Content uploaded by Walter Terkaj
Author content
All content in this area was uploaded by Walter Terkaj on Jan 03, 2017
Content may be subject to copyright.
Available online at www.sciencedirect.com
ScienceDirect
Procedia CIRP 00 (2016) 000–000
www.elsevier.com/locate/procedia
2212-8271 © 2016 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of Scientific committee of the 49th CIRP Conference on Manufacturing Systems (CIRP-CMS 2016).
49th CIRP Conference on Manufacturing Systems (CIRP-CMS 2016)
A telemetry-driven approach to simulate data-intensive manufacturing
processes
Gianfranco E. Modonia,*, Marco Saccob, Walter Terkajb
aITIA-CNR, Institute of Industrial Technologies and Automation, National Research Council, Bari, Italy
bITIA-CNR, Institute of Industrial Technologies and Automation, National Research Council, Milano, Italy
* Corresponding author. Tel.: +39-080-5481265; fax: +39-080-5482533. E-mail address: gianfranco.modoni@itia.cnr.it
Abstract
Telemetry enables the collection of data from remote points to support monitoring, analysis and visualization. It is largely adopted in Formula
One car racing, where streams of live data collected from hundreds of sensors installed on car components are transmitted to the pitwall to be
used as input of real-time car performance simulations. The aim of this paper is to evaluate the potential of a telemetry-driven approach in a
manufacturing environment, where researchers are still looking for efficient methods to perform valuable simulations of the production
processes on the basis of real data coming from the factory. The telemetry could contribute to the implementation of a virtual image of the real
factory, which in turn could be used to simulate the factory performance, allowing to predict failures or investigate problems, and to reduce
costly downtime. This study addresses in particular the efforts to combine and adapt methods and techniques borrowed from the field of
Formula One car racing. Moreover, the investigation of the exploitation possibilities of the factory telemetry is paired with the design of a
software application supporting this technology, starting from the elicitation and specification of the functional requirements.
© 2015 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of Scientific committee of the 49th CIRP Conference on Manufacturing Systems (CIRP-CMS 2016).
Keywords: Factory Telemetry; Factory Image; Cyber-Physical Systems; Simulation
1. Introduction
One of the major issues affecting the current
manufacturing companies is the lack of a full bidirectional
synchronization between the physical world at the shop-floor
and its digital equivalent counterpart (so-called Factory
Image) [1] [2]. When the latter is constantly connected to the
production system, it represents a true reflection of the real
factory which can be used to monitor and simulate the factory
performance, allowing to adjust and optimize processes,
anticipate failures or investigate problems, and thus increasing
efficiency by orders of magnitude and reducing costly
downtime. The current state of the art shows that some
potential solutions are available in literature for transferring
the digital rendered designs to the shop floor in order to build
the right product [3]. However, to the best of our knowledge,
none of the proposed solutions implement the reverse
information flow from the real to the virtual world.
In order to close this loop and provide the full range of
capabilities that synchronization between the real and digital
factory have to offer, a wide variety of different technologies
is needed. A key-enabler is the Cyber-Physical System (CPS)
technology [4], a network of interacting collaborative
elements in constant connection with the surrounding physical
world and its on-going processes. The use of these smart
devices within the manufacturing execution phase allows to
generate a factory telemetry under the form of a large amount
of intensive and multi-source data, which can then be routed
real-time towards various enabled devices connected to the
company network. The major obstacle in the implementation
of such a telemetry approach is the difficulty of handling an
enormous amount of data coming from the real plant to get
aggregate values suitable for the analysis [5].
This paper aims at looking for new strategies to capture,
compare, and view disparate sets of data coming from various
elements of a CPS in order to extract relevant knowledge and
2 Author name / Procedia CIRP 00 (2016) 000–000
provide better insights over the status of the resources at the
shop floor. In this regard, a new vision for processing the
factory telemetry is proposed, by combining and adapting
methods and techniques borrowed from the field of Formula
One (F1) car racing, where telemetry is largely adopted to
collect data from remote points in order to support real-time
car performance simulations. The idea behind this study is
that the envisioned approach allows to generate an integrated
and aggregated view of the factory telemetry that dynamically
augments and enhances the data-driven simulation
applications supporting the manufacturing execution phase.
