Managing Production Complexity with Intelligent Work Orders
Ville Toivonen, Eeva Järvenpää and Minna Lanz
Laboratory of Mechanical Engineering and Industrial Systems, Tampere University of Technology,
Korkeakoulunkatu 6, Tampere, Finland
Keywords: Intelligent Work Order, Decentralised Production Systems, Process Configuration, Information Management.
Abstract: Progress of Industrial Internet of Things is rapidly increasing the amount of data collected from manufacturing
operations. This data can be utilized to control and improve production systems in various ways. Production
control systems play a key role in realizing the potential cost savings and productivity increase. Companies
are required to manage increasing complexity while shortening response times to changes. A concept of
Intelligent Work Order (IWO) is proposed to assist in these challenges. It supports local or distributed
decision-making, and decreases integration complexity between different factory IT-systems. IWOs also
increase information visibility at the shop floor. The IWO structure and functionality are described with a
discussion of the benefits of the approach.
1 INTRODUCTION
Production processes are facing increasingly higher
demands for planning and control that start to
resemble one-off production. The Web-based
consumer trade is pushing the requirement of minimal
lot-size down as the customers want and are capable
for selecting or even configuring digital product
orders at the level of detail that so far has been viable
for large companies placing orders for big production
lots or for skilled sales personnel specifying high-cost
products, such as cars or kitchen furniture, for and
according to the customer.
Allowing the increase in the product
configuration variability also increases the need for
more integrated IT and manufacturing systems along
the ordering-production planning-manufacturing
chain. Likewise, more flexible and accurate
monitoring and control systems on the factory floor
are then needed for maintaining the product quality
and efficiency, ensuring the human-machine safety
under the increased variability in the production
process, and managing the significantly more
complex logistics associating each individual product
to its components, production schedule and the
customer data.
The industrial recognition of the need to increase
digitalization and to introduce novel control systems
has led to German Industry 4.0 initiative (Brettel et
al., 2014) and other similar approaches. A survey
conducted in Finnish manufacturing industry
(Järvenpää et al., 2015) showed similar interest from
the companies but also a large gap between the
academic concepts and the status of the control
system implementations. In this paper a concept of
Intelligent Work Order (IWO) is presented, which is
an effort to bridge that gap and show how existing
Manufacturing Execution Systems (MES) can be
extended to meet the future requirements of
production control systems. In Section 2,
decentralized production control systems are
discussed in general and in relation to IWO. Section
3 presents the structure and functionality of IWO,
while Section 4 is dedicated for information systems
integration. Benefits of IWO are summarized in
Section 5, before closing remarks of Section 6.
2 DECENTRALIZED
PRODUCTION CONTROL
Commonly stated reasons for distributing production
control are managing complexity and gaining the
ability to react quickly to production changes or
disruptions in the system. Decentralized control
systems are built on local decision making and lack a
holistic view over the whole system. In agent based
control systems, such as in (Caridi and Cavalieri,
2004), a coordination process is designed to diminish
Toivonen V., JÃd’rvenpÃd’Ãd E. and Lanz M.
Managing Production Complexity with Intelligent Work Orders.
DOI: 10.5220/0006507801890196
In Proceedings of the 9th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (KMIS 2017), pages 189-196
ISBN: 978-989-758-273-8
Copyright
c
2017 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
the negative effects of having a narrow view of the
whole system, while in holonic manufacturing
systems (HMS) the decision makers are only
connected to higher level controllers (McFarlane and
Bussmann, 2000). This means HMS are always
hierarchical, while the form of agent based systems
can vary more freely. These approaches are
applicable both in a single manufacturing site or a
supply chain (Saharidis et al., 2006). Some often
noted benefits and hindrances of centralized and
distributed control policies are listed in Table 1
(Zannetos, 1965) and (Toivonen et al., 2011).
Decentralised control systems are sometimes
viewed to contradict any kind of forward planning of
production. More practical view is to plan on
aggregate level and allow a decentralised control
system enough decision making power to take care of
execution. Another integrating approach between
planning and control is to have the decentralized
control system forecasting future events (Valckenaers
and Brussel, 2005).
