Advance in Mechatronics
Advance in Mechatronics
Advance in Mechatronics
Advance in
Mechatronics
CHEN,LIANG-XUAN
National Kaohsiung University of Science and
Technology
Department of Mechatronics Engineering
Taichung, Taiwan
[email protected]
I. INTRODUCTION
Abstract— Mechatronics does not have Mechatronics is “the application of
a definite definition; However, microelectronics in mechanical engineering”
Mechatronics could roughly be defined as (the original definition suggested by MITI of
an interdisciplinary engineering with a Japan).
synergistic combination of mechanical Previously, mechatronics just meant
engineering. The portmanteau complementing mechanical parts with some
"Mechatronics" was first coined by Mr.
electrical units, a typical representant being
Tetsuro Mori, a senior engineer of the
a photo camera. Today, mechatronics is an
Japanese company Yaskawa, in 1969.The
area combining a large number of advanced
major advantages of Mechatronics
techniques from engineering, in particular
Systems are that they are simpler,
economical, reliable and versatile systems sensor and actuator technology, with
when integrated than being operated as computer science methods. Figure 1 depicts
individual systems. Due to these major the three areas of mechatronics and their
advantages, its field of application is very overlap.
vast like Automotives, Defence, Medical, Typical examples of mechatronic systems
Smart consumer products, are automotive applications, e.g. advanced
Manufacturing, etc. Cars, CD players, braking systems, fly/steer-by wire or active
washing machines, railways are all suspension, but also DVD-players or
examples of mechatronic systems. The washing machines. Mechatronic systems are
main characteristic (and driving force) of characterized by a combination of basic
recent advances is the progressively mechanical devices with a
tighter coupling of mechanic and processing unit monitoring and
electronic components with software. controlling it via a number of actuators and
In this paper we survey current sensors. The introduction of mechatronics is
developments and discuss future trends in a tight integration of mechanical, electrical
mechatronics, the future of mechatronics and information-driven units.
will specifically see a move towards a high
degree of adaptability and self-
organization.
Keywords— Mechatronics, Actuators,
Automation, Miniaturization,
Modularization, etc.
2022 25th International Conference on Mechatronics Technology (ICMT)
(electrical/electronic technologies).
The second half of the last century saw
dramatic changes in technological processes
and products owing to the rapid extension of
earlier successes in electronic technologies
towards the development of a bewildering
Fig. 1 Areas of Mechatronics array of digital computational units
(computers): general purpose integrated
II. CHANGES IN THE NATURE OF TECHNOLOGICAL
PROCESSORS AND PRODUCTS
chips (IC), application specific ICs (ASIC),
microprocessors (µp), etc. These functional
In the era prior to the invention of the
units are now so small in size (miniaturized)
electromagnetic induction dynamo (1830-
that they can be embedded within the
40) by Michael Faraday, all “machines”
functional units.
(technological processes and products) were
mechanical (M) in nature, i.e., composed III. STATE OF ART
essentially of mechanical units. Since Modeling & Tools: In a certain sense,
mechanical units exhibit large inertia, modeling and even model driven
machines of this era tended to be large, development, i.e. the generation of
cumbersome, slow, “uni-functional” and executable code from a model, has long been
“non-user friendly (difficult to control and existing in the mechatronic world to improve
maintain)”. However, it sufficed for the software quality based on model analysis.
innovators of such machines to be well Code Generation: Based on such a
versed in mechanical sciences and arts. specification, model based development
By the late 19th century, since electrical
ideally requires the generation of code which
(E1) energy can be transmitted and
transformed much more easily than
meets all real time constraints. This requires
mechanical energy, the energy receiving and the code generator to know about all
manipulating units within machines platform specific constraints like speed and
(technological processes and products) started number of processors or available memory.
to be replaced by functionally comparable Only a very few research oriented
electric units. As a result, machines became approaches exist to support a uniform
more compact, controllable and user-friendly. modeling of the behavior of all system
A technological transformation occurred components including the specification of
with the advent of analog electronic (E2) real time constraints and a corresponding
valves in the earlier half of the last century. code generation.