Finally, the investigation of the exploitation possibilities of
the factory telemetry is paired with the design of a software
application supporting this technology, starting from the
elicitation and specification of the functional requirements.
The remainder of this paper is structured as follows.
Section 2 illustrates the analogies in the use of the telemetry
between the worlds of racing cars and factories. Section 3
introduces an overview of a software application supporting
the factory in industrial scenarios and illustrates major
challenges to realize it. Finally, Section 4 draws the
conclusions, summarizing the main outcomes.
2. Towards a factory telemetry: similarities with the
racing car
Analogies can be a valid way of analyzing the performance
of industrial processes in order to understand potential
improvements. A significant example is represented by the
several parallels between biological and manufacturing
systems that have been drawn in literature to solve a series of
problems of the modern manufacturing, through the study of
the structure, control mechanisms, and functions of the
biological systems [6] [7] [8].
The idea behind this study is that various analogies can
also be observed between the worlds of F1 racing cars and
modern factories. In fact, like the F1 cars, a manufacturing
environment comprises a set of processes to be monitored in
near real-time, huge information flows (and corresponding
software applications) from which to take critical decisions in
limited time, and a team of people that has the task of
developing, maintaining, measuring, and adjusting the system
under changing conditions [9]. On the basis of these analogies
(Fig. 1), it is interesting to experiment a transfer of methods
and tools from one field to the other. In this regard, the focus
of this section is on a set of relevant features of the telemetry,
a proven technology of F1 in which important measurements
are made on board of the cars for data recording and
monitoring (Fig. 2). Such features are then analyzed in order
to investigate the potential of the factory telemetry in the
world of the manufacturing. Moreover, it must be emphasized
that F1 represents a relevant reference case, since it is always
on the cutting edge of technological development.
Fig. 1. Similarities between the worlds of F1 racing cars and modern factories
Fig. 2. Telemetry of a F1 car which contains speed, gear and other channels
(SOURCE: Caterham F1 Team/Renault Sport F1)
2.1. Accurate monitoring of the assets for critical decisions
making
Telemetry is a proven technology of F1 through which a
deluge of data is transmitted from the car to the pitwall in
order to allow a team of engineers to monitor accurately and
constantly several parameters about car systems such as
suspensions, engine, transmission, and wheels [10]. In this
way the engineers can watch over the racing car performance
and optimize the vehicle setup, suggesting drivers to change
one of these parameters. Moreover, they can use the telemetry
to analyze tactics and strategies, investigating on which
corners car could go faster.
The accurate monitoring supported by a similar factory
telemetry would be relevant for any manufacturing company
where data provides the basis for critical decisions making. In
particular, there are two major areas of associated benefit: the
management of the allocated resources, and the continuous
improvement between design, development, and manufacture
of the products (enabling a kind of loop between the three
stages). Along the whole factory life cycle, the sensors
connected with the real factory components can provide
detailed information about the performance of various
processes, ensuring a better visibility and control of the used
resources and a more reliable forecasting. Moreover, a proper
Author name / Procedia CIRP 00 (2016) 000–000 3
integration of the data coming from telemetry and from
enterprise systems such as MES or ERP could help operations
managers to analyze the dynamics of the manufacturing
processes and seek to identify potential improvement actions
(e.g. reconfigurations of the input parameters, changes in the
management of maintenance activities, etc.). In this way it is
also possible to identify and address any bottleneck and
ensure a smart utilization of expensive machineries which
allows to maximize the throughput. Finally, the analysis of the
gathered data enables also the check if the product “as built”
is compliant with the specifications and requirements of the
designers, helping the company to adjust and optimize
processes between design and production stages. Specifically
by merging the designer specifications about how the product
is to be manufactured and the information about how the
product is actually being manufactured, it is possible to build
an instantaneous perspective on how the manufactured
product is meeting its design specification goals.
2.2. Feedback from virtual to real to apply corrective
decisions
The F1 two-way telemetry is a bidirectional data flow that
allows engineers to make real time adjustments remotely on
the car even while the latter is running on the track. In this
way it is possible to align the setup of the car with the needs
of the driver also taking into account external conditions.
