Management of Waiting lines

(Queueing models)

Alok Raj
PODS Area
Office: Room No 14, 2nd Floor, Library Building, Tel. 3439
Email: [email protected]
Examples of Queue
Concerns
Have you noticed that when you have nothing to do that
times seems to pass less quickly?

Medical Emergencies – Waiting too long for a critical

surgery or treatment can worsen a patient’s condition.

We can easily spend four to five years of our life waiting in

lines
Queue
❖ A queue is simply a line
✔ Many processes require that people or items wait for service or
processing

❖ We are interested in modeling such processes to answer

questions about
✔ How much time is spent in line?
✔ What proportion of time is a server idle?
✔ What is the average number of people/items waiting?
✔ What is the probability distribution of the number of customers
or of the waiting time?
Queueing Theory
❖ Queueing theory is the mathematical study of waiting
lines
❖ It uses queuing models to represent the various types
of queueing systems that arise in practice.
❖ Queues arise when the short-term demand for service
exceeds the capacity
Most often caused by random variation in service
times and the times between customer arrivals.
If long term demand for service > capacity the queue
will explode!
The Practical View of Waiting Lines
A Cost/Capacity Tradeoff Model
Cost

Total cost
Cost of service

Cost of waiting

Process capacity
 Queueing system
Server

Arriving Jobs Departure

Customer in
service
Parameters/Characteristics of a queueing system:
Arrival Characteristics
Waiting Line Characteristics
Service System Configurations
Arrival Characteristics
Size of the calling population
❖ Finite or infinite

Arrival Pattern of Customers

❖ Predictable
❖Random
• Generally, follows a Poisson distribution
Arrivals can be measured as the arrival rate or the inter-arrival time
❖ Average Arrival Rate (λ): Defines rate jobs arrive to the server.
❖ Mean Inter-arrival Time (1/λ): Average time between successive job arrivals
Arrival Characteristics
❖ Balking: The customer may decide not to enter the
queue upon arrival, perhaps because it is too long.

❖ Reneging: The customer may decide to leave the

queue after waiting a certain time in it.

❖ Jockeying: If there are multiple queues in parallel the

customers may switch between them.
Waiting Line Characteristics
Waiting lines
❖ Limited
❖ Unlimited

Queue discipline: the rule by which customers in

the line receive service
❖ First Come First Serve (FCFS)
❖ Last Come First Serve (LCFS)
❖ Service in Random Order (SIRO)
Service System Configurations
• Number of Service Stages
o Single or multiple
Service Facility Characteristics
Service time distribution
❖ Predictable
• often machine controlled
❖ Random
• Randomly distributed according to a negative exponential probability
distribution

• Make sure that the observations fit the assumed

distributions when applying these models.

❖ Mean Service Time: Average time required to serve a job.

❖ Average Service Rate (µ): The average rate at which jobs are
served.
Queueing models nomenclature

(M / M / s) : (FCFS / K / P)
Arrival Time Dist.
Service Time Dist.
No. of servers.
Service Order
Waiting Space
Calling Population

(M/M/1) : (FCFS/∞/∞)
(M/M/m) : (LCFS/∞/P)
(M/G/1) : (FCFS/K/∞)
Assumptions of M/M/1 Queuing model
•Single queue single server system
•Infinite customer population
•No balking or reneging
•Arrivals are independent of each other but rate is constant
over time
•Arrival rate follows a Poisson distribution (random)
•Arrivals are served on a FIFO basis
•Service times are variable and independent but the average
is known
•Service times follow a negative exponential distribution
(random)
•Average service rate is greater than the average arrival rate
•The system operates under a steady state condition.
 Basic Queuing Model
❖ Arrival process ~ Random(Poisson)

❖ Service time ~ Random(Exponential)

❖ Server ~ Single

❖ Service Order ~ FCFS

❖ Waiting space ~ Unlimited

❖ Calling population ~ Infinite

❖ Nomenclature: M/M/1:FCFS/∞/∞
Arrival and Service Pattern
 Performance Metrics for single server system
❖ Number of jobs in the system: Those in queue plus one being
served.
Average number of units in the system (Ls)

