The Future of Machines: Self-Aware Control Systems

thunderclingΤεχνίτη Νοημοσύνη και Ρομποτική

13 Νοε 2013 (πριν από 3 χρόνια και 10 μήνες)

178 εμφανίσεις

2. http://zone.ni.com/devzone/cda/pub/p/id/1305

The Future of Machines: Self
-
Aware Control Systems

Machine builders have made advances in developing technology that can
complete repetitive tasks with great speed. See how you can integrate the next
generation of machines into your control systems.

When examining machine
-
industry trends, you often encounter new controller
technologies that increase the performance and throughput of high
-
end
machines, motor technologies, or energy
-
efficient algorithms,

or you learn
about tools that help lower the cost of machine design. Over the last few
decades, machine builders have made considerable advances in developing
machines that can complete repetitive tasks with ever
-
increasing speed. There
are other trends a
nd technologies, however, that might have an even more
significant influence on the next generation of machines and the way those
machines are integrated in your work process.

After spending decades optimizing machinery equipment speed, the industry is
run
ning into new limiting boundaries. High speeds and operating machines
running at maximum load are increasing the wear and tear of mechanical
components and tools. This increases the importance of maintenance and
systems that ensure uptime. Additionally, ma
ny tasks in the industry are not
purely repetitive. Solutions to application problems such as picking randomly
shaped parts out of a bin are far from realization. Last, but not least, several
manufacturing processes still involve a significant amount of wo
rk by humans.
The machine industry needs to address safety concerns that arise when
humans work alongside machines and robotic systems.

The availability of data and information about the environment, processes, and
machine parameters is crucial to addressi
ng these new machine industry
challenges. Therefore. sensors and measurement technology that can acquire
this information are playing a significant role for the next generation of
machines. The sensor market was very static for decades, but the last few
ye
ars have brought substantial innovation. Sensor technology advancements
have been adopted into many electronic devices, from smart phones to home
automation systems, and prices have dropped to all
-
time lows.

You can use sensors to create systems that are a
ware of their environment,
perform real
-
time process monitoring, and know every detail about their
mechanical component health. However, sensors alone are worth no more
than the control systems of the past. The key to solving new challenges is
creating con
trol systems that can integrate sensor data, gather information in
real time, and use information from multiple sensors within high
-
speed control
loops. High
-
performance embedded systems with industrial
-
grade ruggedness,
such as

NI CompactRIO
programmable automation controllers (PACs), provide
direct sensor connectivity through modular I/O devices. You can use the
reconfigurable

field
-
programmable gate array

(FPGA) to prepr
ocess sensor
data even before the information is transferred to the real
-
time processor,
which executes custom control or monitoring tasks programmed in the

NI
LabVIEW

graphical development environment.

So how wil
l sensor and measurement technology change the future of
machines?


First, by integrating advanced measurement and sensor technology
into

mechanical systems and machines, you can implement machine
condition monitoring applications. Condition
-
based mainten
ance systems help
decrease unscheduled outages and optimize machine performance while
reducing maintenance and repair costs. Additionally, you can use
measurement and sensor technology to increase machinery and equipment
safety and provide the control syst
em with information about system health at
any time during operation. You can add machine condition monitoring tasks
through embedded subsystems that are connected to the control system via
network technologies or that are integrated into the control syste
m as another
task.


Figure 1.

CompactRIO PACs offer connectivity to hundreds of sensors and
actuators, such as in this application from

Oregon State University
.

Next, with a more seamless integrati
on of sensor information in the controller
you can create self
-
healing or adaptive machines and systems that can adjust,
for example, to the changing characteristics of mechanical components. By
enabling the control system to identify these changing condit
ions caused by
mechanical component wear and tear, you can implement routines that gather
data during a startup process or during the machine’s operation and use
adaptive control algorithms, which improve machine operation.

Finally, by integrating advanced

sensors you can develop dynamic machines
that adapt to changing environmental parameters and monitor process
parameters to ensure near
-
perfect manufacturing results. With additional
knowledge about the process, the machine can detect imperfections or
chan
ges in the raw material, adjusting for vibrations that might appear within a
machined part or dealing with tasks that are different for each of the processed
parts or iterations of the process. Good examples are medical machines and
devices, such as cell
-
s
orting systems or DNA sequencers that must process
unpredictable structures such as naturally growing cells. These systems need
to heavily adapt their process based on information they gather from imaging
sensors, camera systems, or other advanced sensor s
ystems.


Figure 2.

You can use high
-
performance

NI Smart Camera

systems to create
vision
-
guided robotics applications, such as this one from

Ve
traco
.

The increased adoption of measurement and sensor technology for control
applications will also revolutionize the robotics machine industry. If you look
closely at the well
-
orchestrated movement of a line of welding robots, you see
that they
simply perform the same predefined movements again and again.
You need advanced sensor technology to adapt industrial robotics systems
(which have been used in industry for decades) to perform more advanced
applications, such as picking parts from a bin or

handling delicate parts and
goods. You can use tactile sensors, light detection and ranging (LIDARs), and
camera systems to give robots the ability to detect the presence and location
of objects and humans, or control the amount of force and pressure appl
ied to
parts and goods. This opens up new application areas for robotics.

Sensors have always been part of high
-
end machines. With new technologies,
decreasing prices, and high
-
performance control systems with the capability to
incorporate and process info
rmation from multiple sensors, you can create the
next generation of machines and devices that are fully aware of their
environment and monitor all machine and process parameters so that they can
adapt to changing conditions.