From the 2003 season, the two-way telemetry has been
banned from the FIA (Fédération Internationale de
l'Automobile), with the exception of the system for the
activation the DRS (Drag Reduction System), which allows
the driver to adjust the rear wing in order to reduce drag and
increase top speed. In fact, this system is automatically
enabled only in certain circumstances on the basis of the data
coming from the cars telemetry [11].
Similarly, within the factory, a two-way telemetry would
allow project managers and designers to accurately monitor
manufacturing processes progress in real time, enabling them
at the same time to detect problems early (e.g. breakdowns)
and apply corrective decisions based on the information they
receive and analyze. Once these decisions are final, they
would be applied to the real factory, thus implementing the
closed loop between the virtual and real factory [2].
2.3. Integration with Advanced Simulation and Forecasting
Telemetry not only allows F1 teams to collect and monitor
information in real time but also to use them in order to
properly simulate the car for maximizing its performance.
These simulation models have become so advanced that
potential lap time of the car can be calculated, and this time is
what the driver is expected to meet. Moreover, between a race
and another, the F1 teams compute a series of analysis
through which they are able to build predictive models of how
the car will perform with different setups, different tracks
under changing ambient conditions, on the basis of the
collected historical data [12].
If a telemetry-based simulation is used on the factory floor
next to the machineries that it models, it could give operators
a digital representation that looks and acts exactly like the
machine itself. In this way it can offer the capability to
execute the operations through a simulation environment
where the various product components can be inserted and
tested in different configurations across the entire production
chain. Under these conditions, operators can optimize and
validate new processes into state-of-the-art machine, without
taking the latter out of production. In order to realize this
approach, the data telemetry should be fully integrated with
discrete or continuous simulators, which allow to model the
complex dynamics of a manufacturing system. The latter can
refer to the processes of a single cell, a production line, an
entire factory, or several companies interconnected with the
warehouses through a network. Another key success factor of
the approach is the capability to initialize the simulation
models through a snapshot of the real system [5].
2.4. Digital continuity between telemetry historical data
The simulation-based analysis of the F1 car performance
mentioned in the previous subsection can be effectively
exploited only if the digital continuity between telemetry
historical data is guaranteed. Indeed, it must be ensured that
data can be playbacked and passed as input to the simulation
tools in order to perform forecasts against which to compare
the behavior of actual running real-time systems. Digital
continuity is also important from a reliability point of view,
since statistics based on historical data make sure that
installed components not exceed their recommended lifetime
ranges [12]. Finally, digital continuity plays an essential role
in case of an accident, since FIA can determine driver errors
as a possible cause on the basis of the driver inputs that have
been recorded.
Similarly, digital continuity between historical data of
factory telemetry allows to create numerous simulated data
streams that are semantically interoperable with real
operational data. Such data emulation offers a real-world
environment to train personnel, where for example control
room operators can directly interact with the system and
receive real feedback [13]. Specific analytics have to be
performed on the gathered information to extract better
insight over the progress and status of each single machine.
These analytics can provide comparison between machine
performance. Moreover, historical information can be
measured to predict the future behavior of the allocated
machineries.
In order to guarantee the digital continuity between
historical data, Terkaj et al. [14] proposed to use an history
model of factory objects. In this way, historical data can be
collected and stored in a distributed way, while keeping an
overall coherence thanks to a common virtual factory model.
2.5. In situ simulations
The seamless integration of simulation tools and the real
environment of the factory paves the way to in situ simulation
approaches, which takes place in the working environment
and involving those who work there. The in situ simulation is
4 Author name / Procedia CIRP 00 (2016) 000–000
distinct from center-based simulation, which is performed in a
context separated from the work environment [14].
