Flow queue analysis co4.pptx business process management

Contents
1. Flow Analysis
2. Queuing Analysis
3. Simulation
4. Recap
SEITE 1
Chapter 7: Quantitative Process Analysis

Process
discovery
Process
identification
Process
analysis
Process
implementation
Process
monitoring
Process
redesign
Process architecture
As-is process
model
Insights on
weaknesses and
their impact
To-be process
model
Executable
process
model
Conformance and
performance
insights
Process Analysis in the BPM Lifecycle
Define Vision Develop Strategy Implement
Strategy
Manage Personnel Manage Assets
Management Processes
Core Processes
Support Processes
Manage Risk
Manage
Information
Procure
Materials
Procure
Products
Market
Products
Deliver
Products
Manage
Customer
Service
les for BPM lifecycle and process mining
C
1.5h
B
15h
D
E
2h
C D
A B E
A
3m
35h 30h
15m
10m
10min
5m
5m 10m
30m

Contents
1. Flow Analysis
2. Queuing Analysis
3. Simulation
4. Recap
SEITE 3

Process
model
Performance
of each
activity
Process
performance
Flow analysis
4

Refresher: Process performance measures
Process
perform
ance
Time
Cost
Quality

Processing
time
Waiting
time
Cycle
time
Common time-related measures
6
Time taken by
value-adding
activities
Time between start
and completion of a
process instance
Time taken by
non-value-adding
activities

Processi
ng Time
Cycle
Time
Cycle
Time
Efficienc
y
Cycle time efficiency
7

Flow analysis of cycle time
8
1 day
1 day
3 days
3 days
1 day
2 days
Cycle time = X days

Sequence – Example
9
• What is the average cycle time?
Cycle time = 10 + 20 = 30

50
%
50
%
90
%
10
%
Cycle time = 10 + (20+10)/2 = 25
Cycle time = 10 + 0.9*20+0.1*10 =
29
Example: Alternative Paths

Cycle time = 10 + 20 = 30
Example: Parallel paths

Example: Rework loop
100%
0%
Cycle time = 10 + 20 = 30
1%
99%
Cycle time = 10 + 20/0.01 = 2010
80%
20%
Cycle time = 10 + 20/0.8 = 35

Flow analysis equations for cycle time
CT = max(T1, T2,…, TN)
CT = p1*T1+p2*T2+…+ pn*TN
CT = T / (1-r)
CT = T1+T2+…+ TN
T
r
1 - r
T1
...
T2
TN
T1
...
T2
TN
p2
pn
p1
T1 ...
T2 TN

Flow analysis of cycle time
14
1 day
1 day
3 days
3 days
1 day
2 days
Cycle time = 1.25 + 3 + 3 + 1.4 =
8.65 days
20% 60
%
1/0.8 max(1,3) 3 0.6*1+0.4*2
80%
40
%

Flow analysis of processing time
15
2 hours
0.5 hour
2 hours
2
hours
0.5 mins.
Processing time = 2.5 + 3 + 2 + 1.4 = 8.9 hours
20% 60
%
2/0.8 max(0.5,3) 2 0.6*2+0.4*0.5
80%
40
%
3 hours
Cycle time efficiency = 8.9 hours / 8.65 days = 12.9%

Flow analysis: scope and limitations
 We have seen how to use flow analysis for processing & cycle time calculation
 Flow analysis can also be applied to calculate:
 The average cost of process instances (assuming we know the cost of each activity)
 Cf. Section 7.1.6
 The number of times on average each activity is executed
 Can be used to calculate the “unit load” of each task, the resource utilization of each
resource pool, and the theoretical capacity of an “as is” process
 Cf. Section 7.1.5
 But flow analysis has some fundamental limitations…

Limitation 1: Not all Models are Structured

 Cycle time analysis does not consider:
 The rate at which new process instances are created (arrival rate)
 The number of available resources
 Higher arrival rate at fixed resource capacity
 high resource contention
 higher activity waiting times (longer queues)
 higher activity cycle time
 higher overall cycle time
 The slower you are, the more people have to queue up…
 and vice-versa
Limitation 2: Fixed arrival rate capacity