3 INTELLIGENT WORK ORDER
Intelligent work order provides one possible
infrastructure for realising an agent based production
control system in real environments. In this case IWO
can be an information agent or sometimes both an
information agent and a controller. Control decisions
(such as resource allocation and job dispatching) are
done based on the information gathered by the
intelligent work order. In order to fully realise a
decentralised control system, IWO needs to fulfil the
following set of minimum requirements:
Ability to recognize and communicate with
resources
Ability to create work orders for subtasks
Ability to communicate with a higher level
controller
Inclusion of a decision making control logic
Intelligent work order collects information from
several corporate IT-systems and can also be used to
collect and communicate process information for
various purposes.
Table 1: Comparison of decentralized and centralized control hierarchies.
Decentralized Centralized
Benefits Benefits
Increased flexibility due to self-configuration that
also enables efficient response to changes.
Provides decision-making power to lower level
managers or workers with better experience on local
processes.
Efficient development of the local operations and
processes.
Detailed and up-to-date information.
Clearness of goals and responsibilities.
Expanded responsibility and decision-making
authority which often result in increased work
satisfaction, motivation and efficiency.
Lower costs for organizational and transaction
management.
Fast dissemination of information.
Fast decision-making due to one author
controller.
Consistent processes and practices.
Combined control system that enables fast
review of resource allocation.
Disadvantages Disadvantages
Higher costs for organizational and transaction
management.
Possible lack of coordination among autonomous
managers.
Lower level managers decisions could be
damaging, because they do not necessarily have full
understanding of the wider perspective.
Lower level managers may have goals that are
different from the goals of the entire supply chain.
Lower level managers may make decisions that are
not in the organization’s best interests.
Inflexibility that is caused by the complexity of
information systems and organization structure.
Weak ability to respond to changes that is caused
by the complexity of organizations.
Challenges that result from information validity
and integrity
In a strongly centralized organization process
optimization is difficult due to the complexity of
the organization’s structure.
Lack of collaboration that is caused by a
centralized control system which is managed by
a dominant authority.
3.1 IWO Functionality
Traditionally a work order describes either the
process or the end result of the required task. These
work orders are commonly delivered to the factory
floor in paper format (Järvenpää et al., 2015), which
means they have limited information content and are
difficult to maintain. Intelligent work order is digital
and contains up-to-date information of both process
(e.g. instructions, NC-programs) and output (e.g.
specifications, quality control guidance) in a machine
and human readable format, as shown in Figure 1.
IWO is role- and context dependent. It can be
configured based on the operators’ personal
characteristics, preferences and experience. For
instance, the way to present the work instructions may
be modified based on the operator’s native language
and experience.
Figure 1: Principle of Intelligent Work Order.
IWO is created only at the time it is needed, after
which it calls for resources required in the process.
This approach is intended to ensure that changes
made in ISA-95 level 3 and 4 planning systems
(ANSI/ISA-95.00.03-2005) are automatically
considered in the process level. After completion of
the task, process data is aggregated to a desired level
of detail and stored into the memory of IWO. Without
aggregation a lot of the data acquired from the process
might not be directly very useful (e.g. signal values)
to a higher level planning system. The collected
process information can later on be viewed from
resource, time, customer or product point-of-view.
This allows managers to link process information
with businesses processes in a meaningful way.
3.2 IWO Structure
IWO is an information element that has a clearly
defined lifespan from start to the completion of a
specific task. Figure 2 shows a generic IWO structure
and information content that has been deducted from
the requirements. A digital work order is essentially
an information distribution agent that contains all
relevant process information. The additional features
of IWO require a more diverse structure.
Communication interface allows user specific views
to information and the realisation of online
negotiation or alarm systems. Data collection
interface tries to standardise some of the factory floor
data exchange and decrease the related integration
effort. Processing unit contains the local controller in
distributed systems and can be utilised to perform
some operations and reduce complexity of centralised
control systems. Data storage is a short term storage
location for work order data and feedback from the
manufacturing process.
