THE USE OF SIMULATION IN ACTIVITY-BASED COSTING FOR FLEXIBLE MANUFACTURING SYSTEMS

Proceedings of the 1997 Winter Simulation Conference
ed. S. Andradóttir, K. J. Healy, D. H. Withers, and B. L. Nelson
THE USE OF SIMULATION IN ACTIVITY-BASED COSTING FOR
FLEXIBLE MANUFACTURING SYSTEMS
Soemon Takakuwa
School of Economics
Nagoya University
Furo-cho, Chikusa-ku, Nagoya-shi, Aichi, 464-01 JAPAN
ABSTRACT
type of flexible manufacturing system. A procedure for
cost accounting is developed for obtaining the
A framework to design simulation models is described in
manufacturing costs by utilizing simulation results. In
order to perform activity-based costing for flexible
addition, it is concluded that cost reduction could be
manufacturing systems before actual manufacturing
achieved by increasing operation time per day, by
activities.
analyzing the contents of the manufacturing costs.
For illustrating a procedure to perform activity-based
costing, analysis of the random-access type of the
2 THE FLEXIBLE MANUFACTURING SYSTEM
flexible manufacturing system (FMS) is performed from
both efficient and economic standpoints, by simulating.
The flexible manufacturing system (FMS) completely
The flexible manufacturing system considered in this
controls both material and information flows in an on-
paper consists of NC machine tools (i.e., one NC lathe,
line, real-time mode, and is particularly suitable for a job
one turning center, and two machining centers), one
shop concerned with production of versatile items
washing machine, the AS/RS, and AGVs. Workparts are
(Hitomi 1979). The use of machining centers with
machined at the assigned NC machine tool(s) in the
adaptive control devices will increase flexibility of the
predetermined order of operations, and then transferred
system and machine utilization.
by AGVs. Set-up operations are performed by the
The flexible manufacturing system considered in this
industrial robot located inside the AS/RS.
paper comprises four NC machine tools (one NC lathe,
In this study, a simulation model for a FMS is
one turning center, two types of machining centers), one
constructed. Then a procedure for cost accounting is
washing machine, two AGVs, one AS/RS with an
developed for obtaining the unit cost of the products
industrial robot, and so forth (Takakuwa 1995).
through simulation experiment. It is shown that precise
Workparts are transferred by one of the AGVs, and they
cost accounting can be performed before actual
are set up by one industrial robot inside an AS/RS.
manufacturing activities, if kinds of designated products
Furthermore, each workpart is loaded and unloaded by
and their production quantities are specified.
each industrial robot associated with each machine tool,
before and after machining. (The animation layout for
1 INTRODUCTION
this system is shown in Figure 1 (b).)
Operation sequences and the associated processing
Accurate determination of cost components for the
times for the selected workparts are summarized in Table
operation of a factory can play an important role in that
1. As an example, let us trace the movement of product
factory’s success. However, current methods of
1 shown in Table 1. The operation sequence is a NC
accounting in use in most factories do not always
lathe, a turning center, a vertical type of machining
describe or allocate costs accurately. One major reason
center, a washing machine, a horizontal type of
is variance of cost between predicted and actual figures.
machining center, and again a washing machine. Every
In this study the problem mentioned above will be
workpart first comes out from the AS/RS and awaits an
resolved by performing simulation experiments.
available AGVs. Then it is moved to the NC station
Including costs as well as times on every operation
(i.e., the first station). After machining at the station, it
provides an effective implementation of cost modeling
is moved to the turning center (i.e., the second station).
methodologies on the factory floor.
Before machining at each of the two machining centers,
Cost analysis and cost accounting before performing
each workpart should be set up for machining by the
actual operations are performed for a random-access
industrial robot inside the AS/RS.
793

---

794 Takakuwa
Table 1: Operations Sequence and Processing Time
difference between the realized value and the predicted
value of the cost and so forth. Reducing this source of
Part type
NC lathe Turning
V. machining H. machining Washing
variance requires development of a better prediction
center
center
center
machine
Product 1 1 2 3 5 4,6
model. In this paper, a simulation model is adopted as
116.4 306 604.7 793.8 180(x2)
such a prediction model.
Product 2 1 2 3
255.6 510 180

Product 3 1
3.3 Simulation-Based Cost Accounting
386.4

Product 4 1
291.6

Now, current cost-accounting methods are unable to
Product 5 1
account for many of the costs incurred in production
666.6
Product 6 1
because of the difficulty in tracking the parts through the
605.4
entire production operation. It is almost impossible to
Product 7 1 3 2,4
estimate accurate performance of the production
791.4 853.2 180(x2)
Product 8 1 2
operation without simulation. Simulation provides the
796.2 180
ideal tool for cost estimating since it provides a complete
Product 9 1 2
864 180
summary of production activity.

