1.

Hospital Emergency Department

A emergency department visit consists of a sequence of four steps. Each step is processed by one
or more healthcare professionals working in parallel.

I. Initial patient assessment takes 3 minutes per patient and is performed by 2 triage nurses.
II. Patient data entry (insurance and basic information verification) takes 6 minutes per patient
and is performed by 3 receptionists.
III. Consultation and treatment take 18 minutes per patient and is handled by 2 doctors.
IV. Patient discharge processing is performed by 1 discharge nurse. It takes 5 minutes per
patient.

Questions:

1. What is the theoretical flow time of the whole process?

2. What is the capacity rate of the whole process (patients per hour)?
3. What is the bottleneck resource?
4. Suppose the hospital is considering hiring one additional doctor.
o What would be the new process capacity?
o Would you recommend hiring the additional doctor? Briefly justify.

2. Loan Processing & Funding Process

Managers at Coastal Community Bank want to reduce the time it takes for approved loan
applications to be funded. Each activity is performed by a different employee unless stated
otherwise.

Approved loan applications follow the process below:

Step 1 — Document Intake & Sorting

A loan officer checks the application package for completeness and orders the documents.

• Processing time: 12 minutes per application

• Resources: 1 loan officer

Step 2 — Categorization
Applications are classified by loan type (personal vs. commercial) and by amount bracket.

• Processing time: 16 minutes per application

• Resources: 1 analyst

Step 3 — Parallel Activities (done at the same time)

After categorization, two activities start in parallel:
 • Step 3A: Credit & Risk Verification
o Processing time: 14 minutes per application
o Resources: 1 credit specialist
• Step 3B: System Data Entry
o Processing time: 9 minutes per application
o Resources: 1 admin clerk

Step 4 — Compliance Review (with possible rework)

A compliance officer checks for regulatory compliance.

• Processing time: 10 minutes per application

• Resources: 1 compliance officer
• Rework rule: 20% of applications are missing a required disclosure and must return
to Step 1 for correction. The correction takes an additional 6 minutes of loan officer time,
after which the application resumes at Step 4 (not Step 2).

Step 5 — Funding & Final Packaging

Once compliance is cleared, the application is packaged and sent for funding.

• Processing time: 8 minutes per application

• Resources: 1 funding coordinator

Questions

1. Draw a process flow map that clearly shows:

o the parallel activities in Step 3
o the rework loop in Step 4
2. Compute the theoretical flow time (ignore waiting; include the effect of parallel
processing, but ignore the rework probability for this part).
3. Compute the effective average processing time per application at Step 1 considering the
rework rule (i.e., incorporate the extra correction work caused by rework).
4. Determine the capacity rate of the overall process (applications per hour), taking into
account:
o parallel resources
o the rework rule’s impact on Step 1 workload
5. Identify the bottleneck resource and briefly explain why.

3. Hotel Room Service

When a guest calls room service, the room-service coordinator takes down the order. As soon as
the order is taken, three activities begin in parallel: the kitchen starts preparing the food,
the bar starts preparing the beverages, and a waiterprepares the cart and condiments.
The process consists of the following steps:

Step 1 — Take Order & Assign Waiter

• Processing time: 5 minutes per order

• Resources: 1 room-service coordinator

Step 2 — Parallel Preparation (starts after Step 1 and runs in parallel)

• 2A. Kitchen prepares the food

o Processing time: 20 minutes per order
o Resources: 2 kitchen staff working in parallel (each can handle one order at a
time)
• 2B. Bar prepares beverages (alcoholic or non-alcoholic)
o Processing time: 7 minutes per order
o Resources: 1 bartender
• 2C. Waiter prepares the cart (tray setup, cutlery, napkins, condiments)
o Processing time: 11 minutes per order
o Resources: 2 waiters working in parallel (each waiter can prepare one order at a
time)

Step 3 — Deliver to Guest

Once the food, beverages, and cart are all ready, a waiter delivers the order to the room.

• Processing time: 10 minutes per order

• Resources: 2 waiters (same waiters as in Step 2C)

Step 4 — Return & Close Order

The waiter returns to the station and closes the order in the system (billing/charging and
cleanup).

• Processing time: 6 minutes per order

• Resources: 2 waiters (same waiters)

Questions

1. Draw the process flow map (make sure to show the parallel structure in Step 2 and the fact
that Steps 2C, 3, and 4 share the same waiter resource pool).
2. Compute the theoretical flow time of the process for one order (ignore waiting).
3. Compute the capacity rate of the overall process (orders per hour).
4. Identify the bottleneck resource.
5. Management is considering hiring one additional waiter (so there would be 3 waiters total).
o What would be the new process capacity?
o Would you recommend hiring the additional waiter? Briefly justify.
 4. UBC Farm Saturday Market

The UBC Farm operates a Saturday morning produce market.

The counter officially opens at 9:00 AM, but customers begin lining up at 8:45 AM.

Customer arrivals follow this pattern:

• From 8:45 AM to 9:15 AM, customers arrive at a rate of 36 customers per hour.
• From 9:15 AM to 10:15 AM, customers arrive at a rate of 18 customers per hour.
• After 10:15 AM, arrivals drop to 8 customers per hour.