This transformation accelerated after the Processes: The above description focused
1950s owing to the development of on modeling the software part of
transistors, digital electronics and power mechatronic systems.
electronics (E3). Wherever possible, One of the most prominent problems in
electrical functional units were replaced by current industrial development and even
such electronic units so as to attain several research approaches is however the lack of
orders superior performance in terms of size, integration between the different disciplines,
controllability and user-friendliness. The namely mechanical and electrical
synergistic combination of E1, E2, and E3 engineering and computer science or
technologies may be collectively referred to software engineering more specifically.
as E technologies Usually, the mechanical engineer starts with
2022 25th International Conference on Mechatronics Technology (ICMT)
designing the shape and mechanical parts, models in this class are written as timed
then the electrical engineer plans the wiring automata, and a number of tools support
and finally the software engineer has to write verification of timed automata with respect
the code. This approach leads to a lot of to reachability or even temporal logic
design errors and costly rework when it is specified properties. Automation can still
finally noticed that some parts do not fit only partially be achieved; the algorithms
together or the simple layout of processors employed in the model checking are not
and memory make certain software solutions guaranteed to terminate anymore. In order to
impossible. make the actual system fit into the required
Analysis & Tools: A rather large subclass, approximations of the real system
percentage of mechatronic systems are are used.
deployed in safety critical areas (e.g. the
automotive or rail domain). This makes
analysis of mechatronic systems (or first of
all, their models) one of the main areas of
work for software engineers employed in the
design of such systems. Since its invention
in the late 80’s model checking has become
a standard technique for verification, in Fig. 2 Composition of Mechatronics system
particular for hardware systems.
The main advantage of model checking IV. FUTURE DEVELOPMENTS AND CHALLENGES
which makes it interesting for mechatronic
We believe that future mechatronic
systems is its (almost) full automation,
systems will consist of several
providing tool support for analysis.
autonomously acting agents capable of
Notwithstanding recent advances and
monitoring their own physical environment
success stories, the main challenge is still the
as well exchanging information with other
socalled state explosion problem: model
agents. Constructing the software of
checking techniques (most often) rely on a
advanced systems requires a number of
search of the whole state space, and this can
significant changes of current software
grow to arbitrarily large dimensions. For
engineering techniques. In particular, the
Example: SAT solvers are combined with
following issues have to be addressed to
decision procedures (giving socalled SMT-
build the next generation systems properly.
solvers), model checking with specific AI
search methods, bounded model checking is A. Current software design processes
parallelized or model checking combined are tailored towards a particular domain
with static analysis methods. rather than spanning over all involved
Mechatronic systems present a further domains.
challenge for verification as they belong to B. Modeling formalisms allow for a
the area of hybrid systems, characterized by description of static systems but not for their
a combination of discrete and continuous volatility. Model transformations are meant
parts. The software constitutes the discrete for transforming models towards a particular
part, while the continuous dynamics use on a platform but not for describing the
corresponds to the physical system with its change in that model.
sensors and actuators. Verification of hybrid
systems today is still in its infancy. System
2022 25th International Conference on Mechatronics Technology (ICMT)
v. Performing the physical task using limbs engineers will have to possess deeper
(actuators). understanding of natural sciences so as to
A remarkable human ability is to learn cope with the growing variety of sensors.
from the results obtained from past acts so as
to perform better when executing similar
tasks in the future. This learning ability
provides human beings with the ability to act
as autonomous units. A further ability lies in
communicating with other human beings so
as to undertake collective tasks. The above Fig. 4 CNC Machine
description of human abilities provides a
basis for understanding trends in B. Machine learning: Intelligence means
mechatronics. adapting to the environment and improving
performance over time. Within the domain
A. Sensing and sensor fusion (task ii) will of mechatronic engineering, “there has been
be the next capability to be acquired by considerable interest in learning through the
mechatronic systems. Already, many use of ANN and fuzzy logic for applications
mechatronic units possess rudimentary in control and robotics, autonomous guided
sensing abilities. For instance, modern air vehicles (AGV), etc., that require mainly
conditioning units are able to sense air reflective intelligence when performed by
temperature and humidity through separate human operators and tasks, such as machine
sensors and fuse the signals through fuzzy diagnostics, requiring combinations of
logic reasoning. Likewise, sensors in the reflexive intelligence and low level
form of transducers have long been used to reflective intelligence.” This interest will
enable feedback control in machines. continue well into the future.
However, there is still a long way to go.
Sensors produce copious amounts of data
that need to be digested to discover patterns
of interest before control can be effected
through the “actuators”.
Advances in high-speed microcomputers
and signal processing algorithms have now Fig. 5 Example of sensing and sensor fusion: Robots
opened the door for the exploitation of C. Autonomization refers to the
sensors exploiting a wide range of physical, development of the ability to survive and
chemical and, even, biological phenomena. perform robustly while the external
While actuators are limited in variety, the environment changes. With progress in
variety of possible sensors is almost sensor and learning technologies,
unlimited. tomorrow’s mechatronic devices can be
For instance cutting forces in CNC expected to become progressively more
machining (Figure 4) and its consequences autonomous. They will be able to reset their
(e.g., tool fracture) can today be monitored local goals autonomously under changing
and controlled using commercially available external environments so as to meet the
devices capable of sensing machining noise, broad system-level goals set by human
machine vibrations, acoustic emission, drive beings.
motor current, etc. Future mechatronic
2022 25th International Conference on Mechatronics Technology (ICMT)