Lezione 07 - Business Process Modeling PDF

Business Process Modeling - System Dynamics and Agents Cambiamento Organizzativo e Progettazione dei Processi Aziendali Laurea Magistrale in Ingegneria Gestionale Matteo Vignoli, University of Bologna E' vietata la copia e la riproduzione dei contenuti e immagini in qualsiasi forma. E' inoltre vie...

Business Process Modeling - System Dynamics and Agents Cambiamento Organizzativo e Progettazione dei Processi Aziendali Laurea Magistrale in Ingegneria Gestionale Matteo Vignoli, University of Bologna E' vietata la copia e la riproduzione dei contenuti e immagini in qualsiasi forma. E' inoltre vietata la redistribuzione e la pubblicazione dei contenuti e immagini non autorizzata espressamente dall'autore o dall'Università di Bologna. 18 Ottobre, 2024  Strategy Mapping (or casual loop diagrams) 2 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Strategy Map: Capture a Cause Effect Relationship from the Bottom Up Profitability Innovation Customer Satisfaction New Customers Customer acquisition Loyalty Market Share 3 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Key Benefits of Strategy Maps Articulates how the organization creates value for its constituents and legitimizing authority Displays key priorities and relationships between outcomes (the "what") and performance enablers or drivers (the "how") Provides a clear view of "how I fit in" for sub-organizations, teams, and individuals © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Strategy Maps to Communicate Strategy Executive consensus and accountability Building the map eliminates ambiguity and clarifies responsibility. Educate and Communicate Build awareness and understanding of organization strategy across the workforce. Ensure Alignment Each sub-unit and individual link their objectives to the map. Promote Transparency Communicate with and educate constituents, partners, oversight bodies, and the general public. © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna Source: "Using Balanced Scorecard Technology to Create Strategy-Focused Public Sector Organizations", Robert S. Kaplan, April 21, 2004, pg. 20  Modeling Decision Making Processes System Dynamics Source: ILOG, Marcello La Rosa 6 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Il pensiero sistemico e la system dynamics The System Dynamics (SD) is an approach to the study of Social Systems Behavior – It can be applied to companies seen as Dynamical Complex Systems – emphasizes the entanglement of Political Decision-making structures and Delays Objective → emphasize the relationship between the structure of a system and the behavior generated by the decisions taken inside. Main points: – interrelationships rather than linear chains of cause and effect; – processes of change rather than snapshots. 7 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  What is is System Dynamics The neo-classical theory of the firm Single decision maker. The rational choice theory Completely rational decision maker. Profit-maximizing firm. Choose the optimal combination of two fundamental resources: capital Behavioral theory: and labor. Many decision-makers (hierarchy); Negotiation and compromises; Bounded Rationality Management The new theories of the firm theories: The theory of bounded rationality Complex utility function Resource Based Theory Source: Edoardo Mollona Combination of various resources. System 8 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna Dynamics  What is it System Dynamics. Structure of a decision-making business process Dynamic phenomena behavior Flow variables Turnover Stock variables Decision (Resources) Information Motivation Source: Edoardo Mollona tempo 9 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Feedback loops System state action to Feedback change the Information loop status of the system More elementary feedback loops can be connected to form complex architectures that describe the variety of dynamic phenomena that affect Source: Edoardo Mollona businesses. 10 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Connected Feedback Loops Fonte: Edoardo Mollona 11 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles VHS User base + What behavior R would you expect? + Incentives to useVHS time Betamax User base + What behavior R would you expect? Source: Edoardo Mollona Incentive + to use Betamax time 12 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles What these processes have in + common? a Equilibrium state R + b time Acceleration and Source: Edoardo Mollona Reinforcement 13 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles (economic) systems with unstable equilibrium Accelerators of growth Source: Edoardo Mollona positive Externalities Bandwagon effects 14 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles Desired temperature= 10° Difference Actual temperature = 15° 15° between desired and - actual temperature 10° What behavior would you expect? B time Temperature + adjustments Difference Desired temperature = 10° between desired and Actual temperature = 5° - actual temperature 10° What behavior B would you expect? Source: Edoardo Mollona 5° time Temperature + adjustments 15 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles What these processes have in common? - a Equilibrium state B + b time Balance Source: Edoardo Mollona 16 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles (economic) systems with stable equilibrium Limits to growth Control mechanisms (budget) Source: Edoardo Mollona 17 