A similar philosophy can be found in F1 behind the driving
simulator, which is a car cockpit that gives drivers true feel of
a real environment and direct feedback on their actions. The
driving simulator replicates real race track conditions and is
used to test different aspects that affect performance of the car
such as wings and brake settings. The high fidelity of the
simulator allows the driver to feel the difference that
modifications applied to the car setup can produce without the
high acceleration of a real test drive. As the new FIA
regulations limits the number of test days on the track and
also wind tunnel time to reduce costs and level the playing
field, the driving simulator plays a key role for drivers
training, saving at the same time both time and money while
respecting new regulations. Moreover, the driving simulator
can be used to test future car designs and train new drivers on
different circuits.
3. Factory Dynamic Simulator: A dedicated software for
telemetry analysis
A typical infrastructure supporting the telemetry data flow
from the real to the virtual factory should include three main
components (Fig. 3). The first is an embedded controller unit
enabling sensor data collection and logging and corresponds
to the first level of the 5C Architecture for CPS introduced by
Lee at al. in [15]. The second component is the
communication module, which commutates dynamic data
read from real components into a real-time transmission
stream. The third component is a software application that
receives, interprets, persists, integrates, and analyzes the
collected data. This section focuses in particular on the
requirements elicitation of the third component. During this
activity, a valid starting point can be the evaluation of existing
F1 telemetry systems, such as Atlas [16], which is the
standard system, or Wintax [17].
The following list highlights the major features that a
software application supporting factory telemetry should
provide: capability to maintain the links between factory
configurations/layouts and telemetry data; simulation of the
effects of different input parameter values on a given factory
process; and direct comparison of simulated results with real
telemetry data or with other simulations.
Data visualization is an essential task of the envisioned
software tool. XY Charts, waveform and scattered plotting,
statistics and animations permit to show under different views
the data telemetry acquired from the sensors which are
connected to the real factory. In this way, it is possible to
study accurately a particular aspect of the factory. Among the
most significant graphical features, the new environment
should include functionalities to filter and select a part of the
collected data stream in order to provide it as input of a new
simulation. Using a multiscale model as reference, the
envisioned software application should also comprise
capabilities to zoom in and zoom out the selected data in order
to drill down into specific data subsets. Moreover, the
Graphical User Interface should also provide facilities to
change the factory setup which comprises the different input
parameter values for the proper configuration of the factory
processes; the setup can be stored to a database in order to be
used as input for a following simulation. Each new created
setup should be compliant with the previous already saved
setups, allowing in this way to guarantee their Digital
Continuity, as discussed in the previous section.
A typical issue of the data coming from sensors is the noise
errors. As it is better to have a smooth curve to analyze the
factory performance, removing high frequency noise and
spike is a necessary feature for the envisioned software
application. In this regard it is essential to use various
techniques of high frequency noise removal such filtering and
smoothing [18]. Also, the end-users should have the
possibility to introduce a sensor offset/gain or implement a
sensor correction. Combining the digital versions of telemetry
signals and a lot of math/logical/filter/statistical functions,
also through the integration with external commercial tools
such as Excel, Matlab and Simulink, it is possible to create the
so-called virtual channels, which represent a method to
abstract and remap the original telemetry channels (for
example to create alarms). A proper API (Application
Programming Interface) should guarantee the access to
telemetry data, enabling data analysis in external tools (e.g.
Matlab). Finally, a Multicast transmission of the data over the
factory network would allow the software application to
receive the telemetry regardless of the PC where the software
application runs, as long as it is connected to the network and
enabled.
Fig. 3. The components of a typical infrastructure supporting telemetry
4. Conclusions
This paper has highlighted various analogies between the
worlds of F1 racing cars and modern factories. On the basis of
these analogies, the paper has analyzed the benefits of the
technology transfer of the telemetry, proven in F1, to the
manufacturing field. In particular, it is shown that the
exploitation of the factory telemetry could offer various
methods to perform valuable simulations of the production
processes, using as input the data coming from the real
factory. Moreover, the requirements of a software application
supporting the factory telemetry have been elicited. Further
developments of this study will address the difficulty in
Author name / Procedia CIRP 00 (2016) 000–000 5
integrating the telemetry with existing simulation packages,
the semantic interoperability of the data coming from
heterogeneous sensors, and the need for more efficient and
scalable databases for Big Data storage [19].
Acknowledgements
The research reported in this paper has been funded by the
European Union 7th FP (FP7/2007–2013) under the grant
agreement No: 314156, Engineering Apps for advanced
Manufacturing Engineering (Apps4aME).