❖ Number of jobs in queue: No. of jobs waiting

Average number of units in the queue (Lq)

❖ Time in System: = tdept-tarrival

Expected waiting time per customer in the system (Ws)

❖ Waiting time: Time in queue

Expected or average waiting time per customer in the queue (Wq)
Single-Channel Single-Stage Model, Poisson Arrivals, Exponential Service

Times (M/M/1)

Waiting Service rate μ

Line
Arrivals Server
at rate λ Customer Exits
the System

Lq
Customer getting
served
L
Single-Channel Single-Stage Model (M/M/1)
Single-Channel Single-Stage Model (M/M/1)
Problem 1: Part A
Western National Bank is considering opening a drive
through window for customer service. Management
estimate that customers will arrive at the rate of 15 per
hour. The teller who will staff the window can service
customers at the rate of one every three minutes.
Assuming Poisson arrivals and exponential service find
Utilization of the teller
Average number in the system
Average waiting time in line
Average waiting line in the system including service
Problem 1: Part B
Because of limited space availability and a desire to
provide an acceptable level of service, the bank manager
would like to ensure with 95 percent confidence, that no
more than three cars will be in the system at any time.
What is the present level of service for the three-car limit?
What level of teller use must be attained and what must be
the service rate to ensure the 95% level of service?
Problem 2
A vending machine supplier provides food and beverages to a large Indian
university. However, students frequently mishandle or kick the vending
machines out of frustration, leading to frequent breakdowns. On average, three
machines break down per hour, and these breakdowns follow a Poisson
distribution.
❖ Downtime costs the company INR 2000 per hour per machine due to lost
sales and repair expenses.
❖ Each maintenance worker is paid INR 1200 per hour.
❖Maintenance capacity: One worker can repair five machines per hour,
following an exponential distribution. Two workers together can repair
seven machines per hour, exponentially distributed. A team of three workers
can repair eight machines per hour, exponentially distributed.
Question: What is the optimal number of maintenance workers to
minimize total costs while ensuring efficient machine repairs?
Problem 3
In the bustling city of Nagpur, Mehra Retail, a rapidly growing e-commerce business, operates
a call center that takes orders for its popular online mall. During the peak festive
season—Diwali and New Year sales—the call center faces a surge in incoming calls from
eager customers looking to place orders.
Current Situation at the Call Center Call Arrival Rate: average time between call arrivals is 0.5
minutes, but due to high variability, the standard deviation is 0.203 minutes.
Service Rate: Each operator takes an average of 4 minutes to complete an order, but sometimes
calls take much longer, with a standard deviation of 2.5 minutes (suggesting some extreme
outliers).
Current Workforce: The call center employs nine operators to handle these calls.
Problem: Customers often complain about long waiting times before their calls are answered.
Management wants to estimate: How long customers wait before being served under the current
setup. The impact of hiring an additional operator—would it significantly reduce customer
waiting times?
If customers wait too long, they may hang up and abandon their orders, leading to lost sales and
frustration. On the other hand, hiring too many operators could increase costs without enough
return.
Approximating Customer Waiting Time
All you need is the mean and standard deviation to
approximate waiting time
“Quick and dirty” mathematical approximation
No assumption needed about a service distribution
Need data on service times
Need data on time between arrivals
Interarrival time
Impact of Pooling: Economies of Scale
Here is an interesting question: Does combining the two
systems into a single system with one waiting area and
two (m) identical servers lead to lower average waiting
times ?
We refer to such a combination of multiple resources into
one “mega-resource” as pooling.
Impact of Pooling: Economies of Scale
For simplicity, assume that both of them have a customer
arrival stream with an average interarrival time a of 4
minutes and a coefficient of variation equal to one. The
processing time p is three minutes per customer and the
coefficient of variation for the service process also is equal
to one.
Solution
The pooled process can serve the same number of
customers using the same processing time (and thereby
having the same utilization), but in only half the waiting
time!

The intuition for this result is as follows. The pooled

process uses the available capacity more effectively, as it
prevents the case that one resource is idle while the other
faces a backlog of work (waiting flow units). Thus,
pooling identical resources balances the load for the
servers, leading to shorter waiting times
Anylogic Modeling