What are the Differences Between Simulation Software: Discrete, Continuous, and Agent-Based?

Simulation has become an integral part of many industries due to its capacity to provide insight into complex operations and processes. This post deals with the different types of simulation software applications, their capabilities, and application. Here, discrete event, agent-based, and continuous simulation will be defined and the differences across all options highlighted to help enterprises make easy decisions when choosing a simulation software.


Discrete event simulation (DES) models the operation of a system as a sequence of discrete events that occur in different time intervals. The discrete events occur at specific points in time thus marking the ongoing changes of state within the modeled system.

Continuous simulation (CS) models the operations of a system to continuously track system responses through the duration of the simulation. This means results are produced at every point during the simulation and not in intervals. Continuous simulations also produce data in instances where no ongoing changes occur.

Agent-based models (ABM) simulate the actions and interactions of individual agents within a system. The agents can either be a piece of singular equipment or a group of assets working towards a similar goal. ABM simulations are run to determine the effects of these agents on the functions of the entire system an agent is a part of.

An example that highlights the functions of these different simulation techniques is that of a simple check out point in a supermarket. A DES model will view the arrival of a customer and the moment the customer departs as two separate events while the time spent will be represented as a time lapse between both events. The continuous simulation will continuously count the number of customers passing through the checkpoint and its general effect on the checkout system. The ABM simulation sees the customer and checkout point as autonomous agents and tracks their effect on the entire sales process.

With this explanation, it is easy to note that DES technique models physical phenomena or reality excellently as it is able to track occurring events. The agent-based and continuous options are excellent at determining the behavioral pattern of a system. In many cases, a combination of the different simulation techniques provides more-rounded results, especially when modeling complex processes with diverse variables and events.

The Differences in Agent-based, Discrete Event, and Continuous Simulation Features

To highlight these differences a few criteria will be used. These criteria include the following features:

  • What they simulate – This refers to the models they are best at simulating
  • Time step – This refers to how the techniques view the passage of time and time intervals
  • Queuing – This refers to how queue flows are managed
  • Statistical details – This refers to how they define or evaluate events within a system

What they Simulate

Starting with DES, as stated earlier, DES software applications are used to simulate discrete events, needs, and requirements. Continuous simulations are generally applied to flowing continuous processes while ABM is applied to autonomous agents and systems.

Time Step

For DES software, the time step changes according to the occurrence of individual events while for continuous simulation, time steps basically remain unchanged. In AGM software apps, time steps change according to the changing interactions of the autonomous agent.


 DES software applies diverse techniques or systems to manage queues. This includes the use of a first-in-first-out (FIFO) approach or the last-in-first-out (LIFO) approach to managing queues. Continuous simulation software makes use of only the first in and first out system to manage queues. As for ABM, the management of queues is a bit different as it describes a system from the perspective of the agent. But a FIFO or LIFO system can be used to manage queues in ABM simulations.

The Differences in Application

Use cases provide realistic examples of defining or highlighting the differences encountered using these different simulation techniques. Starting with discrete event simulation, the discrete nature of this technique makes it an excellent choice for industrial simulations where events occur.

This includes the manufacturing industry, pharmaceutical production enterprises, plants, and industries with functional logistics systems. Here, the ability to simulate the arrival and departure of entities or queuing problems provide a level of insight into industrial operations in ways other methods cannot. An example is the use of Simio’s DES software to optimize activities within the Nebraska Medical Center. In this example, DES modeling was used to optimize hospital operations by reducing the travel time of surgeons and patients, as well as, the use of operating rooms across the medical facility.

Discrete event simulations are also powerful tools in capital intensive industries due to their ability to perform what-if analysis before pursuing further implementation initiatives. The experimentation capacity it brings to the table can save these enterprises from financial loses on specified business operations. The ability to also speed up or slow down specific phenomena to analyze expansive shifts or systems makes it a powerful tool for business applications.

Other application advantages DES software apps bring to the table include; its use as a training and validation tool in Industry 4.0, and its ability to kick start the digital transformation initiatives of enterprises.

