|
Part Cost in
e
Cost in
e
Com.
FTS
2
5000
10000
-
RTP
1
2500
2500
-
Software
1
10000
10000
Custom
PC
1
2000
2000
-
Input devices
2
250
500
Low cost
Total
25000
-
Annual Costs
Expense
Calculation
Value
Unit
Com.
Investment
-
25000
e
-
Service life
T
5
yr
-
Depreciation
DC
= I/T
5000
e/yr
-
Interest
I
= 0.09I/2
1125
e/yr
9%
/ rate
Maintenance
M
6000
e/yr
-
Total
AC
= D + I + M
12125
e/yr
-
Benefit Calculation
Expense
Calculation
Value
Unit
Com.
Programmer
P
100
e/hr
-
Time saved (%)
T p
45
%
-
Time saved (hr)
T s
= (1/1 − T p) − 1
0.82
hr
-
Usage
U
250
hr
/yr
-
Cost benefit
BC
= P.T s.U
20455
e/yr
-
Payback
period
I
/(BC − AC + DC)
1.87
yr
-
Table 8.3: Details of a benefit calculation based on the third scenario
128
8.5 Programming, sensors and intelligent robots - The big picture
The latter assumptions are translated to an annual financial benefit of 20,455
e. Hence the
payback period amounts to 1.87 years. This however is only an example of the cost benefit
expected from utilizing the approach. A more detailed and representative calculation
should take several other factors into consideration. For instance how complicated is
the given task in terms of motion and force requirements? Is the task planned for batch
production or small size lots? This would result in a very good estimate of the frequency of
reprogramming i.e. the yearly usage and therefore the economical benefit. Additionally, it
is imperative to determine whether a specialized robot programmer (for CIR) is required?
or is it possible that a normal robot programmer do the job?. Such factors constitute the
backbone of any benefit calculation and significantly influence the decision whether or not
to apply the WPBA.
8.5 Programming, sensors and intelligent robots - The big picture
Industrial robots are known for their flexibility compared to other manufacturing machines.
This potential however was restricted by the class of tasks envisioned for them from the
start. By limiting their application domain to spray painting and welding, which require
accurate positional capabilities, their full potential as a human replacement in many other
tasks was not realized. The historical development of industrial class robots set the stage
for the current situation of industrial robots in production facilities. One of the main
factors playing a role in said development was the fact that they were only a
ffordable
for companies producing large volumes (Bilbao et al. 2005). However, the constraints
on the development and deployment of industrial robots are beginning to fade, primarily
due to two reasons; the decreasing costs of robots (compared to human workers) and
the increasing demand on customized products . These factors play an implicit role in
the trends sweeping the robotics research nowadays. Decreasing costs -of automation in
general while increased costs of manual work,- render robots more attractive for a new
segment of production facilities which was not conceivable a decade ago (International
Federation of Robotics - IFR Statistical Department 2009). Additionally, the fact that
they will be utilized to replace manual work is forcing the manufacturers to develop smaller,
more compact and safe robots, in order to allow for direct human-robot collaboration
(Kr¨uger & Surdilovic 2008)(Kr¨uger et al. 2009). The second factor outlines a major
challenge facing production engineering worldwide. Customers hungry for tailored
products are increasingly pushing the boundaries of production lines and their capability
to manage the accompanying complexity (Scholzreiter & Freitag 2007). Despite these
factors, the gap between what is currently needed and what is available w.r.t. deployment of
robot-based production lines, has yet to decrease. By providing simpler user interfaces and
hence enhanced programming capabilities this gap could definitely be bridged. However,
it is the author’s view that the interfaces should not be limited to over-designed GUI or to
abstracting a limited number of functionalities to an input device. But rather, the designers
should -in a holistic view- reconsider the hardware and software components and their
interactions together. In other words, reevaluating their basic approach to industrial robots
129
8 Assessment
as pure positional machines or as enablers of mechanical tasks. This way, designers can
tightly integrate di
fferent sensors in simple and reliable ways to increase the bandwidth
of possible tasks. Additionally, this would allow a sensor to be utilized in di
fferent
capacities during programming and operation. For instance, a camera can be used during
programming to acquire the shape of the object, while during operation to detect the
existence of the same object in a given area. Furthermore, enhanced sensoric capacity
also means implicitly increasing the robots ability to interact to uncertain conditions
during execution in a feel-reason-act loop (Caccavale et al. 2005). Such a development
would lead to blurring the line between programming and operation while providing the
robots with self-optimization capabilities and more independence. By extrapolating this
development path, intelligence is eventually achieved by creating robotic systems capable
of reacting in an autonomous and flexible manner to continuously changing environmental
conditions
2
(Gausemeier 2005)(Gausemeier et al. 2009). What the author advocates here
is attaining intelligence through an incremental approach which takes into consideration
the complexities of standardized interfaces and operation procedures in manufacturing
facilities and not by attempting to achieve intelligent robots with one step. Despite that the
approach and the test-rig developed in this thesis pertain to a certain class of robot tasks
(tightly coupled CIR), the discussion here formed the spirit of the research done in this
work.
2
This can be also termed self-optimizing mechatronic systems
130
9 Conclusion
Do'stlaringiz bilan baham: |