One of the most common means of allocating costs
[Upper: the sequence of operations; Lower: the processing time(sec.)]
to products has been based on proportioning them on
direct labor costs, even though in modern highly
mechanized systems direct labor costs may only be a
small fraction of total cost (Kaplan 1984). A simulation
3 COST ACCOUNTING WITH SIMULATION
can also be used as a basis for defining and calculating

these costs as the product moves through the system.
3.1 Manufacturing Costs
These costs may be associated with stations in the
simulation model through which the product passes and
Manufacturing costs comprise direct costs and indirect
may include the contribution to cost of added materials,
costs in production (the following items are added for
of set-up transactions, of frequency in handling, and
the model of the flexible manufacturing system in this
other detailed considerations that realistically add to the
study).
cost of manufacturing (Zuk et al. 1990).
(1) Direct cost
Production/cost information can be obtained through
direct materials: material cost
simulation as follows.
direct labor: N/A
(1) Information on machining
direct expenses: N/A
machining time (for each machine)
(2) Indirect cost
number of finished parts
indirect materials:
processing time
cost of cutting oil
tool-replacement time
cost of cutting edges
etc.
indirect labor: wages
(2) Information on materials handling
indirect expenses:
travel time
depreciation (machine tools/equipment)
work-in-process inventory
electricity charge
etc.

taxation
In this study, the simulation is performed by SIMAN

(Pegden et al. 1994); however, the basic idea of this
3.2 Cost Variance
procedure can be applied to other simulation languages.

Another important issue is where the determination

Regarding the cost variance, there are at least six sources
of costs should occur when modeling: during the
of variance: (1) inappropriate standard, (2)
simulation or after the simulation. Cost determination
mismeasurement of actual results, (3)
should occur after simulation, because the
implementation breakdown, (4) parameter prediction
“unallocatable” costs are not as well known during the
error, (5) inappropriate decision model, and (6) random
simulation as they are after the simulation.
variation (Horngern and Foster 1987). Each source may
call for different corrective actions, and all of these
3.4 Activity-Based Costing
corrective actions are costly. Especially, regarding the
parameter prediction error, planning decisions are based
Regarding the hierarchy of factory operating expenses,
on predictions, such as future costs, future selling prices,
four activities are separated in activity-based costing,
and future demand. In many cases, there will be a
that is, unit-level activities, batch-level activities,

