Operations Analytics

PROF SANDESH AKRE

Predictive Analysis

• Use of Data Mining, Statistics, Modelling, Machine Learning, Artificial

Intelligence to analyze the data
• Creating Data Model using Statistics, Probabilities and Trends
Process
Time

Reporting Predictive
Data Monitoring
Analysis Analytics

Happened? Happening? will happen?

Business Process and Modelling

Evaluating Creating
• Data Analysis &
Visualization
• Hypothesis
• Statistics
• Monitoring

Validating Testing
Some Predictive Models
• Linear Regression

• Logistic Regression

• Neural Network

• Support Vector Machine

• Naïve Bayes Model

• K-Nearest Neighbour Model

• Decision Tree
Business Applications

• Supply Chain • E-Commerce

• Customer Profiling • Quality Control

• Revenue Performance
• Pricing
• Fraud & Crime
• Human Resource
Detection
• Renewable Energy • Healthcare
• Financial Services • Education
Measuring Forecast
Accuracy
MODULE 3
Important

• A model which fits the data well does not necessarily forecast well.
• A perfect fit can always be obtained by using a model with enough
parameters.
• Over-fitting a model to data is as bad as failing to identify the systematic
pattern in the data
Forecast accuracy measures

• Mean absolute error: MAE = mean(|ei |),

• Root mean squared error: RMSE =

• Mean absolute percentage error: MAPE = mean(|pt |)

• Tracking Signal, Used to pinpoint forecasting models that need adjustment

Service Analytics

MODULE 4
Definition

Most same as other Analytics

improve, extend, and personalize the service provided

how new value is created for both the provider and the
customer.
Service Analytics Method

Waiting line Single & Multi-Server Simulation (Customer Cost Optimization

System Service efficiency)
Structure of Waiting Line Systems

Buffer
Inventory as
Boundry

Population Source in Queuing System (Q. What could be number?)

Arrival Process

Waiting Area, Service Area (Channel)

Queueing Models Notation

The two main inputs for any queue system are:

• λ (lambda): average number or arrivals per time period (i.e. mean arrival
rate)
• µ (mu): average number of customers served per time period (i.e. mean
service rate)
Standard Notations

• A represents the probability distribution for the arrival process

• B represents the probability distribution for the service process

• c represents the number of channels (servers)

• k represents the maximum number of customer allowed in the queueing

system
• m represents the maximum number of customers in total
WLS Costs and Management
Strategies

• Cost of Waiting

• Cost of Service

• Cost of Scheduling

• Cost of losing
WLS Operating Characteristics

1. the probability that no customers (or units) are in the system,

2. the average number of customers in the lines,

3. the average number of customers in the system (customers in line plus those being served,

4. the average time a customer spends in the waiting line,

5. the average time a customer spends in the system (waiting time plus time in the service facility,

6. the probability that an arriving customer has to wait for service,

7. the probability of n customers in the system, where n could be any real integer such as 1 customer, 2, 3, ...
example

A document clerk earns $15/hour, processes an average of 5 documents per

hour throughout the workday, and receives documents at the rate of 4 per
hour. There is a cost of waiting for the documents (by an office paralegal) of
$25.00 per hour. For this problem, the mean or average service rate is mu = 5
documents per hour, and the mean or average arrival rate is lambda = 4
documents per hour.
example

The law firm is interested in knowing the operating characteristics of this

waiting line system, such as average time a document spends in the waiting
line, number of documents in the waiting line, the utilization rate of the
document clerk, the probability of 3 documents in the system, and the cost of
the system.
1. The probability of an idle system (no
documents in the system, P0):
- find probability of atleast one documents in the system
- take inverse of it

P0 = 1 - (lambda/mu)
= 1 - (4/5) = 0.20
There is a 20% chance that the system will be idle at any one point in time.
2. The average number of documents
in the waiting line (Lq):
• Lq = lambda2/ [mu(mu - lambda) ]

42 / [5 (5-4)] = 3.2 documents

3. The average number of units in the
system (L):
• L = Lq + (lambda / mu)

= 3.2 + (4/5) = 4 documents

4. The average time a document
spends in the waiting line:
• Wq = Lq / lambda

= 3.2 / 4 = 0.8 hours or 48 minutes.

