=Paper=
{{Paper
|id=Vol-1690/paper11
|storemode=property
|title=Working Process Quantification in Factory Using Wearable Sensor Device and Ontology-based Stream Data Processing
|pdfUrl=https://ceur-ws.org/Vol-1690/paper11.pdf
|volume=Vol-1690
|authors=Masao Watanabe,Kazunari Hashimoto,Seiya Inagi,Yohei Yamane,Seiji Suzuki,Hiroshi Umemoto
|dblpUrl=https://dblp.org/rec/conf/semweb/WatanabeHIYSU16
}}
==Working Process Quantification in Factory Using Wearable Sensor Device and Ontology-based Stream Data Processing==
https://ceur-ws.org/Vol-1690/paper11.pdf
Working Process Quantification in Factory
Using Wearable Sensor Device and
Ontology-Based Stream Data Processing
M. Watanabe, K. Hashimoto, S. Inagi, Y. Yamane, S. Suzuki, and H. Umemoto
Fuji Xerox Co., Ltd. Yokohama, Japan
Abstract. A method for quantifying working processes on manufactur-
ing floors was established that uses a wearable sensor device and an
ontology-based stream data processing system. Using this method, the
measurement of manufacturing process efficiency from sensor data ex-
tracted from such a device worn by workers on the job was confirmed at
the Fuji Xerox factory.
Keywords: Ontology, Stream data processing, Wearable device, IOT
1 Introduction
A real-time understanding of the situation at production sites is essential for
management in manufacturing companies. Their ability to exploit unstructured
as well as structured knowledge in the production field determines their manufac-
turing competitiveness. Knowledge in the production field includes three factors:
production machinery, productive resources, and humans engaged in production.
The machinery and resource factors, which are easy to measure quantitatively
and can therefore be projected onto structured data, have been thoroughly uti-
lized in manufacturing management. In contrast, the human factor, which is
unstructured and therefore difficult to sense, has been little utilized so far.
The purpose of this study is to establish a method for quantifying the hu-
man factor on manufacturing floors. We first propose to monitor factory job
processes with wearable sensor devices. The meanings of signals detected by
such devices, however, are not uniquely determined; that is, the same signals de-
tected in different contexts can have different meanings. To define the meaning
of each detected signal uniquely in each context, we further propose to exploit
ontology-based complex modeling.
2 Related Work
Kharlamov et al. [1] have advocated ontology-based data access (OBDA) as a
suitable Web-driven technology for effective and efficient access to data accu-
mulated in production fields. Their approach has focused on the machinery and
resource factors but has not been able to address the human factor.
Real-time data addressing is achieved by a data stream management system
(DSMS) or a complex event processing (CEP) system [2]. These can address
large datasets, reactivity, and fine-grained information access, but they cannot
make complex application domains model heterogeneous data. Stream reasoning
technology can solve this problem [2]. However, a trade-off exists between the
complexity of the reasoning method and the frequency of the data stream of
the reasoner. Morph-streams [3] can ease this trade-off using ontology-based
reasoning and the rewriting of queries using the CEP/DSMS query method. A
barrier to the introduction of Morph-streams is that they require well-understood
operational semantics for DSMS and CEP.
3 Approach
Wearable sensor devices were used and optimized to correspond to human activ-
ity specific to the production field. In order to reduce the barrier to introduction,
a CEP system for semantically filtering stream data was implemented.
3.1 Wearable Device Optimization
Fig. 1. Wearable sensor device and sensed signal
Fig 1(a) shows the wearable sensor device used in this study. The device was
developed by LAPIS Semiconductor in the HAPIC project [4] and is designed
to detect human movements of daily living using 10 types of sensors, such as a
three-axis acceleration sensor, a three-axis gyro sensor, a three-axis geomagnetic
sensor, and a barometric pressure sensor.
The character of motion in a factory work process differs greatly from that
in daily living because factory work requires more effective motion for mass pro-
duction. Fig 1(b) shows the acceleration sensor signal for a factory job process
at different sensing frequencies (10 Hz and 50 Hz). For daily living, 10 Hz is the
parameter value, whereas 50 Hz is the parameter value optimized for this work.