Continuous Simulation Software – The continuous nature of this simulation technique makes it a unique tool for analyzing flowing processes or elements with non-linear relationships.

Continuous simulations are generally used within advanced engineering fields where simulator engines are designed. This includes the aviation industry for designing flight simulators, and autopilot programs. It is also used in designing gaming engines for video games such as the Nintendo Wii.

In industrial settings, discrete event simulation software applications are favored but continuous simulation is being used for generative design tasks and managing control systems in the pharmaceutical industry. It is also used in predicting or estimating the probability of natural phenomena such as the occurrence of flooding and hurricanes. These application examples mean continuous simulation is predominantly applied in STEM-related fields.

The advantages continuous simulation brings to the table include; the ability to describe systems with varying activities occurring within the same time interval. Continuous simulations are also used in enhancing artificial intelligence systems due to their theoretical analytical capabilities.

Agent-based Simulation Software – ABM models are generally used in the social sciences. It is extensively used to study interdependencies between different human activities, social and economic systems, and in facilities where the interactions between diverse systems define operations.

The three concepts that define the application of ABM are its flexibility, its ability to capture emergent phenomena, and its ability to define systems. With these abilities comes certain advantages such as the ability to integrate ABM simulations into DES or continuous simulation environments.

Its ability to simulate interactions between autonomous agents also makes it an excellent tool for understanding shop floor behavior. For example, can be used to analyze the cause of shop floor traffic across a facility where both humans and autonomous machines interact. Here, it’s individualistic approach to simulation provides different perspectives from active agents explaining the cause of phenomena such as an unexpected traffic jam within a system.

ABM is actively used in to monitor flowing process such as traffic and customer flow management within physical shops, parks, and recreational centers. An example is its use in a Macy’s store. In this example, ABM was used to estimate the distribution of sales people within its facility and how they interact with customers to enhance its operations.

It is also used to analyze stock market phenomena and operational risk within organizations in diverse industrial niches. Thus, highlighting the versatility and flexibility ABM simulations bring to diverse interactive processes.

In Summary

Simulation provides insight into human relationships, industrial processes, urban and regional planning, and complex systems across every niche. Thus cementing its status as a major data analytics and digital transformation tool designed for every organization.

Although DES, CS, and ABM simulations apply different approaches to simulation, the results they produce optimize human and industrial endeavors in different ways. These ways include planning and implementation, enhancing customer relationships, training staff, developing strategies, and design. The Simio modeling and Simulation software provides an intuitive platform for modeling, running, managing, and sharing DES, CS, and ABM simulations to optimize your organization’s operational processes. You can learn more about specific use cases by browsing through our catalog of case studies.

Evolution of Discrete Event Simulation Software

Today, discrete event simulation (DES) software and the benefits it provides are currently being used across a majority of industries to simplify business operations, make predictions, and gain insight into complex processes. But before modern simulation software such as Simio could be used to create shiny models and execute real-time simulations, there were earlier technologies that formed the foundation built upon by modern simulation software. As you can probably tell, there is definitely a story behind the evolution of simulation software and today, that story is being told.

To accurately tell this story, the evolution must be arranged in chronological order. The traditional order currently in use today is the order outlined by R.E. Nance in 1995. This chronological order will be used here but with slight edits to accommodate the earliest memories of simulation software and the current strides being made. This is because the most referenced order outlined in 1995, did not take into account the efforts of Jon von Neumann and Stanislaw Ulam who made use of simulation to analyze the behavior of neutrons in 1946.

RE. Nance’s chronology which was written in 1995 could not and did not account for the recent paradigm shifts in DES software. This understandable omission will also be highlighted and included in this post. Therefore, this paper on discrete event simulation should be seen as an update of the history and evolution of DES software.