5. The average time a document
spends in the system:
• W = Wq + (1/mu)

= 0.8 + (1/5) = 0.8 + 0.2 = 1 hour

6. The probability that an arriving unit
has to wait for service:
Pw = lambda / mu

= 4/5 = 0.80
7. The probability of n documents in
the system
• Pn = ( lambda/mu)n * P0

For n=3 ?
3
= (4/5) * 0.2 = 0.1024

Calculate for Pn < 3 = P0 + P1 + P2 + P3

8. Costs of the Waiting Line System

Total Cost = Cost of Waiting + Cost of Service

= ( cw L ) + (cs k )
where
k = number of channels
L = number of documents
cw = cost of waiting
cs = cost of service
Optimal Stocking Policies

Module 5
Discussion on Supply Chain &
Consignment Stock
• supply chain management (SCM) deals with planning, designing, executing,

and controlling of people, resources, information, and activities pertaining to

the transformation and movement of raw materials to customers

• consignment stock (CS) is inventory that is not paid for until sold or used.

The customer may return the unsold or unused items to the vendor at any
point of time
Need of Optimization in Consignment
Stock
to achieve the “best” solution relative to a set of prioritized criteria or

constraints. These include maximizing factors such as productivity, strength,

reliability, longevity, efficiency, and utilization

Types of Optimization techniques

• Continuous Optimization versus Discrete Optimization

• Unconstrained Optimization versus Constrained Optimization

• None, One, or Many Objectives

• Deterministic Optimization versus Stochastic Optimization

Classification of Optimization Problems

• Linear Programming

• Quadratic Programming
Linear Programming
the real relationships could be much more complex , but it can be simplified
into linear relationships
• In operation analytics, complex real life problems can be expressed as linear
programming problems
• Many algorithms of optimization problems operate by solving Linear
Programming problems as sub-problems
• Linear programming is still used in departments of planning, production,
transportation, technology etc
Quadratic Programming
Suitable when more than 2 constrains are involved
?
• Image and signal processing

• Optimization of financial portfolios

• Performing the least-squares method of regression

• Controlling scheduling in chemical plants

• Solving more complex non-linear programming problems

• Usage in operations research and statistical work

Solving Linear equations
Linear Optimization
# Company wants to maximize the profit of two products A & B, which are sold
at ₹50/- and ₹40/- respectively.
# There are 3600 resources available everyday
# Product A require 40 units and B requires 24 units
# Production time is 4 minutes and 8 minutes resp.
# Total working hours are 8 per day
* Find production quantity of each Product to maximise the profit.
Supply Chain Metrics

MODULE 6
What are Metrics?

• numbers and ratios a company tracks to measure how efficiently it delivers goods to customers

• assess how your supply chain is performing

• helps stakeholders stay on the same page and prioritize resources

• where to invest to achieve maximum returns

Cash-to-cash cycle time

• This metric tells you the length of time between when you pay suppliers for materials and when
customers pay for the final finished product. You want the cycle time to be as short as possible.
• Tracking this metric will help identify potential causes of cash flow issues. The most efficient
companies have cash-to-cash cycle times of less than one month.

Cash-to-cash cycle time = receivable days + inventory days – payable days

Customer order cycle time

• Customer order cycle time tracks the number of days between your company receiving a
purchase order and completing customer delivery. It helps measure the responsiveness of your
supply chain and how well you’re providing customer service

Customer order cycle time = actual delivery date – purchase order creation date
Supply chain cycle time

• Supply chain cycle time measures how long it would take to complete an order if inventory
levels were zero. This KPI provides an overview of the efficiency of your entire supply chain

Supply chain cycle time = time it takes to order and receive supplies + order fulfillment cycle time
Service rate

• The service rate measures the percentage of product orders that are delivered on time

Service rate = product orders delivered on time / product orders received

Perfect order index or perfect order delivery rate

• The perfect order index measures the percentage of your orders that are error-free from
beginning to end. That means the order was recorded correctly, shipped on time and in the right
quantities and arrived without damage