---

The Use of Simulation in Activity-Based Costing for Flexible Manufacturing Systems 795
product-sustaining activities, and facility-sustaining
manufacturing cost. While overhead as a percentage of
activities (Cooper and Kaplan 1991). Firstly, expenses
total manufacturing costs has steadily increased, the
for unit-level activities consist of direct labor, materials,
percentage of direct labor content has decreased (Ruhl
machine costs, energy, and so on. Secondly, expenses
and Bailey, 1994).
for batch-level activities consist of setups, material
movements, purchase orders, inspection, and so on.
4 SOME IMPORTANT ASPECTS OF
Thirdly, product-sustaining activities consist of process
EFFICIENCY AND COST ANALYSIS
engineering, product specifications, engineering change
notices, product enhancement, and so on. Finally,
Before performing cost analysis it should be stressed that
facility-sustaining activities consist of plant
there exist some important aspects of efficiency and cost
management, maintenance of the building and grounds,
analysis. In this section, issues on “product-mix/
heating and lighting , and so on. Unit-level activities and
machine loading” and “scheduling effect” are selected
batch-level activities could be examined through
especially to be examined, by performing simulation
simulation from among those activities. Therefore, these
experiments.
issues are treated solely for analysis in this study.
The procedure of applying fixed costs to products
4.1 Product-Mix and Machine Loading
through a cost markup percentage, based on some
reasonable measure of activity in a department
Since available resources for production, such as
(machine-hours in fabrication, labor-hours in assembly),
machines and labor, are limited for each individual
had its origins in the financial accounting requirement to
manufacturing firm, it is desirable to effectively allocate
allocate all production costs to items produced. This
and utilize those production resources which determine
system works well at the aggregate level of financial
the optimal kinds and quantities of products to
statements – to obtain values for inventory and cost of
manufacture.
sales – and is generally inexpensive to operate.
Now, suppose that we want to produce some kinds of
However, the system can produce enormous errors in
products from among nine products shown in Table 1
attributing the consumption of production resources to
with the production resources in one shift of operation,
individual products (Kaplan and Atkinson 1989).
i.e., eight hours. To solve this product-mix and
In an activity-based system, the cost of a product is
requirements problem, the following linear programming
the sum of the cost of all activities required to
model is obtained:
manufacture and deliver the product. The allocation
bases used by activity-based cost systems are termed
Maximize 20.1 x1 + 19.8 x2 + 6.70 x3 + 6.70 x4 + 6.50 x5
cost drivers. A variety of cost drivers can be used to
+ 6.50 x6 + 1.91 x7 + 8.60 x8 + 15.8 x9
trace volume-unrelated costs, including:
subject to
n Setup hours.
116.4 x1 + 386.4 x3 + 291.6 x4 28,800
n Number of setups.
306 x1 + 255.6 x2 + 666.6 x5 + 605.4 x6 28,800
n Material handling hours.
604.7 x1 + 510 x2 + 791.4 x7 + 796.2 x8 28,800
n
793.8 x
Number of times handled.
1 + 853.2 x7 + 864 x9
28,800
n
360 x

1 + 180 x2 + 360 x7 + 180 x8 + 180 x9
28,800
Ordering hours.
x
n
1 - x2 = 0
Number of times ordered.
x 0 : integer
n Part number administration hours.
n Number of part numbers maintained.
The objective function is the total profit gained, and
Managing costs across the firm means managing the
28,800 time units (seconds) are available for machine
costs incurred before the product is manufactured
resources. In addition, the last constraint means that the
(upstream costs, i.e., research and development, and
numbers of product 1 and product 2 should be same for
product design, and so on), while the product is
production to put them together.
manufactured (manufacturing costs), and after the
The optimal solution to the above-listed problem is
product is manufactured (downstream costs, i.e.,
obtained as follows: x1* = 25, x2*=25, x4*=88, x6*=23,
marketing, distribution, customer service, and so on).
x9*=9 (pcs.). However, the time needed to produce all
Total manufacturing cost is the sum of the cost of
these products would be approximately 41,000 to 61,000
materials, labor, and applied overhead. If manufacturing
seconds, by performing simulation, and it is found to be
overhead were a negligible portion of total product cost,
much more than 28,800 seconds. Hence, all these
misapplication of manufacturing overhead would not be
products cannot be produced definitely within one shift
a concern. However, in a business environment
of operation.
characterized by high technology manufacturing,
overhead cost is a large percentage of total

