Caron Engineering [6] is
selected for TCM. TMAC supports all central monitoring tasks
expected of TCM technologies based on fluctuations in spindle power
recorded. The recorded spindle power is weighed against a power
representation that
-
9
forms the basis for determining the permissible limits of wear
for specific cutting processes. These limits that determine the
permissible wear on the cutting tool are based on the quality
requirements of the manufactured part.
- HAM aims at providing an accurate description of the quality
of the machine tool based on machine prognostics and experimental
data. The WatchDog Agent [8] was developed by the Intelligent
Maintenance Systems Center at the University of Cincinnati to
support the HAM module, with a goal to achieve a health monitoring
system capable of accurately monitoring and predicting the machine
health for near-zero downtime.
The integration of these thrust area technologies via the
supervisory system aims to overcome drawbacks of individual
technologies to achieve optimal machining performance. We first
discuss the drawbacks of the TCM and IPP and how these drawbacks
can be overcame through supervisory system integration. We then
discuss further supervisory system integration with HAM to provide
required data in real time.
The current principle of almost all TCM technologies, including
TMAC, is based on learning a good cut to set the tool limits to
diagnose defects in the tooling assembly. The essentiality of a
learning cut, which acts as a basis for future monitoring, has the
following drawbacks:
- The machine needs at least one stock part and one new tool for
the learning cycle. - The tool and machine are not protected
against errors or collisions during the learning
cycle. - Given the lack of computerized monitoring during the
learning cut, it is very possible
that the part produced through a learning cut will need to be
scrapped or re-machined. - The subsequent limits to monitor tool
wear, tool breakage and tool presence were
usually set based on historical data rather than a scientific
approach.
On the other hand, IPP technologies utilized physics based
material models to draw up force calculations in their efforts to
optimize NC tool paths to increase machine utilization. However,
the outputs from the ThirdWave production module were only the NC
program with an optimized tool path. The internal force
calculations and physics based material models served as volatile
internal data that was simply put away on subsequent optimizations;
and thus, could not be further utilized in streamlining the
monitoring process of other technologies.
By tapping into the force calculations and the physics based
material models used by ThirdWave AdvantEdge Production Module, it
is possible to actually postulate a fundamental power
representation of the cutting process. This power representation
can be used in place of the power representations picked up by
Caron TMAC technology during its learning cut. Several mappings
were created to within the supervisory system to enable the
integration
-
10
between IPP and TCM. These mapping included ways to tap into the
internal force calculations of the ThirdWave Production Module, and
also methods to customize the scheduled tool condition monitoring
job to look at the power representations built based on those force
calculations instead of the usual representations picked up during
the learning Cut.
As shown in Figure 4, the idea is to use a combination of these
mappings to generate an accurate power representation (peak power)
and force prediction based on material physics. This way, the job
file can be generated automatically and the process of learning a
good cut can be eliminated. However, it must be noted that the
power representation picked up by Caron Engineering during its
learning cut is subjected to processing under various internal
filters during actual monitoring process. To facilitate this
internal filtering by the Caron TMAC technology, the first cut is
instead used as a calibration cut.
(Insert Figure 4 here)
Similarly, while doing away with the manual inputs required by
HAM technology represented by the WatchDog Agent (as described in
Section 3.3), additional functionality was incorporated by using a
set of advanced decision rules based on expert feedback and
historical data, which was utilized to check for tool validity and
determine if the tooling selected by the NC program meets the
conditions (user defined tool profile and machine definitions)
assumed by IPP technologies during their optimization and NC code
generation. The tooling and machine setup used in ThirdWave
Production Module (as well as other IPP technologies) needs to
match the tooling setup that exists on the ATC (Automatic Tool
Changer) of the CNC machine, to avoid potential conflicts during
actual machining. The process diagram illustrating this integration
is shown in Figure 5.
(Insert Figure 5 here)
Figure 6 shows the complete integration approach adopted by the
supervisory system. The interruption type custom macro is used to
handle any cutting tool abnormality (wear beyond permissible
limits, missing tool, and broken tool) detected by Caron TMAC. A
retract program is written in such a way that the cutting tool
retracts back whenever the supervisory system detects an
abnormality through Caron TMAC technology. In this process, the
supervisory system takes the alarm signal from the log file of the
Caron TMAC system, instead of waiting for the alarm to be generated
on the controller screen. This is to facilitate real-time response
to the situation.
