This document summarizes a case study conducted to improve the quality of service in the material receiving system at Tata Engineering & Locomotive Co. Ltd. in Jamshedpur, India. The study found that trucks were experiencing delays of 2.5 hours on average when delivering materials. By analyzing the process and conducting a simulation, researchers determined that segregating documents with errors into a separate queue and adjusting staffing schedules could reduce average wait times by 76.4%. Implementing these recommendations led to 61.5% of trucks experiencing wait times of 1 hour or less, down from only 9.3% previously.

1. Improvement of Service Quality
of
Material Receiving System
Soumyanath Chatterjee
&
Ashit Ghosh
Productivity Services Division
Tata Engineering & Locomotive Co. Ltd.
Jamshedpur - 831 010

2. ABSTRACT :
This paper is a Case study on how principles of Operations Research has been applied to
improve the Quality of Service in Material receiving Area.
In TELCO Jamshedpur, all Trucks arriving with material comes to CRS(Gate). At
CRS(Gate) these trucks are checked for existance of proper purchase Order and Delivery Point
through an On-line inventory control system. Then the trucks are given a "Priliminary Goods
Inward Note ". After this, laden weight of the trucks are taken and they are allowed to go to
store for unloading. After unloading material, unladen weight of the truck is taken and it is
allowed to go out.
This whole process used to take several hours to complete, causing lot of inconvenience to
the transporter. As around two hundred trucks comes in the Works every day for delivering
material the Management felt the need to improve quality of service in this area.
A study was conducted to find out the extent of the problem and the reason of delay. The
study brought out a small lacuna in the queuing system at the very beginning of the process. This
was getting magnified by the physical system and resulting in hours of delay at the end.
The system was suitably redesigned to eleminate those problems. A simulation study was
carried out to validate some of the assumptions made to design the new system.
BACKGROUND:
TELCO Jamshedpur manufactures Commercial vehicle Chasis and Excavator. It has over
fifteen stores spread over an area of six square kilometers. In most of the places stores are
situated adjacent to production shops. This makes the shops vulnerable to pilferage.
A central checking and control agency, CRS(Gate) was set up to check whether material being
brought in belongs to TELCO and the delivery point of these material. In addition this agency
also collects Excise Gate Pass for taking MODVAT credit. All trucks that enter in the works
has to report to CRS(Gate) . At the start of the study on an average it used to take 2.5 Hrs to
inward a truck at CRS(Gate) and 22.8% of the trucks were held up for more than 4 hours. In
totality a truck used to spend 20 Hrs. inside works on an average. This delay causes loss of
revenew to the transporter. Accumulation of large number of trucks inside works also causes
problems for movement inside works.
Top management felt the need to improve quality of service for the material receiving.

3. OLD SYSTEM: -
The system of operation at the begining of study was as follows:
1. Transporter's men deposit challan and other document at the CRS(Gate) counter.
2. At CRS(Gate) the documents are fed in computer and Priliminary GIN are printed. A
Movement Control Form with details of store location, item and quantity are manually
3. prepared for monitoring movement of truck. These are sent to security office.
4. After receiving documents trucks are called by security.
5. Security officer checks incoming material on a sampling plan
6. After checking, laden weight of trucks are taken and a weighment card is given to the
transporter.
7. Trucks then unload material in the store.
8. After unloading material transporter hands over road permit and PGIN to stores Incharge and
obtain signature on MCF.
9. After unloading material, the truck has to get unladen weighment done and then it goes
out after handing over a copy of MCF to security person.
DATA COLLECTION AND ANALYSIS:
In order to analyse causes of delay a sample of thirty seven trucks were taken and time spend
between various points were noted. Upon analysis percentage of delay was found to be: -
LOCATION %GE DELAY
At CRS(GATE) office 14.3
Queue between CRS & Security 6.1
Security checking 0.3
Weighment, travelling to first store 30.9
Movement between store 28.8
Unloading material in store 10.5
Exit weighment & checking 9.1
As weighment was found to consume the maximum time this place was observed physically.
Physical observation brought out the fact that at weighment point trucks comes irregularly. As
a result of this a long queue builds up at weigh bridge. Moreover trucks after unloading come
for unladen weighment. On further exploration it was found that documents are sent from
CRS(Gate) office in bundles. As a result calling of trucks and arrival of trucks at weigh bridge
was irregular. This analysis made it clear, CRS(Gate) was the controling factor in the whole
process. Further work was done in CRS(Gate) office system.
OLD SYSTEM OF CRS(GATE) :
At the time of study the transporters used to deposit all papers at a counter of CRS(Gate) . The
recieving clerk used to a give Seniority number after checking all documents and making entries
regarding truck number, arrival time, number of boxes etc. After this, the clerk used to send the
documents to terminal room for feeding into computer. At terminal room, terminal operator