Time spent
per
resource on
process
work
Time
available
per
resource for
process
work
Resource
utilization
Resource utilization
19
Resource utilization = 60%
 on average resources are idle 40% of
their allocated time

Resource
utilization Waiting
time
Resource utilization vs. waiting time
20
Typically, when resource utilization > 90%
 Waiting time increases steeply

 WIP = (average) Work-In-Process
 Number of cases that are running (started but not yet completed)
 E.g. # of active and unfilled orders in an order-to-cash process
 WIP is a form of waste (cf. 7+1 sources of waste)
 Little’s Formula: WIP = ·CT
  = arrival rate (number of new cases per time unit)
 CT = cycle time
Interlude:
Cycle Time & Work-In-Progress

Exercise
A fast-food restaurant receives on average 1200
customers per day (between 10:00 and 22:00). During
peak times (12:00-15:00 and 18:00-21:00), the restaurant
receives around 900 customers in total, and 90 customers
can be found in the restaurant (on average) at a given
point in time. At non-peak times, the restaurant receives
300 customers in total, and 30 customers can be found in
the restaurant (on average) at a given point in time.
1. What is the average time that a customer spends in the
restaurant during peak times?
2. What is the average time that a customer spends in the
restaurant during non-peak times?

Exercise (cont.)
3. The restaurant plans to launch a marketing campaign to attract more customers.
However, the restaurant’s capacity is limited and becomes too full during peak
times. What can the restaurant do to address this issue without investing in
extending its building?

Contents
1. Flow Analysis
2. Queuing Analysis
3. Simulation
4. Recap
SEITE 24

 Capacity problems are common and a key driver of process
redesign
 Need to balance the cost of increased capacity against the gains of
increased productivity and service
 Queuing and waiting time analysis is particularly important in
service systems
 Large costs of waiting and/or lost sales due to waiting
 Example – Emergency Room (ER) at a Hospital
 Patients arrive by ambulance or by their own accord
 One doctor is always on duty
 More patients seeks help  longer waiting times
 Should we increase the capacity from one to two doctors?
Queuing Analysis
Inspired by an example by Laguna & Marklund (2004)
25

If arrivals are regular or sufficiently spaced apart, no queuing delay
occurs
Delay is Caused by Job Interference
Deterministic traffic
Variable but
spaced apart
traffic
© Dimitri P. Bertsekas
26

Burstiness Causes Interference
 Queuing results from variability in processing times
and/or interarrival times
27

 The queuing probability increases as the load increases
 Utilization close to 100% is unsustainable  too long queuing
times
High Utilization Exacerbates Interference
28

 Common arrival assumption in many queuing and simulation
models
 The times between arrivals are independent, identically
distributed and exponential
 P (arrival < t) = 1 – e-λt
 This distribution is applicable when the next arrival (i.e. the
next creation of a case) does not depend on how long ago
the previous arrival occurred
 In other words, the creation of a case is independent of the creation of
other cases.
The Poisson Process
Inspired by slide by Laguna & Marklund (2004)
29

Negative Exponential Distribution
30

Basic characteristics:
  (mean arrival rate) = average number of arrivals per time unit
 m (mean service rate) = average number of jobs that can be handled by
one server per time unit:
 c = number of servers
Queuing theory: basic concepts
arrivals waiting service

m
c
© Wil van der Aalst
31

Given  , m and c, we can calculate :
 occupation rate: r
 Wq = average time in queue
 W = average system in system (i.e. cycle time)
 Lq = average number in queue (i.e. length of queue)
 L = average number in system average (i.e. Work-in-Progress)
Queuing theory concepts (cont.)