The checkout counter can serve customers at a rate of 24 customers per hour.

The counter continues operating until all customers have been served.

Assume:

• Continuous-time setting
• No balking or reneging
• First-come, first-served discipline

Questions

1. Draw the inventory build-up (queue length) diagram starting at 8:45 AM.
o Clearly show when the queue starts forming.
o Clearly show when it begins shrinking.
o Indicate the time when the line becomes empty again.
2. What is the maximum number of customers waiting in line during the morning?
3. Using the inventory build-up diagram, compute the average number of customers waiting
in line.
4. Using Little’s Law, compute the average waiting time per customer.
5. At what time will the system finally become empty?

5. Bagel Store
Consider a bagel store selling three types of bagels that are produced according to the process
flow diagram outlined below. We assume the demand is 180 bagels a day, of which there are 30
grilled cheddar & veggie, 110 veggie only, and 40 cream cheese.
Assume that the workday is 10 hours long and each activity is staffed with one worker. Moreover,
we assume the following activity times:
 Activity Time
Cut 3min/bagel
Grilled Cheddar 10 min/bagel
Veggies 5 min/bagel
Cream cheese 4 min/bagel
Wrap 2 min/bagel

Activity times are independent of which bagel type is processed at a resource (for example, cutting
a bagel takes the same time for a cream cheese bagel as
for a veggie bagel).
(a) Where in the process is the bottleneck for each kind of bagel? Where is the bottleneck for the
product mix? Is it possible to satisfy the demand for the
whole day?
(b) Given the product mix, how many units can the process produce within one hour?

Question 6 – Surgery Wait Times

The Ministry of Health reports the following statistics for a certain type of elective surgery in
British Columbia. The average number of patients currently on the waitlist is 520 patients. The
average waiting time until surgery is 78 days. The average number of surgeries performed is 5.5
per day. Assume the system is in steady state and that the waitlist includes only patients waiting
to receive surgery (not post-surgery recovery).

Using Little’s Law, determine whether the three reported numbers are internally consistent. If they
are not consistent, compute what the number of surgeries per day would need to be for the reported
average number of patients and waiting time to hold. Alternatively, if the reported surgeries per
day and average number of patients are correct, compute what the average waiting time would
need to be. Briefly explain two operational reasons why real-world reported statistics might not
perfectly satisfy Little’s Law even if each number appears reasonable individually.

The OM Triangle can be interpreted as an illustration of the P-K formula. Recall that the P-K
formula is

ρ" 𝐶#" + 𝐶$"

𝐼! = ⋅
1−ρ 2
where

Iq = inventory in queue,
ρ = long run utilization,
Ca = the coefficient of variation of inter − arrival time, and
Cs = the coefficient of variation of service time.

Discuss how the P-K formula allows us to arrive at an intuitive understanding about the
relationship between capacity, inventory, and information (the OM Triangle). You may limit your
answer to one paragraph.

Question 8 – Call Centre Under Demand Variability

You manage a call centre that handles technical support calls for a software company. On average,
48 calls arrive per hour. Each operator takes an average of 5 minutes per call. Currently, there are
4 operators working. You observe that arrival rates are highly variable during the day (coefficient
of variation of inter-arrival times is greater than 1), and service times are also quite variable
because some calls are simple while others require troubleshooting.

Management is considering two alternatives. The first option is to hire one additional operator.
The second option is to invest in training and standardized troubleshooting scripts that would
reduce the coefficient of variation of service times by 40%, without changing the average service
time.

Using the logic of utilization and the P-K formula, discuss how each option would affect the
average number of customers waiting in queue. Which option would you expect to have the greater
impact on waiting time, and why? Would reducing service-time variability reduce the need to hire
additional capacity? Provide a clear operational explanation rather than only a mathematical one.

Question 9 – Benny the Barber

Benny owns a one-chair barber shop. Based on past experience, customer arrivals follow a Poisson
process and haircut times are exponentially distributed. Benny’s market survey indicates that
customers arrive at an average rate of 2.4 customers per hour, and Benny takes an average of 22
minutes to complete a haircut. Assuming an M/M/1M/M/1 queue with first-come, first-served
service, compute (a) the average number of customers waiting in line, (b) the average time a
customer waits in line, (c) the average total time a customer spends in the shop (waiting plus
service), and (d) Benny’s utilization.

Question 10 – Help Desk with General Variability

A university computer lab operates a help desk that assists students with spreadsheet assignments.
On average, 18 students arrive per hour at the help desk. The standard deviation of inter-arrival
times is 2.5 minutes. The help desk can serve on average, 24 students per hour. The standard
deviation of service time is 2.0 minutes. Assume the system operates as a single-server queue with
general inter-arrival and service-time variability and use the P-K approximation approach
discussed in class.

Compute (a) the utilization of the help desk, (b) the average number of students waiting in line, (c)
the average number of students in the system, (d) the average time a student spends waiting in line,
and (e) the average time a student spends in the system.