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Casual Loop diagrams In a causal loop diagram, variables are connected by arrows showing the causal influences among them, with important feedback loops explicitly identified. Causal loop diagrams are good for quickly capturing your hypotheses about the causes of the dynamics in the system, and communicating it to others (Sterman, 2000). Borshchev, Andrei. The Big Book of Simulation Modeling: Multimethod Modeling with AnyLogic 6 18 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Casual Loop diagrams Work force recruitment clients 1 Financial resource Available budget orders Source: Edoardo Mollona 19 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Casual Loop diagrams Productive capacity pressures Work force orders recruitments + + + + clients 2 Delivery delays 1 - + orders dropout + Available budget + Financial resources + Negative customer satisfaction Source: Edoardo Mollona 20 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles Orders Source: Edoardo Mollona time 21 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles Orders 'reinforcement’ action of circuit 1 ‘balance’ action of circuit 2 + - Source: Edoardo Mollona Time 22 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles ‘balance’ action of circuit 2 orders + a - R + a + b B 'reinforcement ’ action of b - circuit 1 + Source: Edoardo Mollona time 23 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles Sector offer - Equilibrium price B + Incentives to enter actual production the sector capacity + Desired + Time delay production capacity + adjustment 24 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles Actual production capacity in the sector Equilibrium prod. cap. 25 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  System Dynamics principles Negative feedback loop with delay + a B equilibrium b - Limits to grow time + + a R b B c + - time 26 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Structure of system dynamics models There are two types of variables that appear in the SD models: Level: describe the state of a system as a collection of past actions. Flow: collect the information arising from the variable level and contain the information to change the status of the latter. Level Flow Action Information Decision 27 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  New Product diffusion  Case Study Consider a company that starts selling a new consumer product. The addressable market has a known size, which does not change over time. Consumers are sensitive to both advertising and word-of-mouth. The product has an unlimited lifetime and does not need replacement or repeated purchases. A consumer needs only one product. We are to forecast the sales dynamics. Borshchev, Andrei. The Big Book of Simulation Modeling 29 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  30 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  31 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  32 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  33 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  34 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Agent Based Modeling  Agent Based Modeling Simulation composed of one or more classes of agents Each agent corresponds to one or more autonomous entities in the simulated domain Agents have behaviors, often defined by a set of simple rules (computational models of behavior) Agents can communicate with environment and with each other Complexity emerges from simplicity 36 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Agents Agents have: – Representation of its internal data memory (or state) – Ability to change its own internal data representation (perceptions) – Ability to modify their environment (behavior) Types of Agents: – Plants and animals in an ecosystem (Boids) – Vehicles in the traffic – people 37 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  What is it an Agent Based Model? "Boids" are simulations of the behavior of flocks of birds (Reynolds 1987) Three rules of individual behavior – Separation: Avoid colliding with other birds – Alignment: Tip to the mean of the heads of other birds – Cohesion: Move towards the center of the flock The results closely resemble the actual behavior of flocks of birds, schools of fish, etc.... 38 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Why Agent Based Modeling formal – Hypothesis stripped flexible – Cognitively: agents can be "rational" or "adaptive" negotiable – Easier to cope with the complexity (non-linearity, discontinuity, heterogeneity) generative – Makes it easy to create new hypotheses 39 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Problems with Agent Based Modeling Models are too simple – Could be solved analytically (Axelrod, 1984) – The analytical solution is always preferable Models are too complicated – It is not possible to understand the causality (Cederman 1997) Problems Code – Many more lines of lines of code than in a demonstration Analysis of data – In which part of the parameter space is the solution? 