(Insert Figure 6 here)
In addition to the integrations, various correlations pertaining
to process uncertainty and tool tolerances, such as limits for wear
of the tool, were analyzed so that intelligent
-
11
knowledge-based decisions can be made by the supervisory system.
A feed forward process design about the correlation between these
technologies was then developed to simulate the learning curve in
consequent process plans. A number of tests were conducted to
determine a correlation between predicted power (using power
representations from ThirdWave) and measured power (during TMAC
learning cut). This correlation was then used to determine the
predicted power representation for the monitoring process. The
fitting relation for TCM power representation (peak power
prediction) obtained was
Pp = 1.396PTW + 0.574 (R = 0.988)
where,
PTW is the peak power representation predicted from Third
Wave
Pp is the predicted peak power to be used as the power
representation in Caron TMAC
This correlation was generated from tests on 5 different cutting
regimes (each regime being defined as a distinct combination of
speed and feed) and 10 data points on each of the regimes.
Subsequently, a number of tests were conducted to investigate the
uncertainty of prediction based on the ThirdWave power
representations associated with the above equation. An unbiased
experimental design block was generated for two different cutting
regimes.
As per the guidelines for evaluating and expressing the
uncertainty of measurement results provided by the National
Institute of Standards and Technology, a combined standard
uncertainty of 3 ( being the standard deviation in the
observations) would encompass over 99% of the total normal
distribution. Hence, these uncertainties associated with 3 were
used as a starting point limits to determine permissible wear on
the cutting tool during the calibration run.
As has been mentioned before, subsequent to the calibration cut,
the supervisory system continues to monitor the log file of the TCM
technology during all ensuing tool condition monitoring tasks, to
initiate any adaptive action that might be necessary in the case of
any alarm. The communication module of the prototype supervisory
system continues to communicate with one of the subsystem
components, specifically Caron TMAC System, to generate an alarm if
required. The alarm generation would then trigger an interruption
type custom macro in the Fanuc controller which is preprogrammed to
execute a retract program to prevent any further damage to the tool
and the work piece. An USB controlled digital output generator was
also used to aid in the implementation of the interruption type
custom macro. LabVIEW (National Instruments) was used to trigger
the interruption type custom macro.
-
12
The GUI (Graphical User Interface) of the main LabVIEW program
(Figure 7) was developed to create a flexible and operator friendly
application for use on the shop floor. It automated the whole
process with the goal of the supervisory system and First Part
Correct in view. This prototype supervisory system was successfully
demonstrated to the industry a number of times during the years
2008 to 2010. The supervisory system was successful in
demonstrating the generation of job files for TCM technology based
on the inputs (power representations) from the optimized cutter
path generated by IPP technology. It also successfully verified
whether the tooling available on the ATC of the CNC machine was in
line with the requirements of the tooling definitions and machine
profiles assumed during the pre-process by IPP technologies. It
also successfully verified the suitability of the tool holder
assembly to do the prescribed cutting and made changes to the NC
program when appropriate.
(Insert Figure 7 here)
5. Conclusion
The Smart Machine program was developed as a reinvention of the
basic manufacturing process. It aimed at providing an optimal
manufacturing process through the coordination of various disparate
manufacturing control systems. The supervisory system is in charge
of coordinating individual technology areas to deliver an optimal
manufacturing solution in real-time. A prototype of the supervisory
system was developed to demonstrate this functionality. It made use
of available data in technologies employed by intelligent process
planning, tool condition monitoring, and health and maintenance, to
provide an optimized solution by cutting down on time required for
tool verification, metal cutting for learning processes, and
calibration. In the future, the prototype also needs to incorporate
other thrust areas, viz. on-machine probing, and machine tool
metrology, to develop a more robust supervisory system.
Acknowledgement This research was sponsored by the U.S. Army
Benet Laboratories and was accomplished under Cooperative Agreement
Number W15QKN-06-2-0100. The views and conclusions contained in
this document are those of the authors and should not be
interpreted as representing the official policies, either expressed
or implied, of U.S. Army Benet Laboratories or the U.S. Government.
The U.S. Government is authorized to reproduce and distribute
reprints for Government purposes notwithstanding any copyright
notation heron.