The Early Years (1930 – 1950)

Before discrete simulation came to prominence, early mathematicians made use of deterministic statistical sampling to estimate uncertainties and model complex processes. This process was time-consuming and error-prone which led to the early DES techniques known as Monte Carlo simulations. The earliest simulation was the Buffon needle technique which was used by Georges-Louis Leclerc, Compte de Buffon to estimate the value of Pi by dropping needles on a floor made of parallel equidistant strips. Although this method was successful, simulation software as we know it got its origin in 1946.

Sometime in the fall of 46’, two mathematicians were faced with the problem of understanding the behavioral pattern of neutrons. To understand how neutrons behaved, Jon von Neumann and Stanislaw Ulam, developed the Roulette wheel technique to handle discrete event simulations. The light bulb moment came to Ulam while playing a game of Solitaire. Ulam successfully simulated the number of times he could win at Solitaire by studying hundreds of successful plays.

After successfully estimating a few games, he realized it would take years to manually observe and pick successful games for every hand. This realization led to Ulam enlisting Jon von Neumann to build a program to simulate multiple hands of solitaire on the Electronic Numerical Integrator and Computer (ENIAC). And the first simulation software was written.

The Period of Search (1955 – 1960)

The success of both mathematicians in simulating neutron behavioral patterns placed the spot light on simulation and encouraged government agencies to explore its uses in the military. As with all technological processes, the growth of discrete simulation software could only match the computing units available at that time. At that time, analog and barely digital computers were the springing board for development.

Around 1952, John McLeod and a couple of his buddies in the Naval Air Missile Test Center undertook the responsibility of defining simulation concepts and the development of algorithms and routines to facilitate the design of standardized simulation software. In the background, John Backus and his team were also developing a high-level language for computers. The efforts of the multiple teams working independently of one another led to the development of the first simulation language and software that would lead to the evolution of DES software.

It also highlights the general theme of how technological advancements and software evolutions occur which is through advancements in diverse interrelated fields.

The Advent (1960 – 1965)

By 1961, John Backus and his team at IBM had successfully developed FORTRAN, the first high-level programming language for everyday use. The success of FORTRAN led to the creation of a general-purpose simulation language based on FORTRAN. This language was SIMSCRIPT which was successfully implemented in 1962 by Harry Markowitz.

Other general-purpose simulation software and systems also sprang up within this period as competing contractors continued to develop simulation languages and systems. At the tail end of 1965, programs and packages such as ALGOL, General Purpose Simulation System (GPSS), and General Activity Simulation Program (GASP) had been developed. IBM computers and the BUNCH crew consisting of Burroughs, UNIVAC, NCR, Control Data Corporation, and Honeywell were developing more powerful computers to handle complex simulations.

One of the highlights of this period was the successful design of the Gordon Simulator by IBM. The Gordon Simulator was used by the Federal Aviation Administration to distribute weather information to stakeholders in the aviation industry. Thus highlighting the first time simulation was used in the aviation industry.

Here again, the increase in processing speed and the prominent entry of a new term known as computer-aided design was to play a role in advancing the development of simulation software for use. At this stage, early simulation packages and languages were still being used predominantly by the government, as well as, a few corporations. Also, ease of use, intuitive, and responsive packages were slowly being integrated into simulation software such as the GPSS which had become popular in the 60s’.

The Formative Years (1966 – 1970)

The formative years were defined by the development of simulation software for commercial use. By this time, businesses had begun to understand simulation and the role it plays in simplifying business process and solving complex problems. The success of systems such as the Gordon Simulator also got industry actors interested in the diverse ways DES software could be employed.

Recognizing the need to apply simulation in industrial processes, the first organization solely dedicated to simulation was formed in 1967 and the first conference was held in New York at the Hotel Roosevelt. In the second conference, 78 papers on discrete event simulation and developing DES software were submitted. Surprisingly some of the questions asked in the 1968 conference still remain relevant to this day. These questions include:

  • The difficulties in convincing top management about simulation software
  • How simulation can be applied in manufacturing, transportation, human behavior, urban systems etc.

