Description
Simulates a service as entities moving through activities and queues over time, to test design and staffing changes before implementation.
Discrete-event simulation represents a process as a series of events that change system state: arrivals, service starts and ends, routing decisions. Entities (users, tickets, orders) flow through activities with stochastic durations and resource constraints. Unlike queueing formulas, DES handles complex routing, priorities, time-varying demand, abandonment and human behaviour rules. Rigour comes from validating the model against current performance, running multiple replications for confidence intervals, and documenting assumptions. Service designers can use DES to test new service blueprints, triage rules or self-service options before costly pilots.
Attributes
- Discipline
- Systems & Complexity
- Loss type
- Unmigrated
- Data
- Quantitative
- Effort
- Weeks
Steps
- Map the process with a service blueprint or process mining output.
- Collect input distributions: arrival patterns, task durations, routing probabilities, abandonment.
- Build the model in a DES tool (e.g. SimPy, AnyLogic, Simul8) and verify logic.
- Validate against historical performance metrics (waits, throughput) before experimenting.
- Run scenarios with multiple replications, warm-up periods and confidence intervals.
- Communicate results with animations and key metrics; pilot the best scenario.
Sources
- Law, Simulation Modeling and Analysis (5th ed., 2015)
- Robinson, Simulation: The Practice of Model Development and Use (2nd ed., 2014)
Also known as
- DES
- Process simulation
- Service simulation modelling
Categories
- Strategy & Collaboration
- Systems & Complexity