-
13
Reference 1. Atluru S, Deshpande A, Huang SH, Snyder JP (2008)
Smart machine supervisory system:
concept, definition and application, Society for Machinery
Failure Prevention Technology 62nd Meeting Conference, May 6-8,
Virginia Beach, VA.
2. Atluru S, Deshpande A (2009) Data to information: Can
MTConnect deliver the promise? Transactions of NAMRI/SME, 37:
197-204.
3. CMTI (2004) Smart Machine Technology Initiative (SMPI)
technology plan, Coalition on Manufacturing Technology
Infrastructure, TechSolve Inc., Cincinnati, OH.
4. Caron Engineering (2010) TMAC: Tool Monitoring Adaptive
Control, http://www.caron-eng.com/tmac/
5. Djurdjanovic D, Lee J, Ni J (2003) Watchdog agent: an
infotronics-based prognostics approach for product performance
degradation assessment and prediction, Advanced Engineering
Informatics, 17(3/4): 109-125.
6. Furness RJ, Ulsoy AG, Wu CL (1996) Supervisory control of
drilling, ASME J. Eng. Indus., 118(1): 10-19.
7. Fussell BK, Jerard RB, Hemmett JG (2001) Robust feedrate
selection for 3-axis machining using discrete models, ASME J.
Manuf. Sci. Eng., 123(2): 214-224.
8. Intelligent Maintenance Systems Center (2003) Development of
Smart Prognostic Agents (WATCHDOG AGENT),
http://www.imscenter.net/Research/core_research/WD_onepage_
Nov17_2003.pdf
9. Jerard RB, Fussell BK, Xu M, Schuyler C (2006) Development of
a testbed for research on smart machine tools, Proceedings of 2006
NSF Design, Service, and Manufacturing Grantees and Research
Conferences, July 24-27, St. Louis, Missouri.
10. Jerard RB, Fussell BK, Ercan MT, Hemmett JG (2000)
Integration of geometric and mechanistic models of NC machining
into an open-architecture machine tool controller, ASME
International Mechanical Engineering Congress and Exposition, Nov.
5-10, Orlando, FL.
11. Katz R, Min BK, Pasek Z (2000) Open architecture control
technology trends, ERC/RMS Report 35, University of Michigan, Ann
Arbor.
12. Landers RG, Ulsoy AG (1995) A supervisory machining control
example, International Conference on Recent Advances in
Mechatronics, Istanbul, Turkey, August 14-16, pp. 990-997.
-
14
13. Landers RG, Ulsoy AG (1997) Supervisory machining control on
an open-architecture platform, CIRP International Conference on
Design and Production of Dies and Molds, Istanbul, Turkey, June
19-21, pp. 97-104.
14. Landers RG, Ulsoy AG (1998) Supervisory machining control:
design and experiments, Annals of the CIRP, 47(1): 301-306.
15. Lapira E, Deshpande A, Lee J, Snyder JP (2008) Tool assembly
prognostics, The International Manufacturing Science and
Engineering Conference (MSEC), Evanston IL, October 7-10.
16. Liang SY, Hecker RL, Landers RG (2004) Machining process
monitoring and control: the state-of-the-art, ASME J. Manuf. Sci.
Eng., 126(2): 297-310.
17. MTConnect (2011) MTConnect Overview,
http://mtconnect.org/index.php?option=com_ content
&task=view&id=95&Itemid=26
18. Third Wave Systems (2011) Production Module,
http://www.thirdwavesys.com/products/
advantedge_production_module.htm
19. Ulsoy AG, Koren Y (1993) Control of machining processes,
ASME J. Dyn. Syst. Meas. Control, 115(2B): 301308.
20. Xu XW, Newman ST (2006) Making CNC machine tools more open,
interoperable and intelligent: a review of the technologies,
Computers in Industry, 57(2): 141152.
-
15
Figure 1 : Schematic of supervisory system architecture.
-
16
Figure 2 : The interruption type custom macro (from GE Fanuc
documentation)
-
17
Figure 3: Architecture of MTConnect interface for health and
maintenance technology
-
18
Figure 4: Process diagram for TCM-IPP integration
-
19
Figure 5: Process flow of integration with health and
maintenance technology
-
20
Figure 6: Flowchart showing the overall integration approach
-
21
Figure 7: GUI of the supervisory system