Contributed on September 22, 2014

7.1K

3.4K

Business process simulations: from GREAT! to good, Razvan Radulian, Sept 2013

Description

This was my presentation at the World Congress for Project Managers and Business Analysts in Orlando (2013). While the title/teaser was simply playing with our fascination with the book "From good to great", the subject was very serious and pragmatic: using process simulations to learn, understand, explore, and ultimately decide.

Transcript

Business Process Simulations
From GREAT! to GOOD
Razvan Radulian, MBA
Independent Consultant/Coach/Trainer
Why-What-How Consulting, LLC
W5HY.com
Why don’t we just jump right into it?
i.e. Let’s skip the part when we sell our services…

PICTURE
THIS…
TriAgilus (fictitious spin-off subsidiary of large global CRO*)
Challenge: Reduce RFP+Start-up processes from 30 to 15-days
Some of the MANY variables:
• Huge stakes:
• Sponsor: $600K to $8M loss per delayed-day
• CRO: WIN (if profitable) or DIE (if miscalculating operational
timelines and/or costs)!
• Large number & diversity of players
• Many AND very complex processes
• Aggressive competition
• Global operations
• Complex Regulatory environment
* CRO: Clinical Research Organization
How would we (YOU) handle that?
Could it be… “Hmm” …?!?
Lets’ try another approach…
“Yeap, of course,
WHY NOT?”
Can it be… …?
WHY SIMULATION(S)?
• Reasons (past)
• Purposes (future)
WHY: Reasons (past)
The world without Simulations:
• Complexity
• Uncertainty
• Interdependencies
• Myths of the No-Sim world…
WHY: Purposes (future)
• What are we trying to achieve?
• Understand & Communicate
• Validate the Process Model(s)
• Exploration & Experimentation
• Facilitating Decisions:
• Design choices
• Manage Operations
• Rule out the UNEXPLAINED
• Seek to explain the UNEXPECTED
WHY/WHEN NOT?
• Alternative methods that are…
• Better
• Cheaper
• Easier to implement/execute
Unclear
Goals
Can’t Model
Better
Alternatives
• Not sure why we [want to] do it…
Reality
• Risks too high to tinker with
something that looks like “reality”,
BUT is not
• More… SIMULATIONS MIGHT NOT BE
THE BEST OPTION!
WHAT IS IT?
Part of a larger process…
Discover
Goals &
Objectives
Model
Optimize
Analyze Simulate
WHAT IS IT?
Types of Simulation
Process Validation
Structural Analysis
Time Analysis
Resource Analysis
Calendar Analysis
Operations
Management
Design
Exploratory
Areas of Use
Types of Analyses
WHAT IT IS NOT
Related but not the same…
Not the
Real
“Thing”
Process
Modeling
Pure
Mathematics
Pure
Testing
Guess-work
WHAT WE FEAR IT IS
“Some” of the fears/concerns…
Simulation vs. Alternatives
Experimenting with the Real Thing/Process:
• Do we have a/the REAL THING?
• Technically/operationally possible?
• Convenient? Any distractions?
• Practical? Costs, resources…
• Too risky? Can we control it?
• Do we have enough time?
• Any processes too long to even try it?
• Politically/strategically possible?
• Any possible misinterpretations?
• Can competitors find out about it?
Simulation vs. Alternatives
Using Mathematical Models:
• Possible?
• Do we have the Knowledge?
• Do we have the “Rocket scientist”?
• Do we have the Data?
• Too complex?
• Any moving parts?
• Any complex interdependencies?
• Predictable? Any uncertainties?
• Certain about the… uncertainties?
• Can we model the data behind the model?
Simulation vs. Alternatives
• Guessing (“What do we THINK it would happen?”)
• Possible?
• Can we tell if we are right?
• How risky (if we are wrong)?
• Who knows enough (for smart guesses)?
• Any differences of opinion?
• How many chances do we have
(before we get it right)?
Myths
• Limited applicability
• Too complex
• Too expensive (to setup and run)
• Too expensive (tools)
• Only highly specialized people (the SIM-experts) can do it
• There are ALWAYS easier ways to do it (using other methods)
• Simulations are THE ONLY WAT to do it…
HOW
• Process Modeling
• Scenarios & What-If Analyses
• Data
• Multi-level Analysis
• Standards & Best Practices
• Tools
Process Model(s) [BPMN]
Core Process Elements:
• Activities:
• Tasks
• Sub-Processes
• Decisions/Gateways:
• Exclusive work (either A or B)
