SOME COMMUNICATION PROBLEMS OF KB CONTROLLED

MANUFACTURING SYSTEMS
George L. Kovács and János Nacsa
CIM Research Laboratory, Computer and Automation Research Institute of HAS,
1518 Budapest POB 63, Hungary

ABSTRACT (Franklin et al.,1994). Some experts claim that

there is nothing that really could be called
According to our recent view knowledge based or intelligent control. Most of them are speaking only
intelligent control of Flexible Manufacturing about process control, and not about discrete
Systems (FMS) means a continuous or frequent manufacturing or robot control, however the
observation and evaluation of the status and control tasks and problems of manufacturing
condition of the system performance, decision systems are basically similar to those of batch-like
making based on the evaluation results and on pre- process control. Now, without going into
defined knowledge, and then the operation discussions in this issue we accept the need and
according to the decisions. In the case of normal necessity of intelligent control which we intend to
scheduled operation, there is no need for solve by means of knowledge based (KB) systems,
intelligence, neither for interactions to modify the which we often call expert systems. We admit that
operation. The above given procedure may help some commercial expert systems provide good
(Smith, et al.,1992) if any kind of disturbance or problem description and software development
irregularity happens, what is rather common in the tools where the programming is closer to the
case of highly sophisticated, complex systems. Our problem to be solved and to the user and limited
paper deals with control and communication real-time facilities are supported, too (Laffey et al.
problems of flexible manufacturing systems using 1988).
programs based on a real-time expert system The next chapter shortly introduces our basic
environment taking into consideration the system the developments started from. The third
requirements of openness of up-to-date systems. chapter will discuss the communication problems
Two robot applications will be discussed as of expert systems in FMS applications and give a
demonstrations. problem decomposition. Later on there is a brief
description of the features of the MAP based
Virtual Manufacturing Device model in intelligent
1. INTRODUCTION control. Finally there is a presentation of the
detailed basic definitions, and the way we went on
Recent discrete manufacturing and/or assembly to end up with something what we call "intelligent
systems (FMS/FMA) are more and more often robot cell controller with open communication".
using MAP/MMS (Manufacturing Automation
Protocol, Manufacturing Message Specification)
(Brill and Gramm, 1991), because this technology 2. A KNOWLEDGE BASED HYBRID
is widely available from many vendors and gives a SIMULATION-SCHEDULER SYSTEM
real safe and open solution according to the
demands of OSI (Open System Interconnection). We had developed a knowledge based simulation-
Many users do not exactly know that they have scheduler system called SimSched-Q (Kovács et
such interconnections, they just enjoy the useful al., 1993). In the SimSched-Q system the
features of MAP. We must also admit that the simulation was implemented in a world-wide
MAP or related solutions are still rather expensive. accepted, but 'traditional' simulation/animation
On the other hand the so called intelligent control system called SIMAN/Cinema. It was surrounded
is getting to be a general demand. There is a vivid by expert system modules (preparation, on-line
discussion in literature and in private quality and scheduler advisor and evaluation)
communications among control engineers about implemented in the same G2 Ver. 3.0 object-
the existence and need of intelligent control oriented real-time intelligent environment. The
study in this paper only refers to the SimSched-Q support clear and easy programming to reach
system as resulting scheduling data are the basic objects, to call procedures, to set and get variables,
input for (intelligent) FMS control. Similar efforts etc. The interfaces are dedicated to specific
are also being done (e.g. Somlo and El Buzidi, software tools of the ES and they are general
1996). towards the external world without being able to
There are four expert systems: two of them are take into account the requirements of the given
shallow coupled with the simulation (Preparation application. So nearly each CIM implementation
and Evaluation), two of them are deep-coupled requires special software development to cover this
(Quality and Scheduler advisors). The structure and gap between the external world and the ES.
the functionality of the system were designed and Clearly the communication functions depend on
developed in a way that later the simulation might the capabilities of the expert system. The way of
be changed to a real FMS environment, and the ES learning and knowledge handling determine the
would be the real-time controller. logical levels of the communication. Three
The prototype application was developed with the different types of working mode and different
real data (layout, capacities, process plans, machine levels of the communication of an intelligent cell-
parameters, etc.) of some existing FMS, as for controller in a CIM environment are defined.
example the Pilot FMS of the Technical University (Nacsa and Kovács, 1994). These levels are
of Budapest. This system contains four ‘holon-like’ implemented - of course - within the same
cells (storage with AGV, measurement, assembly protocol.
and metal-cutting consisting of a CNC lathe and a The lowest level (Data Acquisition and Control
CNC machining center served by robots) on a Level) has the basic control and data acquisition
network. type messages. The other two levels (Knowledge
To speak about real-time control we need all Acquisition and Knowledge Communication
scheduling data, as the starting and finishing date Levels) have messages if and only if the
of all operations of all equipment (machine-tools, ‘intelligence of the cell controller is not hidden’.