m
c
Wq,Lq
W,L
© Wil van der Aalst
32

M/M/1 queue

m
1
Assumptions:
• time between arrivals and
processing time follow a
negative exponential
distribution
• 1 server (c = 1)
• FIFO
L=r/(1- r) Lq= r2/(1- r) = L-r
W=L/=1/(m- ) Wq=Lq/=  /( m(m- ))
μ
λ
Capacity
Available
Demand
Capacity
ρ 

33
Inspired by a slide by Laguna & Marklund (2004)

m



r
*
c
Capacity
Available
Demand
Capacity
• Now there are c servers in parallel, so the expected
capacity per time unit is then c*m
W=Wq+(1/m)
Little’s Formula  Wq=Lq/
Little’s Formula  L=W
M/M/c queue
34

 For M/M/c systems, the exact computation of Lq is rather complex…
 Consider using a tool, e.g.
 http://www.supositorio.com/rcalc/rcalclite.htm
 http://queueingtoolpak.org/ (for Excel)
Tool Support
0
2
c
c
n
n
q P
)
1
(
!
c
)
/
(
...
P
)
c
n
(
L
r

r
m




 


1
c
1
c
0
n
n
0
)
c
/(
(
1
1
!
c
)
/
(
!
n
)
/
(
P











m



m


m

 
35

 Situation
 Patients arrive according to a Poisson process with intensity  ( the time
between arrivals is exp() distributed.
 The service time (the doctor’s examination and treatment time of a patient)
follows an exponential distribution with mean 1/m (=exp(m) distributed)
 The ER can be modeled as an M/M/c system where c = the number of
doctors
 Data gathering
  = 2 patients per hour
 m = 3 patients per hour
 Question
– Should the capacity be increased from 1 to 2 doctors?
Example – ER at County Hospital
36

 Interpretation
 To be in the queue = to be in the waiting room
 To be in the system = to be in the ER (waiting or under treatment)
 Should we increase the capacity from one to two doctors?
Queuing Analysis – Hospital Scenario
Characteristic One doctor (c=1) Two Doctors (c=2)
r 2/3 1/3
Lq 4/3 patients 1/12 patients
L 2 patients 3/4 patients
Wq 2/3 h = 40 minutes 1/24 h = 2.5 minutes
W 1 h 3/8 h = 22.5 minutes
37

 Can be used to analyze waiting times (and hence cycle times), but not cost or
quality measures
 Suitable for analyzing one single activity at a time, performed by one single
resource pool. Not suitable for analyzing end-to-end processes consisting of
multiple activities performed by multiple resource pools.
 These limitations are addressed by process simulation
Limitations of basic queuing models
38

Contents
1. Value-Added Analysis
2. Queuing analysis
3. Simulation
4. Recap
SEITE 39

 Versatile quantitative analysis method for
 As-is analysis
 What-if analysis
 In a nutshell:
 Run a large number of process instances
 Gather performance data (cost, time, resource usage)
 Calculate statistics from the collected data
Process Simulation
40

Process Simulation
Model the
process
Define a
simulation
scenario
Run the
simulation
Analyze the
simulation
outputs
Repeat for
alternative
scenarios
41

Elements of a simulation scenario
1. Processing times of activities
 Fixed value
 Probability distribution
44

 Fixed
 Rare, can be used to approximate case where the activity
processing time varies very little
 Example: a task performed by a software application
 Normal
 Repetitive activities
 Example: “Check completeness of an application”
 Exponential
 Complex activities that may involve analysis or decisions
 Example: “Assess an application”
Choice of probability distribution
47

Simulation Example
Exp(20m)
Normal(20m, 4m)
Normal(10m, 2m)
Normal(10m, 2m)
Normal(10m, 2m)
0m
48

Elements of a simulation model
 Fixed value
2. Conditional branching probabilities
3. Arrival rate of process instances and probability distribution
 Typically exponential distribution with a given mean inter-arrival time
 Arrival calendar, e.g. Monday-Friday, 9am-5pm, or 24/7
49