---

796 Takakuwa
4.2 Scheduling Effects
system is summarized in Table 3. Miscellaneous costs
such as system controllers are assigned adequately to the
In the dynamic situation, workpieces arrive at the shop
corresponding items listed in Table 3. The service life is
randomly over time, so that the shop itself behaves like a
needed to calculate depreciation for each equipment unit.
network of queues. In this context, scheduling is
There are two major stages: (1) simulation and (2)
generally carried out by means of dispatching decisions:
activity-based costing in the procedure. The scheme of
at the time a machine becomes free a decision must be
this procedure is indicated in Figure 1. By using this
made regarding what it should do next. These
generative system, activity-based costing would start
scheduling decisions are unavoidable in the operation of
automatically without inputting any data, immediately
such a system (Baker 1974). In this section, seven
after a simulation experiment is finished and the
scheduling rules , i.e., LWKR(Least Work Remaining),
corresponding Excel file is opened. Thus all of the
MWKR (Most Work Remaining), TWORK (Total
required calculation will be made by the system.
Work), LTWORK (Least Total Work), FCFS (First
Come First Served), SPT (Shortest Processing Time),
and FASFS (First Arrival at the Shop First Served) are
Item Price Service life
applied to all queues in the system, when performing
($1,000) (years)
simulation.
NC lathe 373.6 10
A list of maximum flow times to process the set of
Turning center 436.7 10
products (i.e., 25 pieces of product 1, 25 pieces of
V. Machining center 461.6 10
product 2, 88 pieces of product 4, 23 pieces of product 6,
and 9 pieces of product 9) obtained in the previous
H. Machining center 501.1 10
section is summarized in Table 2, adopting each
Washing machine 175.8 10
scheduling rule. It is observed that the maximum flow
AGV system 1,025.6 12
time under the MWKR (Most Work Remaining) rule is
AS/RS 732.5 10
the minimum among those under any other scheduling
Tool management 293.1 10
rules in this case. This result demonstrates substantial
Total 4,000.0 -
differences on the maximum flow time by applying the
dispatching procedure; it is important to seek out the
Table 3: Price and Service Life of the System
decision rules that promote good performance.
Figure 1 illustrates the general procedure to execute
Table 2: Maximum Flow Time under Major Scheduling
simulation experiment with the resultant external files
Rules
and generate a series of activity-based costing tables in
one Excel file. One external file is required to perform
simulation in advance, as shown in Figure 1 (a). This
(sec.)
file contains all data on the machining time for each
Scheduling rule Maximum flow time
specified workpart with each cutting edge at each
LWKR (Least Work Remaining) 51,778
MWKR (Most Work Remaining) 41,402
machine tool. The animation layout for the flexible
TWORK (Total Work) 51,796
manufacturing system is shown in Figure 1 (b). The
LTWORK (Least Total Work) 41,408
numbers of finished workparts are indicated on the
FCFS (First Come First Served) 61,078
screen as well.
SPT (Shortest Processing Time) 48,044
After simulation is performed, one external file will
FASFS (First Arrival at the Shop First Served) 61,078
be generated; it contains the summary of machining time
with each cutting edge at every machine tool. The
external file is shown together with the ARENA
5 SEMI-GENERATIVE PROCEDURE OF
summary report in Figure 1 (c). Immediately after the
ACTIVITY-BASED COSTING WITH
corresponding Excel file is opened, the required data
SIMULATION
obtained by simulation will be automatically inputted to
the file. Four sheets of the Excel file are selected and
5.1 The Procedure
shown in Figure 1 (d). Then, activity-based costing will
be performed, by filling the required cells of a series of
A semi-generative procedure of activity-based costing
Excel sheets sequentially and automatically. Finally, the
with simulation is proposed in this study, especially for
unit manufacturing cost for each product will be
the flexible manufacturing system described in section 2.
obtained in the final sheet of the Excel file. Thus
The price (numerical examples) and the service life of
activity-based costing will be done, by making use of
each system component of the flexible manufacturing
simulation results.

---

The Use of Simulation in Activity-Based Costing for Flexible Manufacturing Systems 797
(a) External File (Input)
(b) Performing Simulation
Figure 1: A Semi-Generative Procedure of Activity-Based Costing with Simulation (Continued)

---

798 Takakuwa
(c) Summary Report and External File (Output)
(d) Semi-Genetive Activity-Based Costing System
Figure 1: A Semi-Generative Procedure of Activity-Based Costing with Simulation