Perfect order = [(total orders – errors) / total orders] x 100

On-time delivery

• The percentage of orders that arrive as scheduled

On-time delivery = [(total orders – orders that do not arrive on time) / total orders] x 100
In-full delivery

• The percentage of sales orders that are delivered completely in the first shipment

In-full delivery = [(total orders – orders that aren’t complete or are incorrect in first shipment) / total
orders] x 100
Damage-free delivery

• The percentage of orders that are delivered without any damages

Damage-free delivery = [(total orders – orders that arrive damaged) / total orders] x 100
Accurately documented order

• Percentage of orders in which all documents relating to the order are accurate

Accurately documented order = [(total orders – orders without accurate documentation) / total
orders] x 100
Gross margin return on investment (GMROI)

• The gross margin return on investment measures how much money a company makes on a
specific inventory investment. Tracking this metric gives your company insight into which
inventory items are especially poor or especially good performers. In general, a GMROI of 200 to
225 is considered respectable

Gross margin return on investment = gross profit / [(opening inventory in the period – closing
inventory in the period) / 2] x 100
Total supply chain management cost as percentage of
sales
• Total supply chain management cost as percentage of sales is fairly self-descriptive. It measures
the total cost of your supply chain operations compared with your overall sales

Total supply chain management cost as percentage of sales = (total supply chain costs / total sales)
x 100
Supply chain cost per unit sold

• Supply chain cost per unit sold measures your supply chain costs compared with how many of a
given item your company sells

Supply chain cost per unit sold = supply chain costs for a product over period / number of units
sold in that period
Day sales outstanding

• Day sales outstanding measures how quickly you collect revenue from your customers. A low
day sales outstanding number means you are generating revenue more quickly, which ultimately
improves your cash flow

Day sales outstanding = (receivables / sales) x number of days in a period

Inventory metrics and KPIs

• Use these metrics and KPIs to measure how well your inventory operations are performing. It’s
worth periodically running an inventory analysis exercise to help find ways to fill customer
orders while keeping costs as low as possible
Inventory days of supply (IDS)
• IDS represents the number of days it would take a company to run out of inventory if it didn’t add to
its supply.
• It’s important to track this number to ensure your company doesn’t keep too much inventory on
hand, which ties up cash, but that it has enough to satisfy customer demand. That sweet spot is
sometimes called the par level.
• Inventory days of supply helps your company understand and maintain its par level

Inventory days of supply = (average inventory in a month / monthly product demand, in dollars) x 30
Days sales of inventory (DSI)

• Days sales of inventory calculates the average number of days that inventory remained in stock
over a certain period. It’s a measurement of how long it takes a business to sell the items it
makes or buys

Days sales of inventory = (ending inventory / cost of goods sold) x number of days in period
Inventory-to-sales ratio (ISR)

• The inventory-to-sales ratio compares the inventory you carry to your overall sales. It’s a
measurement of the financial stability of your company. The ratio is closely related to your
inventory turnover ratio

Inventory-to-sales ratio = inventory value / sales value

Turn-earn index (TEI)

• The turn-earn index combines inventory turnover and gross margin to evaluate your company’s
profits and use of inventory. In essence, the index recognizes that your company can get by with
less inventory turnover if it makes a lot of money on that inventory. But it needs more inventory
turnover if it realizes a lower margin on the inventory.
• Generally, most businesses will want a turn-earn index of at least 150

Turn-earn index = (inventory turnover ratio x gross profit percentage) x 100

Inventory turnover ratio (ITR)

• The inventory turnover ratio measures how often a company’s entire inventory is sold in a
specific period. What comprises a “good” inventory turnover ratio depends on the industry. But
in general, a lower inventory turnover ratio means a company may have excess inventory due to
lagging sales.