The Expansion Period (1971 – 1978)

 The expansion period was dedicated to the simplification of modeling process when using simulation software and introducing its use in classrooms. At this stage, diverse industries had begun to understand the use and benefits of simulation software to their respective industries. This, in turn, led to discussing the need to prepare students for a world that integrates simulation.

Also, advancement in technology such as the introduction and wide spread use of the personal computer made the case for developing simulation software for dedicated operating systems. This led to the development of the GPSS/H for IBM mainframes and personal computers. The GPSS/H also introduced interactive debugging to the simulation process and made the process approximately 20 times faster than previous simulation packages. In terms of technological evolution, the GASP IV also introduced the use of time events during simulations which highlights the growth in simulation software available to industries at that time.

By the fifth simulation conference tagged the ‘Winter Simulation Conference’ of 1971, diverse tutorials on using simulation packages such as the GASP2 and SIMSCRIPT had become available to the public. The growing popularity of simulation also led to increased commercial opportunities and by 1978, simulation software could be purchased for less than $50,000.

The Consolidation and Regeneration (1979 – 1986)

The consolidation age was defined by the rise of the desktop and personal computer which led to the widespread development of simulation software for the personal computer. Simulation software also witnessed upgrades through the development of simulation language for alternative modeling (SLAM). The SLAM concept made it possible to combine diverse modeling capabilities and obtain multiple modeling perspectives when handling complex processes.

These upgrades or development made simulation for production planning possible and the manufacturing industry began to take a keen interest in simulation software. The increase in computing and storage capacity also led to the creation of factory management systems such as the CAM – I. CAM – I effectively became the first simulation software used solely for closed-loop control of activities and process within shop floors.

By 1983, SLAM II had been developed and this was an industrial-grade simulation package ahead of its time. SLAM II provided three different modelling approaches which could also be combined at need. These approaches included discrete event modeling approach, network modeling, and the ability to integrate discrete modeling and network modeling in a particular simulation model. More importantly, SLAM II cost approximately $900 which was relatively cheap at that time. This can be signified as the moment where discrete event simulation came into its own as commercial software options for discrete event simulation modeling became available to the general public

The Growth and Animation Phase (1987 – 2000)

The 90s’ witnessed a consolidation of the strides made in the earlier years and many interrelated technologies and processes also came off age within this decade. This era focused on simplicity, the development of interactive user-interfaces, and making simulation available for everyone including non-technical individuals.

In the mid-nineties, simulation software was being used to solve even more complex issues such as simulating every event and process in large-scale facilities. The Universal Data System example was a first in those days. Universal Data System was stuck with converting its entire plant to a hybrid flow-shop which enhanced production. To achieve this, the company made use of GPSS and the end result was a successful flow that enhanced daily operations and the entire process was modeled and simulated within 30 days.

In 1998, vendors began to add data collection features to simulation software. These features included the automation of data collection processes, the use of 3D graphics and animation to make the simulation process more user-friendly and non-technical. Needless to say, the technological advancements in animation, modeling, graphics design, and UI building played roles in enhancing simulation software during this period.

The Flexibility and Scalability Phase (2000 – 2019…)

Finally, we come to the last evolutionary phase of the DES software as we know it. Once again, advancement in interrelated technologies have made scaling simulation and speeding up its process possible. The evolution that came with the new millennium saw DES vendors leverage the use of cloud computing, AI, and high-performance computing to take simulation to greater heights.

Other changes that came within these decades was the evolution of production-based scheduling process to a simulation-based scheduling process. This shift allowed for real-time simulation scheduling, processing, and decision-making. This shift also comes with the fourth industrial revolutions were data collection, automation, and interconnectivity rule. Simulation software of this generation has evolved to become tools capable of digitization and the development of digital twins.

Discrete event simulation software such as Simio are examples of the comprehensive simulation technologies that are needed to drive Industry 4.0. This is because new age DES software must be able to collect and store its own data, model accurate 3-D graphics, animation, manage real-time scheduling, and digitization. They must also be equipped with features that market it possible to leverage cloud computing, integrate enterprise resource planning applications, and high-performance computing. These features all work together to ensure the most complex simulations are executed to deliver accurate answers or insights when applied in professional settings.