The information content of IWO is closely related
to the element structure. In addition to the process
data, IWO contains metadata defining how
information is shown, collected and distributed:
Parent – Defines the process structure
Process Information Task specific information
e.g. work instructions and NC-programs
Information Distribution Defines connections
and integration to other IT-systems
Information Collection Defines reporting
activities and connections and integration to the
control logics
Processing and Acting Decision making and
communication block that allows raising alarms,
calling other services etc. This part defines the low
level controlling logic.
The information content should be fairly similar
for any digital work order. The main difference in
IWO is the processing block that allows it to be used
for decision making.
Figure 2: IWO Structure.
3.3 IWO Creation and Storing
IWO constructor is responsible for creating and
configuring work orders. The main information
content of work orders exists in MES and other higher
level information systems but the configuration can
also be based on the status of the production system.
As an example, a work order could be configured to
always link with a machine that has the shortest setup
time.
IWO constructor needs integration to systems that
maintain the work order related information. While
system integration is not discussed in detail in this
Section, differences of the two main approaches
should be noted. Integrating the constructor solely
with MES is cheaper and easier than creating
connections to several higher level systems.
However, it is important to understand the depth of
the existing MES integration. Alarming systems and
other real-time control methods require a deep
integration to ensure up to date information and rapid
response capability. If the existing MES
implementation contains mainly historical data for
aggregate planning or the information content is
otherwise limited, the constructor might require
additional integration effort to shorten the time delay
of information updates.
Table 2 summarises the elements required for
creating IWOs.
Table 2: Key elements for creating IWOs.
Constructor
Contains a tool for defining
IWO content, see IWO
structure for details
Data storage for created IWO
models
Integration to valid IT-
systems for collecting
information
Information
element (IWO)
A digital work order for
production tasks
Short-term storage for
documenting the process
Communication ability to
real-time applications
Distribution cente
r
A database or a server
Long or mid-term data
storage for process
information
An interface for information
systems to access process
information
After a task has been completed, results are
collected and stored for later use. The proposed
concept does not describe what should be done with
the process data after it has been stored. However, the
information should be made available for different
IT-systems in a way that does not depend of the MES
integration.
Figure 3: Control hierarchy.
4 SYSTEM INTEGRATION
The approach presented here is meant to reduce
integration complexity and improve system
connectivity while the related industrial standards are
still being under discussion and development. In
order to achieve this, the factory floor integration has
been reversed to a bottom-up process as shown in
Figure 3. This means that production process defines
how work orders are brought to and how information
is collected from the process. In addition to the lower
level integration layer, IWOs form a layer of
abstraction between the higher level information
systems and the process data. System complexity
increases in the lower levels of control hierarchy
while the higher level system interfaces stay usually
fairly constant. One important aspect of IWO is to
provide accurate process level information for later
analysis and development purposes. Mapping the
feedback information and sensor data against IWO
combines process plan, end result and resource,
environment and customer data into a single
information object.
The link between two work orders describes a
process chain where a result from a prior task affects
the configuration of the IWO in a later task. The direct
communication between work orders is more relevant
when one is functioning as the controller of subtasks.
The implementation of the system defines how much
of the controlling power is given to IWOs and which
decisions are made by a higher level controller. The
higher level systems are used for planning ahead,
which means that the utilized data is always
somewhat aggregated and delayed.
4.1 Data Collection
In manual reporting the data must often be written
down first and later entered into the system -
sometimes by a different person than the one who
recorded it in the first place. Typographical and
transcription errors are common. Once these errors
become part of the data set, they become difficult to
detect and eradicate, making all the resulting reports
less reliable. At the shop floor of manufacturing
companies, there are thousands of operations without
any link with the current Enterprise Resource
Planning (ERP) and MES solutions. In addition to
that, the relationship between shop floor actions and
data input into MES is in majority of cases off-line.