Inventory turnover ratio = cost of goods sold in period / [(opening stock in period – closing stock in
period)/2]
Inventory velocity (IV)

• nventory velocity is the portion of inventory that your company projects it will exhaust within the
next specified period. The metric helps your company set optimum inventory levels so that you don’t
carry too much inventory but retain enough to satisfy projected sales in the coming period.
• An inventory velocity of 60% to 70% is a solid benchmark, with up to 80% for fast-moving inventory
items. An IV above 80% is high and could lead to excess inventory, while inventory velocity below 60%
is low and could presage shortages

Inventory velocity = opening stock / upcoming period’s sales forecast

Months on hand

• Months on hand indicates how many months of inventory you have at your disposal if you
purchase no more stock while your sales continue as forecast

Months on hand = (average inventory for year / cost of goods sold for year) x 12
Stock rotations

• Stock rotations, also called stock life or stock coverage, refers to the number of days, on average,
that it takes for you to run out of your inventory stock. It is best to use the past 52 weeks to
figure this calculation. That allows for enough time to show seasonal and other variations. If you
have a low stock rotations number, you risk running out of stock and failing to meet customer
demand

Stock rotations = (average stock in a period / total sales in a period) x number of days in the period
Average payment period for production materials

• The average payment period for production materials is the average timespan between when
you receive materials and when you pay for them. By tracking this metric you can favor suppliers
that offer more favorable billing terms and thus improve your cash flow

Average payment period for production materials = (materials payables / total cost of materials) x
days in period
Supplier on-time delivery

• Supplier on-time delivery measures the percentage of time your suppliers deliver products to
you in the agreed-on timeframe

Supplier on-time delivery = (number of products supplier delivers on time in a period / total items
supplier ships in the period) x 100
Return reason

• The return reason measurement shows the top issues that cause customers to send items back.
This can help you assess weaknesses in your product lines or operations and is important to
track given the high cost of reverse logistics operations.
• The return reason measurement will quantify, in percentages, the top reasons customers send
goods back and display them in a chart that’s easily consumable by supply chain and business
unit leaders, who can then take action to fix problems
Order to cash

• While order to cash covers some processes outside of the supply chain, the performance of the
supply chain is an important component. Order to cash is a measure of the company’s entire system
to receive, process and complete orders. It covers the time from when a customer places an order to
when the company receives and records the customer’s payment.
• What constitutes a good order-to-cash ratio depends on the industry, and this metric takes into
account the efficiency of your order management, logistics and accounts receivable functions

Order to cash = date of customer order – date of receipt and recording of customer payment for order, after
receiving order
How can you use KPIs in supply chain analytics?

• Data analysis tools are the key to making KPIs actionable. With the right metrics and predictive
analytics, companies can increase forecast accuracy, keep inventory levels optimized, improve
logistics and shipping operations, maximize cash flow, and minimize customer returns.
• The goal of supply chain analytics is to find and fix inefficiencies in the supply chain, from raw
materials to finished goods and all points in between.
Balanced Score Card

MODULE 9
 BSC allows managers to look at the
What is BSC business from important perspectives

BSC includes measures that tell the

results of actions already been taken

BSC enables performance in several

areas simultaneously

BSC links performance measures

 Customer
Financial
Centric
Perspective
approach
BSC Links • Goals &
• Goals &
Measures
Measures

Internal
Innovation
Business
& Learning
Process
• Goals &
Measures • Goals &
Measures
 How do customers see us? (customer
The perspective)

Balanced
What must we excel at? (internal
Scorecard perspective)

Links Can we continue to improve and create

Performan value? (innovation and learning
perspective)
ce
Measures How do we look to shareholders? (financial
perspective)
Customer Perspective: How Do Customers
See Us?

Customer
Focus on Delivering
centric
Lead Time "Value"
Mission
Internal Business Perspective: What
Must We Excel at?
• Understand all sub-Micro process
• Fine tune Technological capabilities
• Individual level KPIs
• Suitable Info Systems
Innovation & Learning : Can We Continue
to Improve and Create Value?

Improvement in existing product & processes

Less Defects

No Missed Deliveries

Continuous Improvement

Stability ?
Financial Perspective: How Do We Look
to Shareholders?
• profitability, growth
• Shareholder Value Analysis
• Competitor Analysis
• New Market Segments, Products
Example