This result shows that the daily living sensing frequency cannot detect the oper-
ations of factory workers. The 50-Hz frequency requires a higher communication
bandwidth, but there is a trade-off between high communication bandwidth and
battery management. We resolve this conflict to optimize the communication
data profile, device profile, and system profile.
3.2 Ontology-Based Stream Data Processing
There are many workers (more than several thousand) and a wide variety of job
processes (more than several hundred) on the production floor. State control of
such huge and complex events is difficult with a conventional DSMS/CEP mod-
ule. We propose a method for semantically filtering stream data to reduce such
complexity. Fig 2 shows the architecture of our system. Sensed data are collected
in the sensor/actuator control hub and transferred to Key-Value stream data.
Key data are created from the time stamp and the device address using RFC4122,
and Value data are created from the sensed raw data. Stream data correspond-
ing to reactivity and scalability are stored in the Key-Value Database (KVDB).
The service module semantically filters the stream data from the KVDB us-
ing ontology-based reasoning and query rewriting to key select queries. The
reduced quantity of stream data can be accessed by a simple CEP module such
as embedded event processing (eep). The CEP module has pre-processing sub-
modules such as a low path filter, feature extraction submodules such as a fast
Fourier transform (FFT), classification submodules such as a hidden Markov
model (HMM), and post-processing submodules.
Fig. 2. Ontology-based stream data processing system
3.3 Ontology
In order to address the working process context for the sensed data, we created
a manufacturing ontology from the operation instruction manual (OIM), which
describes in detail the assembly motions, procedures, and related parts informa-
tion. The manufacturing ontology consists of the basic motions, procedures, and
product component ontology. Manufacturing context reasoning was produced to
match the basic motion ontology to the classified sensed data.
4 Scenario and Results
Many factory workers in Japan have performed an improvement activity called
Kaizen. Conventional Kaizen begins with a working process quantification that
is carried out via job analysis and process time measurements. The working
process quantification requires much time and a high level of skill because the
job analysis requires a deep knowledge of work processes.
We assume a simple scenario for real-time manufacturing process quantifi-
cation as follows: An assembly worker puts the wearable sensor device on his
or her wrist and collects job process data with it. The manufacturing engineer
select the job to improve. System shows directly selected job process efficiency
value (activity efficiency, motion efficiency) and standard value. If job effiency
value differ widely from standard value, they go to improvement process.
We examined the above scenario in the Fuji Xerox Printer assembly line.
The manufacturing engineer queried the job and collected the data, which were
filtered using queries rewritten from it. The filtered data components such as ac-
celeration, gyro, and geomagnetic data were then used to calculate the direction
and magnitude of displacement that occurred during the job process. Activity
efficiency (derived from integrating the magnitude of displacement) and motion
efficiency (derived from the anisotropic ratio of the direction) were able to be
viewed in real time and compared to standard one.
Table 1 shows activity efficiency and motion efficiency value at different pro-
duction lot compared to standard one. This result shows big activity efficiency
dispersion is exist in the job, and that indicate the job process containes redun-
dant motion and needs to improve it.
Table 1. Manufacturing process quantification results
indicator job@1st-production-lot job@2nd-production-lot job@Standard
activity efficiency 58.7 131 100
motion efficiency (%) 55 44 58
5 Conclusion
We confirmed that our method can be used for quantification of a manufactur-
ing process in Fuji Xerox. Our system allows real-time manufacturing process
improvement using human factor information, which has not been satisfactorily
utilized to date. Our next step will be to apply it in other scenarios such as the
creation of links between process quantification data and enterprise data.
References
1. Evgeny Kharlamov, et al. : How Semantic Technologies Can Enhance Data Access
at Siemens Energy. In: ISWC2014 (2014)
2. M. Balduini, et al. : Stream Reasoning for Linked Data. In: ISWC2013 (2013)
3. Jean-Paul Calbimonte : Morph-streams: SPARQLStream OBDA in Action. In:
ISWC2014 (2014)
4. LAPIS Semiconductor Co., Ltd. (http://www.lapis-semi.com) :
HAPIC Project, http://www.coi.titech.ac.jp