This means that workers are feeding information of
processed jobs manually and often with a time lag to
the actual events. That is why, traditionally ERP and
MES solutions cannot be used to process real time
information from shop floor or to compare real
actions with planning. The lack of real time
information into the MES and ERP systems are
leading to a very low rate of response from the
management team, which is leading to a high
percentage of scrap, useless consumption of raw
materials and energy, increase of waste in process,
and nevertheless losses for SME’s.
The long and continuing effort of removing waste
from manufacturing systems should be directed to
information management as well. This does not only
involve the system side but even more importantly the
practices of producing and consuming information in
the factory floor. In this approach a large portion of
the information management procedures are defined
in the process, allowing similar streamlining and
continuous process development as has been
executed with Lean methods in manufacturing. This
is different from the typical approach where the IT-
systems are defined top down and users are
constricted to those definitions.
4.2 Integration of Individual
Controllers
The trend of digitalization is rapidly increasing the
amount of available data from the factory floor. At the
same time, the complexity and cost of MES
integration are also increasing in the same fashion.
This means many companies create subsystems or
‘digital islands’ in domains where the direct benefits
of digitalization are greatest. In manufacturing
industry this can be seen for example in supplier
specific automated solutions that can offer great
flexibility but are difficult to integrate under a single
controller. IWOs can be used in two different ways to
improve the situation:
1. IWO can assist in connecting these islands
by providing a common communication
interface. In this case defined information
can be exchanged between individual
systems, but there is no controller linking
them as an integrated system.
2. Both systems provide an integration to allow
IWO to function as a controller. This takes
more effort but allows extended automation
and process control. In such an approach
IWO can be seen as an integration platform
to reduce the complexity and costs of
integration.
In a similar fashion, IWOs can be utilized in
human-machine interfaces to reduce human input to
the machine controller and to communicate the
current status and future actions of the machine to the
worker. Flexible integration is especially important
for increasing the level of automation in industries
where it has been difficult.
5 IWO VALUE PROPOSITION
Some of the main benefits and potential applications
of IWO are listed here as a summary of the concept.
The list is categorised to follow timeline of a process;
planning, execution and reporting. MES
implementations are always company or
manufacturing site specific, which means that the
following list is suggestive at best:
Quality of Information
Several practices have been considered to
improve the information quality of IWOs. IWO
creator is integrated to the original source of
information as closely as possible. This is meant to
enforce ‘one owner policy’ and reduce harmful data
replication to several systems. From the process side
the aim is to reduce manual input by increasing
automated data collection and also to give control
over reporting to the process where information is
created.
Targeted Quality Control
One benefit of the Intelligent Work Order is the
ability to reconfigure work process in real-time. This
can be utilised for example in quality control
practices. In case of deviations, such as appearance of
a defected part, an additional inspection can be
assigned for each task that involves a part from the
same batch. IWO allows to focus corrective measures
quickly and precisely to the orders that might be
affected.
Real-Time Process Control
All information needed to accomplish tasks in the
workstation is delivered automatically to the
workstations without the need of worker
involvement. The information of task status and
realized production is always available for production
planning and control, which enables fast reaction to
possible variations and disturbances. The application
of rapid response in quality control can also be
extended to increase system responsiveness in
general. IWO allows order and task specific process
configuration, which means even individual
production orders can be controlled in real-time.
Worker Specific Work Instructions
IWO allows user based configuration of the
process information. In practice this can mean, for
example, choosing the language of work instructions
based on the recognised worker. Similarly new and
less experienced employees might get more detailed
instructions and be required to sign more additional
verifications during the process. Context aware
applications can also involve physically readjusting
the work place to assist in the operation.
Improved Traceability
Mechanical engineering industry has lagged many
other industries in building traceability chains
through their supply chains. IWOs provide a
systematic approach for recording in-house actions in
detail and as such improve product traceability.
Automated data collection
One key driver of digitalization is to reduce
manual handling of information. This is essential for
reducing mistakes and shortening the time delays
between events and information availability. IWO
supports introduction of novel reporting and sensor
technologies by simplifying their integration to the
MES.