robots, transfer units, as AGV, etc.) together with Hidden intelligence means in this term that the
the possibilities of downloading CNC control knowledge based technology is applied only inside
programs to all equipment, and to accept and the cell controller and it has no specific actions via
evaluate different signals from them, etc. The time the communication channel. A typical example of
of the loading/unloading should be calculated, too. the hidden case if a KB system is built up on the
top of a traditional control system using its original
communication. On the highest logical level the
3. COMMUNICATION WITH EXPERT knowledge acquisition and the so-called knowledge
SYSTEMS communications were separated. The first one
contains specific data for modifying or verifying
The practical problems of the communication of the knowledge of the given controller. When a KB
expert systems in CIM (Computer Integrated system shares its knowledge (new or modified) it
Manufacturing) applications can be divided into belongs to the knowledge communication level
two parts. One is the hardware-software connection (Buta and Springer, 1992). The communication
(physical) and the other one is the logical messages of most real and pilot KB applications
connection between the controller(s) and controlled for FMS control belong to the lowest, possibly to
devices. This decomposition was very useful both the middle logical level.
in the design and implementation phase during the
last projects of our CIM Laboratory. If this
decomposition is not so sharp many problems may 4. MMS - AN OBJECT ORIENTED
occur during the development and specially in COMMUNICATION TOOL
maintenance later on.
There are relatively easy programming interfaces Going back to the so-called physical connection
(etc. C/C++) in most available ES shells. These there are many possibilities to follow. Most
interfaces provide data transfer and communication controller and controlled device vendors offer
possibilities with external tasks, stations, etc. They
good (proprietary) solutions to communicate and 5. AN EXPERIMENTAL ROBOT CONTROL
also vendor independent standards are available. SYSTEM
Arcwelding Robot VMD R_ARM As the goal of this work is to provide real-time,
(KUKA-ESS) domain
R_SF intelligent control of FMS systems the next step
Vendor, Model, Revision
Physical & Logical Status
was to apply all our development results and to
R_MP49
R_MP50 integrate the independently working and tested
R_GO
prog. inv. R_MP51 software-hardware modules.
domain As a first experimental set-up (Fig. 2.) a simple
configuration was chosen. From earlier projects a
R_WELDST
Mitsubishi robot was available with serial DNC
R_STATE
connection to a PC that had a MAP interface. The
R_ERRCODE R_OPERID trivial TIC-TAC-TOE game was served by the
R_VLOCAL R_AUTO robot according to the following manner: a user
Fig. 1. Virtual Manufacturing Device model of a could play against the G2 system. Of course it is
robot (an example) not a problem for that such an intelligent system is
needed. But this well-known game is able to
Also in the CIM area there are more accepted clearly demonstrate the 'intelligent' features of the
models or modeling tools to describe the objects of control. On the other hand the full capability of
an FMS. In the communication point of view the MMS networking was used.
most promising one is the object oriented view of The structure of the setup shows that a special
the so-called MMS (Manufacturing Message gateway was developed between the G2 base
Specification), which is originally an application controller and the robot. It couples the G2 system
layer protocol in the MAP OSI networks. It was interface (called GSI) to the MMS standard
realized that this specification is good on the higher communication. In this case it was the 'gap
level of the FMS (Nagy and Haidegger, 1994) to software' mentioned in the chapter 3. This special
give a communication oriented view about the gateway of the G2 allows to develop any discrete
network elements and their resources. manufacturing applications to develop in G2 using
MMS gives a so-called VMD (Virtual MMS standard interface without having to modify
Manufacturing Device) view about each resource the interface. In the demonstration version the
of the FMS. All the objects of the MMS (e.g. gateway supports the context management and
VMD, domains, different types of variables, some program invocation services of MMS.
program invocations) are object oriented with For this solution a special object hierarchy called
special sets of attributes and actions (called MLGO (MMS like G2 objects) and a procedure set
services in MMS) belonging to a given object type. were needed within the G2 to make the knowledge
Fig. 1. shows a simple, but rather general type of base to handle the MMS interface. Three solutions
VMD :an arc-welding robot VMD. In a VMD it is were examined before developing our MLGO.
possible to create, read and write different types of  It was examined how the complete MMS
variables, to up- and download domains which can object structure could be built up within the
be programs or machine data, to start and stop G2. It took a lot of efforts and gave a relatively
different tasks (with program invocations) and to complicated and not too useful structure.
handle events. The object-oriented view of MMS Because of the different services and their
allows to design the VMD model of a certain parameters many external procedure calls were
device, and it is immediately possible to use this defined with big parameter sets. All
model as a specification of the communication implemented MMS services have their own
where the services (what one can do with a given GSI procedures and all MMS object types have
object of the VMD) are defined and they are their internal G2 object descriptions. The
working in the MMS networks. complexity of the inner structure was the same
as the MMS itself. It did not give a clear and
easy-to-use tool for the G2 user.
 INTELLIGENT GATEWAY
GSI GSI
CONTROLLER
obj. & C
(G2
ENVIRONMENT) proce- inter-
dures face
MMS like G2 objects MMS client

