A

Project
on
Failure Mode Effect Analysis
of
Telecom Expenditure Management Process

Done by
Nalini Parthiban
DoMS, Anna University Chennai
 FMEA Analysis of Telecom Expenditure Management Process (TEM)

 Failure Mode Effect Analysis (FMEA) is a powerful quality tool adopted by

companies worldwide.

 The FMEA process begins by identifying "failure modes," the ways in which
product, service or process could fail.

 A key element of FMEA is analyzing three characteristics of failures:

1.How severe they are?
2.How often they occur?
3.How likely it is that they will be noticed when they occur?

 Risk Priority Number (RPN):

RPN = (severity) x (frequency of occurrence) x (likelihood of detection)

 The idea is to focus improvement efforts on the failures that have the biggest
impact on customers.
 The Project

 ‘FMEA analysis of Telecom Expenditure Process’ was a project undertaken

by Ms. Nalini Parthiban , a student of the 5th batch of ‘Six Sigma Green Belt
certification’ course.

 The project carried out at HCL Technologies-BPO services, Chennai was a

project focused on ‘Quality in the Service Set up’.

PURPOSE OF TEM

 Big companies which have lot of telecom lines cannot manage their
telephone expense.

 According to a research, it is seen that about 30 - 35% of companies

revenue is spent in telecom expense.
 A brief intro to TEM

ORIGIN OF HEMS

 Earlier, Bill Manager and Bill Tammer were the two companies who managed
Telecom expense.

 Later, CPS (Control Point Solutions) took over these two companies.

 Recently, HCL owned the CPS and formed as HEMS (HCL Expense
Management System).

TEM Process Flow

 Scanning
 Indexing
 Data Analysis
 Invoice Management
 Carrier Relations
 Auditing

Here FMEA analysis for the Data Entry process has been presented.
 Data Analysis Process

 Data Analysis
In this process Data from the invoice is analyzed and captured for further
processing.

The Following Data are to be found out from the invoice:

 BAN (Billing Account No. which is unique to each client),
 Invoice No.
 client and vendor info. etc.

 After finding, these details are entered into the particular fields in an application
specific to the company.

 The details are however entered manually which forms the major drawback of
this data entry process.
 The Project Flow

 The project was done using the Six Sigma DMAIC framework which is a very
powerful methodology by itself.

The objectives of the project were first defined as:

 To study the TEM process and identify key process parameters contributing to
process errors.
 Bring about Process improvement by error reduction.

As a part of the Measure phase, the errors were measured by the student.

 The possible failure modes in the process were identified & using the pre-existing
company database, the frequency of errors in each process parameter was
measured
 The Project Flow

FMEA for Process Analysis

 FMEA was used as the main quality tool to analyze the errors & boil down to
the most critical errors based on the RPN No. that could be focused on for
process improvement.

 The FMEA table was designed by identifying :

1.The possible effects/severity of failure.

2.The frequency of each error as determined in the measure phase.
3.Likelihood of detection for each of the errors.

 Severity, Occurrence & Detection rating scales were designed on a scale of

1-5
 Severity Rating Scale for Data Analysis Process

Rating Effect/Severity

5 Client incurs financial loss due to delayed/non-

payment of bills. Client cuts service

4 Client incurs financial loss due to payment of

incorrect amount. Client warns of detachment.

3 Client's reporting requirements affected. May result

in audit exceptions. Client highly dissatisfied.

2 No impact on clients but unnecessary escalations

(to the on sight team) & internal processing time

1 No effect.
Occurrence Rating Scale for Data Analysis Process

Rating Frequency of Occurrence of Errors

5 39 to 48

4 29 to 38

3 19 to 28

2 9 to 18

1 <=8
 Detection Rating Scale for Data Analysis Process

Rating Likelihood of Detection of Errors

5 Very low: Very low likelihood that current controls will detect or
prevent potential failure before reaching the next customer.

4 Low: Low likelihood that current controls will detect or prevent

potential failure before reaching the next customer

3 Moderate: Moderate likelihood that current controls will detect or

prevent potential failure before reaching the next customer.

2 High: High likelihood that current controls will detect or prevent

potential failure before reaching the next customer .

1 Very High: Very High likelihood that current controls will detect or
prevent potential failure before reaching the next customer .
A sample of the FMEA table for Data Analysis Process
 To illustrate a few Terms….

 Circuit Id/No : This is similar to the telephone No. for which the billing would
be done.

 MRC : Indicates errors while entering the Monthly Rental Charges. This is
the minimum fixed charge which has to be paid as rental irrespective of the
usage.

 Usage: This is the time duration for which the calls have been made

 OC&C : Other credits & charges is the adjustments which have to be

considered while calculating the current charges to be paid.

Ex: If the client has already paid an additional amount the previous month,
then this amt. has to be deducted while calculating the total amt. to be paid
for the current month. These adjustments are called OC&C

 There are roughly about 50 parameters that have been analyzed. Only a
sample has been presented here.
 Process Improvement
The improvement Phase

 After having identified the critical parameters through FMEA, certain

suggestions were given as shown in the FMEA table of which ‘The
Checklist’ idea given by the student was implemented for a period of 2
weeks.

 The Checklist was administered to a focus group of Freshers. It proved that

this helped them to do the much needed double checking of each
parameter.

 End result was that the error percentage of these new employees reduced
when they made use of the checklist .
The Sample Checklist
Other Quality Tools Used
A sample of the 5 Why Analysis done to identify the Root Cause
A sample of the cause & Effect Diagram

The 4M concept has been used

1.Men-Human Error
2.Machine-Software Errors
3.Materials-Invoice Related Errors
4.Methods-Process Issues
The Pareto Analysis of Data Analysis Process Errors based on
Frequency of Occurrence of Errors
 Other Suggestions
For the control Phase

Date Picker:
 It was observed that lot of errors are happening in the ‘date fields’. Hence, a
suggestion was made to use the date picker instead of entering the dates manually to
improve accuracy.

PDF to Excel conversion software

 A suggestion was made to convert the invoice image to excel sheet using the ‘PDF-
Excel Conversion’ software as conversion to excel for analysis simplification.

Reference Library:
 Another aspect which the operations department can focus on is preparation of
‘Reference Library’. Current Observation is that the analysts spend more time in
consulting their colleagues to clarify any process related doubts. Instead if a habit of
looking into the Reference library is inculcated, it will aid in conceptual strengthening
of the process.

 Currently this process in on & it this library is almost in ready-to-use status but this
has been done only for GA clients. A similar library containing all the updates needs
to be prepared for NJ clients as well.