40 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Agent Based Modeling Approach 1. Write the model 2. Solve analytically as much as possible 3. Justify the simulation 4. Use the real world to "refine" the model 5. forecast 6. Check weather with reality 7. Perform comparative statistics with parameters of the real world to verify the causality 41 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Statechart A statechart is a visual construct that enables you to define event- and time-driven behavior of various objects. Statecharts consist of states and transitions. A state can be considered as a “concentrated history” of the object and also as a set of reactions to external events that determine the object’s future. The reactions in a particular state are defined by transitions exiting that state. Each transition has a trigger, such as a message arrival, a condition, or a timeout. When a transition is taken (“fired”) the state may change and a new set of reactions may become active. State transition is atomic and instantaneous. Arbitrary actions can be associated with transitions and with 42 entering and exiting states. © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  A laptop running on a battery 43 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna Epidemic model  Case Study As an example we will build an agent based model of the spread of contagious disease. Here is the problem statement: Consider a population of 10,000 people. They live in an area measuring 10 by 10 kilometers, and are evenly spread throughout the area. A person in the area knows everybody who lives within 1 kilometer of him, and does not know anybody else. 10 random people are initially infected, and everybody else is susceptible (none are immune). If an infectious person contacts a susceptible person, the latter gets infected with probability 0.1. Having been infected, a person does not immediately become infectious. There is a latent phase that lasts from 3 to 6 days. We will call people in the latent phase exposed. The illness duration after the latent phase (i.e. the duration of the infectious phase) is uniformly distributed between 7 and 15 days. During the infectious phase, a person on average contacts 5 people he knows per day. When the person recovers, he becomes immune to the disease, but not forever. Immunity lasts from 30 to 70 days. We are to find out the epidemic dynamics -- namely, the number of exposed and infectious people over time. 45 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  46 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna Appendix A Project Mars  Project Mars The group of the Project Mars is owned by a company that is dedicated to the development of aerospace software and consists of IT technicians. Each employee has a great deal of autonomy in relation to the quantity of hours worked each week. The employees also have freedom to distribute these hours between three basic tasks. On one hand, the production of IT programmes to deal with current projects and on the other hand the analysis of IT tools – training in order to improve productivity. As this sector is very dynamic, it is important to dedicate a lot of hours to the improvements of productivity because technology can quickly become obsolete. Every week, the employees receive comparative information between the real production and the desired production they develop. Based on the difference of production perceived, pressure is applied to vary the production. This pressure can be understood in terms of an immediate adjustment to the working hours in production. Pressure is also applied to improve the productivity as a key aspect. In this way it is intended that more hours are planned for productivity. © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna 48  How to work more and bette Conflict between short and long term goals – Predictive maintenance – Quality Control Policy – Training vs productivity “nobody ever gets credit for fixing a problem that never happened” Reppening Sterman 2001 Time lag between actions and results 49 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Causal Diagram 50 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Flow Diagram 51 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Additional informations employee usually dedicates 35 hours a week to production tasks and 5 hours to tasks that improve productivity The desired production is 3500 units The productivity per working hour is 100 units/week It is estimated that 20 weeks of delay exists between the hours dedicated to improve the productivity and in the time when this is manifested. PRODUCTIVITY IS STABLE 52 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Model Equations Desired production: = 3500 Actual Production = Productivity*Working hours in production Productivity = +Improvements-Obsoleteness Initial value= 100 Difference in production = Desired production – Actual Production Working hours in production = Pressure to vary the production Initial value: 35 Hours in improvement of productivity =Pressure to improve the productivity Initial value: 5 Pressure to improve productivity = Difference of production / 500 Pressure to vary production = Difference of Production / 100 Improvements = DELAY 3I (Hours in improvements of productivity, 20, 5) Obsolescence = 5 53 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Simulation Results 54 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Simulation Results Desired production = 3500+step(1000,10) 55 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Link between short and long term 56 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Equations Hours in improvements of productivity = Pressure to improve productivity – pressure to vary the production/4 Working hours in production = pressure to vary the production – (Hours in improvement of productivity – 5)/5 57 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna  Result 58 © Matteo Vignoli - ALMA MATER STUDIORUM Università di Bologna