CONTROL CONTROL
RESPONSE REQUEST

MMS server of robot (PNIU)

Fig. 2. Knowledge Based control of a robot with OSI network

 It would have been also possible to define A simple example of the method can be seen in the
independent procedures to all cases when an Fig. 3. When the user wants to take a box to a
MMS services should be used. In this case the given place he or she should push the button of the
whole MMS functionality would be mouse above the place. A G2 'user-defined-action'
implemented within the gateway. It is the belongs to the place which starts the next-move
solution without any MLGO support. procedure. It will pick up the parameters of a start-
 The third solution reduced the number of the program-invocation MMS service from the
necessary objects and procedures to some MLGO and call the GSI mms-serv external
general terms. It uses limited number of procedure that sends the request to the robot across
external GSI procedure calls also, but give a the gateway. This method is rather general and
rather general (few limitations) interface other MMS Application Enabler programs at the
towards MMS. market are using similar solutions (EasyMAP,
1993).
MLGO
PLACE3, a place
Operation "StartPI"
Color gray Param 3
Weight 30
a place Position 3 Last value -1 6. KNOWLEDGE BASED CONTROL OF AN
Remote appl "pc.mms"
table
step Mms object type prog-inv
ARC-WELDING ROBOT
Mms object name "R-GET"
start next-move Domain ""
(user,PLACE3) Named type "" In the last two years a CAD/CAM integrated robot
programming system (PROARC) was developed in
co-operation with German and Hungarian
participants in the frame of an EC East-West
project (Weck et al, 1995). The goal of the system
call mms-serv(operation, is to assist arc-welding with robots in SMEs mostly
remote-appl,object,domain, for small and medium lot sizes with a PC based,
type, param) across mygsi-1 interactive, off-line robot programming system.
Fig. 3. Realization of MMS within the G2 system Most commercially available systems are either too
expensive, or cannot provide enough features for
arc-welding. CAD/CAM integration means that
after adding the necessary technological especially not for the first two modules. In the
information to the AutoCAD based geometric Modelling Module basically the welding-cell and
design the appropriate programs provide automatic its components are inputed into the system. The
program generation and remote control for the Welding Planning Module provides the capabilities
applied robot to execute the welding tasks. to associate seams to the modelled workpieces and
The PROARC system can be divided into different to generate all movements during the welding
modules: Modelling, Welding Planning, Program session. The Program Generation generates the
Generation and Communication. In this functional robot specific code including some system and
division of the system the different phases of an environment specific features as well. The
offline session are mapped. However, the Communication Module provides the connection
underlying structure is not that separable, to the robot controller.