---

The Use of Simulation in Activity-Based Costing for Flexible Manufacturing Systems 799
5.2 Cost Accounting for Various Types of Products
manufacturing costs of these five types of parts are
approximately $31.93, $17.27, $7.69, $11.16, and
Table 4 shows the detailed contents of cost accounting to
$27.73, respectively, as shown in Table 4. It is found
produce the specified numbers of pieces of the five types
that precise activity-based cost accounting can be
of products (i.e., product 1: 25 pcs., product 2: 25 pcs.,
performed before actual manufacturing activities. If the
product 4: 88 pcs., product 6: 23 pcs., and product 9: 9
more reasonable ratios could be adopted for allocating
pcs. shown in Table 1 in this case), when applying the
fixed costs, the values might be substituted by them.
SPT scheduling rule on every queue such as the queue
Applying seven scheduling rules described in section
for AGVs at the exit of AS/RS. Although there are a lot
4.2, activity-based costing is performed for each
of subsidiary tables as shown in Figure 1, major results
scheduling rule, together with corresponding simulation
on accounting are summarized in Table 4.
experiment. The unit manufacturing costs for all
Table 4 comprises two major parts, that is, fixed
products are summarized in Table 5. It is found that the
costs and variable costs. A fixed cost is a cost that
manufacturing cost might vary, depending on the
remains unchanged in total for a given time period
adopted scheduling rule.
despite wide changes in the related total activity or
volume. In this case, depreciation, cutting oil, electricity
6 SUMMARY
charge, and so on, are classified into fixed costs. On the
other hand, a variable cost is a cost that changes in total
A framework of the semi-generative procedure of
in direct proportion to changes in the related total
activity-based costing with simulation is proposed.
activity or volume. For example, material costs are
Activity-based costing analysis of the random-access
variable costs. The contents of all cost items are
type of the flexible manufacturing system (FMS) is
summarized and used for calculating the unit
performed from both efficient and economic standpoints,
manufacturing costs of products.
by performing simulation; and numerical examples are
The term Percentage of value added stands for the
demonstrated, based on an real flexible manufacturing
relative ratio of value added for each product which is
system. In addition, it is shown that precise cost
produced through this manufacturing system. In this
accounting can be performed by utilizing simulation
case, these values are used as the rates for allocating
before actual manufacturing activities are performed, if
fixed costs (Sakurai 1995). Now, it is assumed that the
kinds of designated products and their production
values of 39, 19, 11, 12, and 19 are assigned in
quantities are specified.
percentage for each product respectively. The unit
Table 4: Summary of Activity-Based Costing in Applying SPT Rule
Item NC lathe Turning
V.machining H.machining Washing
AS/RS AGV system Management System
Total
center
center
center
machine
administration
Cutting oil ($) 1.68 1.25 1.20 1.17 0.83 0.00 0.00 0.00 0.00 6.13
Depreciation(equipment) ($) 49.78 58.19 61.51 66.77 23.42 97.60 113.88 0.00 39.05 510.21
Depreciation(building) ($) 3.29 3.29 3.29 3.29 3.29 3.29 3.29 3.29 3.29 29.61
Electricity charge(fixed) ($) 4.88 3.62 3.50 3.41 2.40 1.03 0.90 0.13 0.13 20.00
Taxation ($) 1.98 2.21 2.21 2.21 0.89 3.97 4.28 0.00 0.00 17.77
Total(fixed cost) ($) 61.61 68.57 71.71 76.85 30.83 105.89 122.35 3.42 42.48 583.72
Total(to be allocated) ($) 61.61 68.57 71.71 76.85 30.83 105.89 122.35 3.42 42.48 583.72
Wages ($) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 131.51 131.51 263.01
Electricity charge(variable) ($) 27.37 20.33 19.61 19.10 13.47 5.78 5.05 5.05 5.05 120.81
Total(variable cost) ($) 27.37 20.33 19.61 19.10 13.47 5.78 5.05 136.55 136.55 383.82
Total(to be allocated) ($) 88.98 88.89 91.32 95.95 44.30 111.67 127.41 139.98 179.03 967.54
Item Product1 Product2 Product3 Product4 Product5 Product6 Product7 Product8 Product9 Total
Percentage of value added (%) 39 19 0 11 0 12 0 0 19 100
Total(to be allocated) ($) 227.65 110.91 0.00 64.21 0.00 70.05 0.00 0.00 110.91 583.72
Total(variable cost) ($) 164.59 71.36 0.00 77.47 0.00 42.03 0.00 0.00 28.36 383.82
Cutting edges ($) 256.12 99.60 0.00 7.11 0.00 6.62 0.00 0.00 56.33 425.79
Material cost ($) 150.00 150.00 0.00 528.00 0.00 138.00 0.00 0.00 54.00 1020.00
Manufacturing cost ($) 798.36 431.87 0.00 676.79 0.00 256.70 0.00 0.00 249.60 2413.33
Quantities (pcs.) 25 25 0 88 0 23 0 0 9 170
Unit manufacturing cost ($/pc) 31.93 17.27 0.00 7.69 0.00 11.16 0.00 0.00 27.73 -