Development and Analytics
The information generated during the production
processes, e.g. recordings, task duration,
measurement and quality data, and other data
collected by various sensors, are linked to the
intelligent work order. Analysing existing products
and processes is facilitated by linking the product,
resource and operational process data. The generated
expressive information object can later be used for
different analytics and for increasing the accuracy and
quality of planning and control.
6 DISCUSSION
We have proposed a concept of intelligent work order
(IWO) to tackle increasing complexity, to improve
real-time control and to allow a better integration
between different factory IT-systems. The trend of
digitalization has increased the interest of
manufacturing industry in similar approaches and
industrial implementations do exist. We believe the
additional effort required for MES development and
integration has a very short payback time in most
manufacturing environments that have a dedicated
production control system in place. The concept
should be applicable to different control hierarchies
and MES implementations, and allow cherry picking
the benefits that are chosen as key drivers for the
investment.
The concept is providing tools for closing the gap
between current MES implementations and the future
needs from the control systems. There is a recognized
need for convergence of factory IT and operational
technology. Tools and methods are required
especially to facilitate integration of legacy hardware
in factories. These should allow both collecting real-
time information of production processes and moving
decision making power closer to the process while
maintaining a holistic view of the production.
This research has started from interests of our
industrial partners and we are hoping for this work to
contribute to industrial adaptation of the presented
ideas and digitalization of manufacturing industry in
general. At the moment, industrial demonstrations are
being planned in order to further advance this
development. Interesting applications could also be
found in rapidly developing fields such as
collaborative robotics.
ACKNOWLEDGEMENTS
This research was carried out as part of the Finnish
Metals and Engineering Competence Cluster
(FIMECC)’s MANU programme in the LeanMES
project.
REFERENCES
ANSI/ISA-95.00.03-2005, Enterprise-Control System
Integration, Part 3: Models of Manufacturing
Operations Management.
Brettel, M., Friederichsen., N., Keller. M., and Rosenberg.
M., 2014. How virtualization, decentralization and
network building change the manufacturing landscape:
An Industry 4.0 Perspective. International Journal of
Mechanical, Industrial Science and Engineering, Vol.
8, No. 1, pp.37-44., 2014.
Järvenpää, E., Lanz, M., Tokola, H., Salonen, T., and Koho,
M. 2015. Production planning and control in Finnish
manufacturing companies Current state and
challenges, In Proceedings of the 25th International
Conference on Flexible Automation and Intelligent
Manufacturing, Wolverhampton, UK.
Caridi, M., & Cavalieri, S. 2004. Multi-agent systems in
production planning and control: an overview.
Production Planning & Control, 15(2), 106-118.
McFarlane, D. C., & Bussmann, S. 2000. Developments in
holonic production planning and control. Production
Planning & Control, 11(6), 522-536.
Järvenpää, E., Lanz, M., Tokola, H., Salonen, T., & Koho,
M. 2015. Production planning and control in Finnish
manufacturing companies–Current state and
challenges. In Proceedings of International Conference
on Flexible Automation and Intelligent Manufacturing
(FAIM), Wolverhampton, UK.
Saharidis, G. K., Dallery, Y., & Karaesmen, F. 2006.
Centralized versus decentralized production planning.
RAIRO-Operations Research, 40(2), 113-128.
Toivonen, V., Väistö, V., Perälä, T., and Tuokko, R. 2011.
Decentralized Production Planning and Control in a
Collaborative SME Network, In Proceedings of the 21st
International Conference on Flexible Automation and
Intelligent Manufacturing (FAIM), Taichung, Taiwan.
Valckenaers, P., & Van Brussel, H. 2005. Holonic
manufacturing execution systems. CIRP Annals-
Manufacturing Technology, 54(1), 427-432.
Zannetos, Z. S. 1965. On the theory of divisional structures:
Some aspects of centralization and decentralization of
control and decision making. Management Science,
12(4), B-49.