4. would fill purchase order number and other details to print Priliminary Goods Inward note. In
case there was some error in Purchase Order number then the terminal operator will correct it
before printing PGIN. For non computerised material PGIN were made by typing and updated
on computer at a later date. Printed PGIN were passed on to supervisor for checking and
signature. After signature the bunch of PGIN were passed on to security for calling trucks.
Only Excise Gate pass used to be retained at CRS(Gate) office for further processing. Shift-wise
manning of the office was:
SHIFT TIME %GE STAFF
6:00 AM - 2:00 PM 12 %
7:00 AM - 4:00 PM 64 %
8:00 AM - 5:00 PM 24 %
Against this arrival and release pattern of the trucks were as follows :-
TIME INTERVAL ARRIVAL %GE RELEASE %GE
(CUMM) (CUMM)
6:00 - 7:00 10.1 2.5
7:00 - 8:00 22.4 5.0
8:00 - 9:00 40.2 16.0
9:00 - 10:00 57.0 30.0
10:00 - 11:00 71.0 45.0
11:00 - 12:00 80.5 50.0
12:00 - 1:00 85.8 55.0
1:00 - 2:00 94.0 70.0
2:00 - 3:00 98.5 80.0
3:00 - 4:00 99.7 95.0
4:00 - 5:00 100.0 100.0
Graph of arrival and release of trucks are given in Figure 2.
ANALYSIS OF THE SYSTEM:
It may be observed from the graph of Fig. 1. that arrival rate and service rates do not match.
Moreover these rates changes with time. At 6:00 AM when the office starts approximately ten
percent of the vehicles comes within one hour. Whereas only 2.5% of the vehicles are cleared.
This difference generate a queue right from the starting of the day.

5. As little work has been done in the field of "time-non-homogenious queue" graphical
analysis was done at first to understand the mechanism later on a simulation study was carried
out.
Graphically, total time spent by all trucks is the area enclosed by the arrival and release curve
in Fig. 2. for FCFS queue. From the graph it is clear that slow rate of release between 6:00 AM
to 8:00AM and 11:00 AM to 1:00 PM attributes for large increase of area enclosed by the
curves. This area can be reduced by 76.4% by maintaing steady release rate of 16 trucks per
hour which is already achieved by CRS(GATE). As a result the waiting time will also be reduced
by 76.4%. This proposed curve is shown in Fig 3.
In order to achieve this change of service rate Shift wise manning was changed as follows :-
SHIFT TIME %GE STAFF
6:00 AM - 2:00 PM 56 %
7:00 AM - 4:00 PM 16 %
8:00 AM - 5:00 PM 28 %
During study it was also found that alongwith manual documents some erroneous documents
also comes. Correcting these erroneous documents takes lot of time. It is a well known fact that
mixed queue takes more waiting time than queue with seperate channel for large and small
service time. However to validitate result of time-homogenious queue on time-non-
homogenious queue a simulation study was carried out.
SIMULATION:
The simulation was carried out to test the hypothesis - "In a multiple server situation
seggregating queue for slow processing and fast processing will result in reduced waiting time"
holds good for the time-non-homogenious queue prevalent at CRS(GATE) .
Following assumptions are taken for simulation: -
1. Arrival rate of trucks follows Poission distribution where
mean arrival rate is a function of time.
2. Service rate is time-homogenious and follows Gauessian
distribution.
3. Ten percent of trucks has erroneous documents which takes ten
minutes to process and it also follows Gauessian distribution.
4. There are seven server.

6. Two hypothesis were tested. These are as follows: -
HYPOTHESIS - I
In the first case both normal and erroneous documents were processed in the same queue.
Allocation of the trucks to the server was as follows: -
1. When a truck arrives it is first allocated to first server. If the first server is busy then it is
allocated to second and so on.
2. When all server are busy the truck waits in a queue and it is allocated to a server when it
becoms free.
HYPOTHESIS - II
In second case normal and erroneous documents are processed in different queue. On the
basis of equal utilisation six server are ear marked for normal document and one server is ear
marked for erroneous documents. In order to take care of excegency there is a provision of
overlapping amongst these. Allocation scheme to server is as follows :-
a) When a truck arrives with normal document then it is allocated to first server. If first
server is busy then it is allocated to second server and so on.
b) For erroneous document allocation is first made to seventh server then to sixth and so on.
c) In case all servers are busy the document is put on a queue. There are seperate queue for
normal and erroneous documents.
d) When a server becomes free documents are allocated from the queue. For server one to five
first preference is given from normal queue if normal queue is empty documents are alloted from
erroneous queue. Similarly for seventh preference is given for erroneous documents. When
queue for erroneous is empty then normal documents are allocated to seventh server.
With this logic the simulation is carried out for a period of thirty days.
Average waiting time for Case-I is 1.58 time units and for Case-II is 1.50 time units.
Distribution of waiting time for the two is shown in Figure-4. which shows superiority of Case-
II over Case-I. Flow-chart of the Simulation program is given in Figure-4(a).
RECOMMENDATION:
1. Shift-wise manning at CRS(GATE) office should be changed to -
SHIFT TIME %GE STAFF
6:00 AM - 2:00 PM 56 %
7:00 AM - 4:00 PM 16 %