• Parallel work (A and B)
• Optional work (A and/or B and/or C)
• Events:
• Start + Trigger
• End or Terminate
• Intermediate (e.g. Delays, Messages,
etc.)
• Interrupting (e.g. Timers, Exceptions,
Cancellations, etc.)
Process Modeling (2)
Work coordination:
• Orchestration (within Processes)
• Collaboration (between Processes)
Multi-Levels of Analysis
Process Validation:
• Deadlocks
• Infinite loops
• Branches that are never used
• Logic faults (e.g. unexpected work items/tokens)
Time & Throughput Analysis:
• Processing & Cycle Times
Resources & Constraint Analysis:
• Resource Capacity & Over/Under Utilization
• Delays (due to waiting on available resources)
Calendar:
• Resources availability by Calendar Periods
Simulation Data
Inputs:
• Historical data
• Estimates (constant)
• Probabilistic/Statistical
(distributions)
Simulation Data
“Real-Time” (accelerated) data:
• Animation
• Items processed
• Items waiting (to be processed)
• Utilization Graphs
• Etc.
Simulation Data
Outputs/Results:
• Time Estimates
• Resource Estimates
• Capacity Estimates
• What-If Comparisons
Scenarios & What-If Analyses (“2 or 3 Nurses?”):
Times…
Scenarios & What-If Analyses (“2 or 3 Nurses?”):
Resources…
Standards and Best Practices
• BPMN: http://www.bpmn.org/
• BPSim: http://www.bpsim.org/
• Patterns and Anti-Patterns
Tools
Many available tools:
• Commercial
• Open-source*
But, be aware…
… look for GOOD tools, not “just” GREAT!
* The tool and some of the examples used in this presentation were from Bizagi (Bizagi.com)
DEMOS & EXAMPLES
• “Simple” case: Patient Assistance Process*
• Complex cases: just imagine (see next slide)
* Source: Bizagi.com
LEVEL 0: BPMN VALIDATION
What are we looking for?
• BPMN Specification Violations
LEVEL 1: PROCESS STRUCTURE, ANTI-PATTERNS
What are we looking for?
• Deadlocks
• Infinite loops
• Branches that are never used
• Logic faults (e.g. unexpected work items/tokens)
LEVEL 2a: TIME ANALYSIS
Assumptions:
• Unlimited resources
• No Costs associated with either Resources or
Activities
What are we looking for?
• Cycle-Time
• Processing Time
LEVEL 2b: TIME + COST ANALYSIS
Assumptions:
• Unlimited resources
• Variable-Costs associated with Resources (e.g. hourly
rate)
• Fixed-Costs associated with Activities (e.g. materials)
What are we looking for?
• Cycle-Time & Processing Time (L2a)
• Costs incurred while/to perform the process
LEVEL 3: RESOURCE ANALYSIS
Assumptions:
• Level 2 assumptions +
• Limited Resources
• Resources: either People or other kind of Resources
What are we looking for?
• Level 2 analysis results +
•Waiting times (delays) due to Resources not being
available
• Resource Over and Under-Utilization
LEVEL 4: CALENDAR ANALYSIS
Assumptions:
• Level 3 assumptions +
• Resources Availability varies by Calendar Periods (e.g.
Shifts)
What are we looking for?
• Level 3 analysis results +
• Optimizing Costs & Resource Utilization
• Identify Critical Time/Calendar Periods
LEVEL 5: SCENARIOS & WHAT-IF ANALYSIS
Assumptions:
• Level 4 assumptions +
• Some scenarios may work better (or worse) than others
What are we looking for?
• Level 4 analysis results +
• Optimization:
• Compare various alternate scenarios
• Chose BEST alternative/option
NOW,
IMAGINE
THIS…
“Remember us?”
TriAgilus (fictitious spin-off subsidiary of large global CRO*)
Challenge: Reduce RFP+Start-up processes from 30 to 15-days
Some of the MANY variables:
• Huge stakes:
• Sponsor: $600K to $8M loss per delayed-day
• CRO: WIN (if profitable) or DIE (if miscalculating operational
timelines and/or costs)!
• Large number & diversity of players
• Many AND very complex processes
• Aggressive competition
• Global operations
• Complex Regulatory environment
* CRO: Clinical Research Organization
SO…
What does all this mean to you?
Business Process Simulation and…
• Business Analysis
• Risk Management
• Project Management
Some ADVISE…
• Start SMALL
• Start SIMPLE
• EXPERIMENT, SHARE, INTERPRET
• ADJUST
• Add MORE
• Add COMPLEXITY
• REPEAT
Q&A
THANKS & CONTACT INFO
Mr. Razvan Radulian
… better know as Razz:-)
razvan.radulian@why-what-how.com or razvan@w5hy.com
LinkedIn: whywhathow
Twitter: @w5hy
Blog: w5hy.com