References
[1] Grieves, M.. Digital twin: manufacturing excellence through virtual
factory replication (2014). [Online]. Available from
http://www.apriso.com/library/Whitepaper_Dr_Grieves_DigitalTwin_Ma
nufacturingExcellence.php. [retrieved: Jan, 2016].
[2] Kádár B, Terkaj W, Sacco M (2013) Semantic Virtual Factory supporting
interoperable modelling and evaluation of production systems. CIRP
Annals Manufacturing Technology, 62(1):443-446.
[3] Ben Khedher, A., Henry, S., & Bouras, A. (2011). Integration between
mes and product lifecycle management. In Emerging Technologies &
Factory Automation (ETFA), 2011 IEEE 16th Conference on (pp. 1-8).
IEEE.
[4] Monostori, L. (2014). Cyber-physical production systems: Roots,
expectations and R&D challenges. Procedia CIRP, 17, 9-13.
[5] Kádár, B., Lengyel, A., Monostori, L., Suginishi, Y., Pfeiffer, A., &
Nonaka, Y. (2010). Enhanced control of complex production structures by
tight coupling of the digital and the physical worlds. CIRP Annals-
Manufacturing Technology 59.1 (2010): 437-440.
[6] Christo, C., & Cardeira, C. (2007, June). Trends in intelligent
manufacturing systems. In Industrial Electronics, 2007. ISIE 2007. IEEE
International Symposium on (pp. 3209-3214). IEEE.
[7] AlGeddawy, T., & ElMaraghy, H. (2010). Co-evolution hypotheses and
model for manufacturing planning. CIRP Annals-Manufacturing
Technology, 59(1), 445-448.
[8] Ueda, K., Vaario, J., & Ohkura, K. (1997). Modelling of biological
manufacturing systems for dynamic reconfiguration. CIRP Annals-
Manufacturing Technology, 46(1), 343-346.
[9] McCullen, P., Saw, R., Christopher, M., & Towill, D. (2006, June). The
F1 supply chain: adapting the car to the circuit–the supply chain to the
market. In Supply chain forum: an international journal (Vol. 7, No. 1, pp.
14-23). KEDGE Business School.
[10] Cocco, L., and P. Daponte (2008). Metrology and formula one car.
Instrumentation and Measurement Technology Conference Proceedings,
2008. IMTC 2008. IEEE.
[11] Federation Internationale de l'Automobile, 2011 formula one technical
regulations (2010). Tech. Rep., Section 3.18. [Online]. [retrieved: Jan,
2016].
[12] Waldo, J. (2005). Embedded computing and Formula One racing.
Pervasive Computing, IEEE 4.3: 18-21.
[13] Capozzi, F., Lorizzo, V., Modoni, G., & Sacco, M. (2014). Lightweight
Augmented Reality Tools for Lean Procedures in Future Factories. In
Augmented and Virtual Reality (pp. 232-246). Springer International
Publishing.
[14] Terkaj, W., Tolio, T., & Urgo, M. (2015). A virtual factory approach for
in situ simulation to support production and maintenance planning. CIRP
Annals-Manufacturing Technology, 64(1):451-454.
[15] Lee, J., Bagheri, B., & Kao, H. A. (2015). A cyber-physical systems
architecture for industry 4.0-based manufacturing systems. Manufacturing
Letters, 3, 18-23.
[16] Advanced Telemetry Linked Acquistion System [Online]. Available
from http://www.mclaren.com/appliedtechnologies/products/item/atlas/.
[retrieved: Jan, 2016].
[17] Wintax4 [Online]. Available from
http://www.magnetimarelli.com/business_areas/motorsport/software/wint
ax4. [retrieved: Jan, 2016].
[18] Vaseghi, S. V. (2013). Advanced signal processing and digital noise
reduction. Springer-Verlag.
[19] Modoni, G. E., Sacco, M., & Terkaj, W. (2014, June). A survey of RDF
store solutions. In Engineering, Technology and Innovation (ICE), 2014
International ICE Conference on (pp. 1-7). IEEE.