The future of discrete event simulation is by no means set in stone as the experiences from previous eras have shown. This means with the advancement in interrelated technologies and simulation software, more industrial concepts and business models will be disrupted in the coming decade. 

Simio Partner Finalist in Franz Edelman Award

The prestigious Franz Edelman Award for Achievement in Operations Research and the Management Sciences was presented at the Edelman Gala on April 16th, 2018 in Baltimore, Maryland. The Franz Edelman competition honors distinction in the practice of Operations Research and Analytics, by both individuals and companies, with emphasis on the beneficial impact of their achievements.

To reach the finals, companies are required to demonstrate how their use of technology is transforming the approach to some of the world’s most complex problems.

Simio is proud to be the provider of the simulation that facilitated one of this year’s finalists, Europcar, through our partner, ACT Operations Research (ACTOR). Our combined technologies have been used to develop Opticar which provides forecasting, simulation and optimization of the processes relating to vehicle rental for the leading European car rental company.

Simulation for Decision Making

The vehicle rental industry is a huge, complex, constantly changing market, with cultural variations across countries. In order to meet dynamic demand, decisions are continuously made regarding fleet assets, their locations, usage and pricing. The combination of AI, statistical modeling and simulation allows all eventualities to be considered and evaluated in order to establish optimum processes and make informed decisions.

Simulation can be used to model the possibilities with respect to both capacity and revenue, helping managers of car rental companies to reduce their risks in terms of planning for optimal fleet saturation. By making quality decisions, they can constantly maximize business opportunity for the company and ensure consistent financial and service performances.

At Simio, we are constantly solving business problems of this kind through simulation. When complex system schedules and decisions are required, we deliver leading edge solutions across many industries, from manufacturing to transportation and logistics.

Simio is proud to congratulate our partner ACTOR, with Europcar, on their outstanding achievement of becoming a Franz Edelman Award finalist.

Optimizing the Smart Factory

In the same way that a product development involves prototyping, the production process for manufacturing that product should also be optimized for maximum efficiency and productivity.Discrete Event Simulation (DES) software approximates the manufacturing process into individual events, so can be used to model each step in manufacturing process for overall performance optimization.

The IT innovations of Industry 4.0 allow data collected from its digitalized component systems in the Smart factory to be used to simulate the whole production line using Discrete Event Simulation software.

Real time information on inventory levels, component histories, transport, logistics and much more can be fed into the model, developing different plans and schedules through simulation. In this way, alternative sources of supply or production deviations can be evaluated against each other while minimizing potential loss and disruption.

When change happens, be it a simple stock out or equipment breakdown or an unexpected natural disaster on a huge scale, Discrete Event Simulation software can produce models showing how downstream services will be affected and the impact on production. Revised courses of action can then be assessed and a solution implemented.

The benefits of using Discrete Event Simulation software to schedule and reduce risk in an Industry 4.0 environment include assuring consistent production where costs are controlled and quality is maintained under any set of circumstances.

Success in Simulation Introduction


We will be using this space to help each other become more successful in our simulation projects. We will be sharing information and initiating discussions that will prove interesting and helpful to both experienced and novice simulationists.

When I say “we”, it is because I cannot do this alone – I need active participation from the user community. Your comments, topic suggestions, and article submissions are welcome.

This blog is not about any particular product, nor is it intended to be in any way commercial or sales-oriented.

Success in Simulation is available to all simulationists in industry, service, military, academic and other application areas, as well as anyone who wants to become a simulationist or who just wants to learn more about simulation. While I intend to focus mainly on discrete event simulation, articles on the related fields of agent-based modeling, system dynamics, and emulation will also be included.

The articles will be intentionally be kept short for a quick read, and hopefully most will be written in an informal style for easy reading. I encourage you to subscribe to the RSS feed so that you will automatically receive new articles as soon as they are posted.

Dave Sturrock
VP Products – Simio LLC