Branching probability and arrival rate
Arrival rate = 2 applications per hour
Inter-arrival time = 0.5 hour
Negative exponential distribution
From Monday-Friday, 9am-5pm
0.3
0.7
0.3
9:00 10:00 11:00 12:00 13:00 14:00
35m 55m
50

 Fixed value
3. Arrival rate of process instances
4. Resource pools
51

Resource pools
 Name
 Size of the resource pool
 Cost per time unit of a resource in the pool
 Availability of the pool (working calendar)
 Examples
 Clerk Credit Officer
 € 25 per hour € 25 per hour
 Monday-Friday, 9am-5pm Monday-Friday, 9am-5pm
 In some tools, it is possible to define cost and calendar per resource,
rather than for entire resource pool
52

 Fixed value
3. Arrival rate of process instances and probability distribution
4. Resource pools
5. Assignment of tasks to resource pools
53

Resource pool assignment
Officer
Clerk
Clerk Officer
Officer
Syste
m
54

Process Simulation
Model the
process
Define a
simulation
scenario
Run the
simulation
Analyze the
simulation
outputs
Repeat for
alternative
scenarios
✔ ✔ ✔
55

Output: Performance measures & histograms
56

Process Simulation
Model the
process
Define a
simulation
scenario
Run the
simulation
Analyze the
simulation
outputs
Repeat for
alternative
scenarios
✔ ✔ ✔
✔
57

Tools for Process Simulation
 ARIS
 Bizagi Process Modeler
 ITP Commerce Process Modeler for Visio
 Logizian
 Oracle BPA
 Progress Savvion Process Modeler
 ProSim
 Signavio + BIMP
58

BIMP – bimp.cs.ut.ee
 Accepts standard BPMN 2.0 as input
 Simple form-based interface to enter simulation scenario
 Produces KPIs + simulation logs in MXML format
 Simulation logs can be imported to the ProM process mining tool
59

BIMP Demo
60
https://www.youtube.com/watch?v=TjXl6yASCSc

 Stochasticity
 Data quality pitfalls
 Simplifying assumptions
Pitfalls of simulation
61

 Simulation results may differ from one run to another
 Make the simulation tiemframe long enough to cover weekly and seasonal
variability, where applicable
 Use multiple simulation runs
 Average results of multiple runs, compute confidence intervals
Stochasticity
62

 Simulation results are only as trustworthy as the input data
 Rely as little as possible on “guesstimates”
 Use input analysis
 Deriver simulation scenario parameters from numbers in the scenario
 Use statistical tools to check fit the probability distributions
 Simulate the “as is” scenario and cross-check results against actual observations
Data quality pitfalls
63

 That the process model is always followed to the letter
 No deviations
 No workarounds
 That there is no multi-tasking (the same resource performs multiple tasks
concurrently) nor batching (tasks being accumulated and performed in a single go)
 That resources work constantly and non-stop
 Every day is the same!
 No tiredness effects
 No distractions beyond “stochastic” ones
Simulation assumptions
64

Contents
1. Value-Added Analysis
2. Queuing analysis
3. Simulation
4. Recap
SEITE 65

 Assuming we have performance measures for each activity in a process, flow
analysis allows us to calculate the following performance measures for an “as is”
process:
 Cycle time, processing times, cycle time efficiency of a process
 Average cost per process instance
 It can also be used to calculate the theoretical capacity of an “as is” process and
the resource utilization of resource pools
 But it is not suitable for “what if” analysis
 Queing analysis is a suitable technique for “what if” analysis of waiting times and
cycle times, suitable for analyzing individual activities performed by one resource
pool
 Simulation is a versatile technique for “what if” analysis of entire processes,
covering waiting times, cycle times, and costs.
 Particularly useful for identifying bottlenecks
SLIDE 66
Recap

Flow queue analysis co4.pptx business process management

More Related Content

Similar to Flow queue analysis co4.pptx business process management (20)

Recently uploaded (20)

Flow queue analysis co4.pptx business process management