7. 8:00 AM - 5:00 PM 28 %
2. All error transaction should be processed separately. Two terminal may be ear marked for
this. Document Allocation scheme to terminal will be as follows: -
a) When a truck arrives with normal document then it is allocated to first server. If first
server is busy then it is
allocated to second server and so on.
b) For erroneous document allocation is first made to seventh server.
c) In case all servers are busy the document is put on a queue. There are seperate queue for
normal and erroneous documents.
d) When a server becomes free documents are allocated from the queue. For server one to six
first preference is given from normal queue if normal queue is empty documents are alloted from
erroneous queue. Similarly for seventh server preference is given for erroneous documents.
When queue for erroneous is empty then normal documents are allocated to seventh server.
3. Administrative control should be excercised to reduce formation of bunches as far as
possible.
ACTUAL ACHIEVEMENT:
This study result was implemented in the year 1989-90. In the table below delay distribution of
trucks at CRS(Gate) is given for period before and after the study recommendations were
implemented.
DELAY IN UNITS BEFORE STUDY AFTER STUDY (1990)
OF TIME (1989)
0 - 1 9.3 % 61.5 %
1 - 2 30.2 % 24.6 %
2 - 3 18.6 % 7.7 %
3 - 4 2.3 % 1.5 %
4 - 5 7.0 % 0.0 %
5 - 6 9.3 % 3.1 %
6 and more 23.3 % 1.5 %
TOTAL 100.0 % 100.0 %
AVERAGE 3.26 1.18

8. Figure - 5 clearly shows the reduction in average waiting time and the shift of distribution
pattern.
However, it will be unwise to attribute the big improvement solely on th Operations Research
study. The actual improvement is a combined effect of greater motivation of the staff of
CRS(Gate), awareness of the total system and the implication of delays and recommend changes.
ACKNOWLEDGEMENT:
The authors wish to thank the top management of TELCO for the
continued support for improving the service quality of material
recieving function. Mr.M K Sengupta, AGM(Central Materials)
and Mr.B N Sahay, Asst. Manager CRS(Gate) and the staff of
CRS(Gate) for support in providing data and effective
implementation of the recommendations. Without their support this study would have been
impossible.
BIBLIOGRAPHY
1. Gordon Geoffrey; " System Simulation " [1980]; Prentice-Hall
(India)
2. Joseph J Moder and Salah E Elmaghraby; "Handbook of Operations
Research - Foundations and Fundamentals" [1978]; Von Nostrand
Reinhold Co.
3. Wagner H. M.; " Principles of Operations Research "[1969];
Prentice-Hall (India).

9. ANNEXURE - I
AREA ENCLOSED BY ARRIVAL AND RELEASE CURVE
==========================================
EXISTING CURVE :-
TIME ARRIVAL RELEASE HEIGHT OF
% % CURVE
06:00 - - 0%
07:00 10.1 2.5 7.6 %
08:00 22.4 5.0 17.4 %
09:00 40.2 16.0 24.2 %
10:00 57.0 30.0 27.0 %
11:00 71.0 45.0 26.0 %
12:00 80.5 50.0 30.5 %
13:00 85.8 55.0 30.8 %
14:00 94.0 70.0 24.0 %
15:00 98.5 80.0 18.5 %
16:00 99.7 95.0 4.7 %
17:00 100.0 100.0 0%
Area by Trapizoidal rule 210.7
PROPOSED CURVE :-
TIME ARRIVAL RELEASE HEIGHT OF
% % CURVE
06:00 - - 0%
07:00 10.1 10.1 0%
08:00 22.4 21.9 0.5 %
09:00 40.2 33.7 6.5 %
10:00 57.0 45.5 11.5 %
11:00 71.0 57.3 13.7 %
12:00 80.5 69.1 11.4 %
13:00 85.8 80.9 4.9 %
14:00 94.0 92.7 1.3 %
15:00 98.5 98.5 0%
16:00 99.7 99.7 0%
17:00 100.0 100.0 0%

10. Area by Trapizoidal rule 49.8
Reduction in area = (210.7 - 49.8) ö 210.7 x 100 = 76.4 %