PROARC system G2
environment
AutoCAD R13 ADS function calls Welding
features
management
AutoCAD requests
Modelling and welding
planning module
Workpiece
STEP file families

RCM3 file Knowledge based

Program generation module control system
with MMS interface

Communication module

MMS network

Fig. 4 Structure of the dual system: PROARC and G2 based KB environment

Fig. 4. shows the connections of the KB based can assist of re-using welding settings or even
control and the PROARC system where the pieces of robot program parts.
original communication module of PROARC has  The present system supports already the
been replaced with the KB control module. This unskilled user with offering default values of
modification will lead us to integrate more parts the different welding variables, but a
into the KB and then the real features and power of knowledge based tool could grow the capacity
the integrated KB tool will appear. of this feature.
There are several possibilities to increase the  Recently the system uses only a simple tactile
‘intelligence’ of the recent PROARC system: sensor to provide safe welding. The intelligent
 The similarities of the workpiece families evaluation of sensor-information could help in
could be managed and analyzed in a more making the system more reliable by means of
sophisticated way even if this feature is a analysis and forecasting procedures.
special one in the PROARC. The intelligent  The area we have already dealt with assists in
comparison of workpieces in the design phase performing the control tasks of the system. It
means that the Communication Module of the
 PROARC system is completely replaced a G2 8. REFERENCES
based control with MMS interface.
As the PROARC is completely PC based and the Buta P, Springer S: Communicating the knowledge in
our recent G2 version is running on SUN some knowledge-based systems, Expert systems with
additional interfacing problems had to be solved, applications, 1992, 5, pp. 389-394
which are not shown on the Fig. 4. The big arrows Brill M., Gramm U.: MMS: MAP application services
show the data flow through the system in the first for the manufacturing industry, Computer Networks and
ISDN Systems, 1991, 21, pp. 357-380
version, the little ones the potential or alternate
Franklin G., Meystel A. et al.: private discussion by e-
connections. mail in Architectures for Intelligent Control Systems
Discussion List, 1994
<AICS-L@UBVM.CC.BUFFALO.EDU>
6. CONCLUSIONS Kovács G. L. et al.: Integrated application of Real-time
expert systems for FMS evaluation and control,
The paper deals with intelligent real-time control of Applications of Artificial Intelligence in Engineering
FMS. As example the simulation of a real FMS VIII , ed. Rzevski, Pastor, Adey, Vol.2, pp. 835-847,
and then a robotics cell (where the robot is defined Elsevier, 1993
as a virtual manufacturing device (VMD)) and Laffey T.J. et al: Real-time knowledge based systems, AI
finally an arc-welding robot controlled by a real- Magazine, 1988, 9, 1, pp.27-45
time computer system are presented. The control Nacsa J., Kovács G.L.: Communication problems of
expert systems in manufacturing environment, in
software is implemented using a real-time
AIRTC'94 (ed. Crespo), pp. 377-381, Preprints of the
intelligent environment (G2). Symposium on Artificial Intelligence in Real-time
An interfacing problem was solved when the G2 Control, Valencia, Spain, 1994, IFAC
system and the MAP/MMS network were Nagy G., Haidegger G.: Object Oriented Approach for
connected to reach real-time intelligent control. Analyzing, Designing and Controlling Manufacturing
These results can be very useful and interesting in Cells, in AUTOFACT’94 (ed. Moody), pp. 12/1-10,
the future when they will be applied for more Proceedings of AUTOFACT’94, Detroit, USA, 1994,
complex and sophisticated large scale SME
manufacturing and assembly systems which hardly Smith P. E. et al: The use of expert systems for decision
can be designed and controlled without using support in manufacturing, Expert systems with
simulation and expert systems. applications, 1992, 4, pp. 11-17
EasyMAP User Guide, CRI Ltd., Denmark, 1993
MMS-EASE User Guide, SISCO Inc., MI USA, 1994
Weck M. et al: A PC-based Offline Programming
7. ACKNOWLEDGEMENT System to Assist Robotized Welding, ICRAM’95, pp.
533-538. Proceedings of the Int. Conf. On Recent
The PROARC project (Ref. No. 7831) was partly Advances in Mechatronics, Istanbul, Turkey, 1995
supported by the EC in the frame of Somlo J., El Buzidi A.: Reverse Scheduling, a New
Copernicus'93. The authors are grateful to the Approach to the Solution of FMS Scheduling Problems,
partners, the WZL of RWTH Aachen and RAAD’96, pp 417—425, Proc. of 5th Int. Workshop on
University of Veszprém for their contributions. Robotics in Alpe-Adria-Danube Region, Budapest,
This work was partly supported by the Hungarian Hungary, 1996
Research Fund (OTKA), projects T 4122 and T
14549.