---

800 Takakuwa
alone?,” Harvard Business Review, March-April
Table 5: Comparison of Unit Manufacturing Cost
1986: 87-93.
Kaplan, R. S. and A. A. Atkinson, 1989. Advanced
Management Accounting, 2nd edition, Prentice Hall,
($/pc.)
Inc., Englewood Cliffs, New Jersey: 191-194.
Scheduling rule
Product1 Product2 Product4 Product6 Product9
Pegden, C. D., R. E. Shannon, and R. P. Sadowski. 1994.
LWKR 33.16 17.86 7.80 11.56 29.20
Introduction to Simulation Using SIMAN, 2nd ed.,
MWKR 30.45 16.57 7.56 10.68 25.97
McGraw-Hill, Inc., New York, New York.
TWORK 33.17 17.86 7.80 11.56 29.20
Ruhl, J.M. and T. A. Bailey. 1994. “Activity-based
LTWORK 30.47 16.58 7.56 10.69 25.99
FCFS 34.99 18.74 7.97 12.15 31.38
costing for the total business,” The CPA Journal,
SPT 31.93 17.28 7.69 11.16 27.73
February 1994: 34-38.
FASFS 34.99 18.74 7.97 12.15 31.38
Sakurai, M. 1995. Management of Indirect Costs, Chuo
Keizai-sha, Ltd., Tokyo: 95. (in Japanese)
Takakuwa, S. 1995. Economic Analysis of FA/CIM,
ACKNOWLEDGMENTS
Chuo Keizai-sha, Ltd., Tokyo: 80-91. (in Japanese)
Zuk, J. S. et al. 1990. “Effective cost modeling on the
The author wishes to thank Mr. K. Hirano of Kodo
factory floor: taking simulation to the bottom line
Polytech Center in Chiba, Japan, for his courtesy in
(Panel),” Proceedings of the 1990 Winter Simulation
referring to the flexible manufacturing system. In
Conference, Institute of Electrical and Electronics
addition, the author wishes to thank Mr. T. Fujii for his
Engineers, Piscataway, New Jersey: .590-594.
assistance.
AUTHOR BIOGRAPHY
REFERENCES
SOEMON TAKAKUWA is a Professor of Economics
Baker, K. R., 1974. Introduction to Sequencing and
at Nagoya University in Japan. He received his B. Sc.
Scheduling, John Wiley & Sons, New York, New
and M. Sc. degrees in industrial engineering from
York: 213-233.
Nagoya Institute of Technology in 1975 and from Tokyo
Horngren, C.T. and G. Foster, 1987. Cost Accounting,
Institute of Technology in 1977 respectively. His Ph.D.
6th edition, Prentice-Hall, Inc., Englewood Cliffs,
is in industrial engineering from The Pennsylvania State
New Jersey: 815-817.
Univeristy. He has prepared the Japanese edition of
Cooper, R. and R. S. Kaplan, 1991. “Profit priorities
Introduction to Simulation Using SIMAN. He was
from activity-based costing,” Harvard Business
awarded by Japan Foundation for the Promotion of
Review, May-June 1991: 130-135.
Machine Tools their Prize in 1984, and by SCS,
Hitomi, K. 1979. Manufacturing Systems Engineering,
International for the Best Paper Award of “Modelling
Taylor and Francis Ltd., London: 230.
and Simulation 1992”. His research interests are in the
Kaplan, R. S. 1986. “Must CIM be justified by faith
area of optimization of manufacturing systems and
simulation

---

Document Outline

• THE USE OF SIMULATION IN ACTIVITY-BASED COSTING FOR FLEXIBLE MANUFACTURING SYSTEMS
• ABSTRACT
• 1 INTRODUCTION
• 2 THE FLEXIBLE MANUFACTURING SYSTEM
• 3 COST ACCOUNTING WITH SIMULATION
  • 3.1 Manufacturing Costs
  • 3.2 Cost Variance
  • 3.3 Simulation-Based Cost Accounting
  • 3.4 Activity-Based Costing
• 4 SOME IMPORTANT ASPECTS OF EFFICENCY AND COST ANALYSIS
  • 4.1 Product-Mix and Machine Loading
  • 4.2 Scheduling Effects
• 5 SEMI-GENERATIVE PROCEDURE OF ACTIVITY-BASED COSTING WITH SIMULATION
  • 5.1 The Procedure
  • 5.2 Cost Accounting for Various Types of Products
• 6 SUMMARY
• ACKNOWLEDGMENTS
• REFERENCES
• AUTHOR BIOGRAPHY

---